JYVÄSKYLÄ STUDIES IN BIOLOGICAL AND ENVIRONMENTAL SCIENCE 181

Jarkko Routtu

Genetic and Phenotypic Divergence in virilis and D. montana

Esitetään Jyväskylän yliopiston matemaattis-luonnontieteellisen tiedekunnan suostumuksella julkisesti tarkastettavaksi yliopiston Ambiotica-rakennuksen salissa (YAA 303) elokuun 31. päivänä 2007 kello 12.

Academic dissertation to be publicly discussed, by permission of the Faculty of Mathematics and Science of the University of Jyväskylä, in the Building Ambiotica, Auditorium YAA 303, on August 31, 2007 at 12 o'clock noon.

UNIVERSITY OF JYVÄSKYLÄ

JYVÄSKYLÄ 2007

Genetic and Phenotypic Divergence in and D. montana

JYVÄSKYLÄ STUDIES IN BIOLOGICAL AND ENVIRONMENTAL SCIENCE 181

Jarkko Routtu

Genetic and Phenotypic Divergence in Drosophila virilis and D. montana

UNIVERSITY OF JYVÄSKYLÄ

JYVÄSKYLÄ 2007 Copyright © , by University of Jyväskylä ABSTRACT

Routtu, Jarkko Genetic and phenotypic divergence in Drosophila virilis and D. montana Jyväskylä: University of Jyväskylä, 2007, 34 p. (Jyväskylä Studies in Biological and Environmental Science ISSN 1456-9701; 181) ISBN 978-951-39-2890-2 Yhteenveto: Geneettinen ja fenotyyppinen erilaistuminen Drosophila virilis ja D. montana lajien mahlakärpäsillä Diss.

Genetic and phenotypic differentiation of Drosophila virilis and D. montana populations has proceeded along evolutionarily different pathways. D. virilis, “a human commensal”, does not show clear diversity or population structure in mtDNA haplotypes, which refers to population growth. Microsatellites group the laboratory strains of this into four clusters according to their geographical origin, indicating recent population differentiation. In contrast, D. montana, a “wild” species, has populations in Europe and North-America that are genetically diverged. In this species the mtDNA haplotypes are mixed among the two North-American populations, while microsatellites separate all study populations from each other. The courtship songs of the D. virilis laboratory strains showed significant inter-strain and geographic variation in several song traits. The genetic distances and the song divergence of the strains did not show significant association, which suggests that the songs have not diverged solely as a side-effect of genetic divergence. In D. montana the songs of the laboratory strains from different continents showed the highest divergence in song frequency, while the songs of the wild populations varied most prominently in the remaining pulse characters. D. montana populations also demonstrated divergence in male wing and genital morphology. The phenotypic divergence among populations did not coincide with the extent of their genetic divergence, suggesting that the first-mentioned traits are not evolving neutrally. Finally, we constructed a species recognition method for North European D. virilis group species to ease the identification of wild- collected and to detect possible misclassifications of laboratory strains.

Keywords: Genital morphology; male courtship song; natural selection; phylogeography; sexual selection; speciation; wing morphology.

Jarkko Routtu, University of Jyväskylä, Department of Biological and Environmental Science, P.O. Box 35, FI-40014 University of Jyväskylä, Finland Author’s address Jarkko Routtu Department of Biological and Environmental Science P.O. Box 35 FI-40014 University of Jyväskylä Finland [email protected]

Supervisors Prof. Anneli Hoikkala Department of Biological and Environmental Science P.O. Box 35 FI-40014 University of Jyväskylä Finland

Dr. Maaria Kankare Department of Biological and Environmental Science P.O. Box 35 FI-40014 University of Jyväskylä Finland

Reviewers Dr. Helena Korpelainen Department of Applied Biology P.O. Box 27 FI-00014 University of Helsinki Finland

Prof. Perttu Seppä Department of Biological and Environmental Sciences P.O. Box 65 FI-00014 University of Helsinki Finland

Opponent Prof. Leo Beukeboom Evolutionary Genetics Group Centre for Ecological and Evolutionary Studies University of Groningen P.O. Box 14 NL-9750 AA Haren The Netherlands CONTENTS

LIST OF ORIGINAL PUBLICATIONS ABBREVIATIONS 1 INTRODUCTION ...... 9 1.1 Genetic divergence ...... 10 1.2 Phenotypic divergence...... 11 1.3 Why to study Drosophila virilis group species?...... 12 1.4 Objectives of the project...... 14 2 MATERIALS AND METHODS ...... 16 2.1 Study species ...... 16 2.2 Data analysis ...... 16 2.2.1 Molecular data ...... 16 2.2.2 Phenotypic data ...... 17 2.3 D. virilis group species identification...... 18 3 RESULTS AND DISCUSSION ...... 19 3.1 Phylogeographic patterns...... 19 3.1.1 Mitochondrial haplotypes ...... 19 3.1.2 Microsatellites ...... 20 3.2 Phenotypic differentiation...... 21 3.2.1 Male courtship songs in D. virilis and D. montana ...... 21 3.2.2 Wing size and shape in D. montana...... 22 3.2.3 Genitalia size and shape in D. montana ...... 22 3.3 Genetic species identification in the D. virilis group ...... 23 4 CONCLUSIONS...... 24 Acknowledgements...... 26 YHTEENVETO (RÉSUMÉ IN FINNISH)...... 28 REFERENCES...... 29 LIST OF ORIGINAL PUBLICATIONS

The thesis is based on the following articles which are referred to in the text by roman numbers. I have participated in sequencing and analysing mtDNA for articles I and III and recording the songs for article II. In addition I took part in writing these articles. I have contributed significantly to planning, data collection, analysis and writing articles IV and V, where I am the first author.

I Mirol, P.M., Routtu, J., Hoikkala, A. and Butlin, R.K. (2007). Signals of demographic expansion in Drosophila virilis. Submitted manuscript.

II Huttunen, S., Aspi, J., Schlötterer, C., Routtu, J. and Hoikkala, A. (2007). Variation in male courtship song traits in Drosophila virilis: the effects of selection and drift on song divergence at intraspecific level. Submitted manuscript.

III Mirol, P., Schäfer, M.A., Orsini, L., Routtu, J., Schlötterer, C., Hoikkala, A. and Butlin, R.K. 2007. Phylogeographic patterns in . Mol. Ecol. 16 (5): 1085-1097.

IV Routtu, J., Mazzi, D., van der Linde, K., Mirol, P., Butlin, R.K. and Hoikkala, A. 2007. The extent of variation in male song, wing and genital characters among allopatric Drosophila montana populations. J. Evol. Biol. 20 (4): 1591-1601.

V Routtu, J., Hoikkala, A. and Kankare, M. (2007). Microsatellite-based species identification method for Drosophila virilis group species. Submitted manuscript. ABBREVIATIONS

AFLP amplified fragment length polymorphism AMOVA analysis of molecular variance ANOVA analysis of variance CN the number of cycles in a sound pulse DA discriminant analysis DNA deoxyribonucleic acid EFD an elliptical Fourier descriptor FRE the carrier frequency of the song FST genetic differentiation of neutral markers GPA generalized procrustes analysis IPI the interpulse interval LM a landmark mtDNA mitochondrial DNA OL an outline (pseudo)landmark PC a principle component PCA principle component analysis PL the length of a sound pulse PN the number of sound pulses in a pulse train PTL the length of a pulse train QST genetic differentiation of quantitative traits RAPD random amplification of polymorphic DNA RFLP restriction fragment length polymorphism

1 INTRODUCTION

The role of allopatry in speciation events has been demonstrated early on in evolutionary theory (Darwin 1859), and Mayr (1963) has especially emphasized its significance in speciation. Mayr (1954) showed, for example, that the mainland populations of New Guinean kingfishers do not exhibit subpopulation level differentiation, while the neighbouring geographically isolated islands harbour phenotypically differentiated subspecies. Maynard- Smith (1966) demonstrated that under special circumstances speciation may in theory occur in a sympatric situation, and Bush (1969) presented the first empirical evidence of the Rhagoletis fruit fly’s sympatric speciation on different host species. While the majority of empirical data supports allopatric speciation, there is now abundant evidence that both types of speciation events occur (Coyne & Orr 2004). Areas with complex topography that hinder migration are often hotspots for speciation. The Hawaiian islands and Amazonian rainforest are examples of areas where mountain ridges, the ocean and meandering rivers create obstacles for migrating individuals (Salo et al. 1986, Wagner & Funk 1995). Boreal areas, which lack the ample energy resources of wet tropical areas, may experience lower speciation rates even in the presence of complex geography. These areas have, on the other hand, been influenced by expanding and retreating glaciers of the Pleistocene (Hewitt 1999). Given enough time, genetic and phenotypic differences between populations accumulate and at some stage lead to speciation. Biological speciation is the development of a reproductive isolation barrier between differentiated populations before (prezygotic) or after (postzygotic) fertilization of an egg cell, fuelled by an increase of developmental disturbances in hybrids. In an allopatric situation, pre- and postzygotic barriers accumulate at a near equal pace, while in sympatry, prezygotic barriers become the dominant form of isolation (Coyne & Orr 1997). Development of isolating barriers in sympatric or parapatric situations is a direct consequence of continuous hybrid formation, which creates a strong selective response to reduced fitness of interbreeding individuals. Any behavioural, morphological or physiological changes in characters that prevent mating with non-conspecifics are favoured (Ritchie et al. 10

1989, Butlin & Ritchie 1991). The process called reinforcement enhances and stabilises species-specific mating signals especially in areas where populations that have diverged come in to secondary contact (Butlin et al. 1991). The role of sexual selection is essential in speciation of promiscuous species. Fisher’s run-a-way process and selection for “good genes” are examples of mechanisms that may enhance changes associated with sexual signals (Fisher 1930, Zahavi & Zahavi 1997). The theories postulate that the trait(s) for which a female has a preference for will increase in frequency in the given population. However, a character that has initially arisen through sexual selection may get fixed in a population and lose its information value to the female in selection for “good genes”, which may lead to changes in the target of female preference (Klappert et al. 2007). These selective processes are characterized by an arms race of both sexes, as the benefit of a male often does not coincide with that of a female (Parker 1979). It can be said that the line between sexual selection and conflict is subtle (Arnqvist & Rowe 2005). In this framework, it is not surprising that speciation rates of promiscuous species are elevated (Arnqvist et al. 2000). Integration of genetic methods to speciation and evolutionary ecology studies has made the previously unattainable estimation of neutral population differentiation, size, history and migration straight from DNA sequences possible. This has opened up new possibilities in speciation studies, where the estimation of genetic differentiation and gene flow between populations and species is of primary concern. In fact, the estimation of neutral genetic differentiation forms the base line for inferring the levels of selection that different characters undergo (Merilä & Crnokrak 2001).

1.1 Genetic divergence

A wide variety of molecular markers is used to estimate genetic differentiation of populations. These include RFLP, RAPD, AFLP, microsatellites and mtDNA sequences. The two latter markers have the highest impact in ecological and evolutionary studies of species. The level of neutral genetic differentiation of populations is best estimated with neutral or nearly neutral markers like mtDNA sequences and microsatellites. Recently, new methods, like the DNA microarray technology, have become available for ecological and evolutionary studies (Gibson 2002). Although still very costly, microarrays reveal the exact genes involved in a particular phenotypic change by measuring their expression patterns. Long term population demographic history and species phylogenies are best inferred by mtDNA haplotypes. There are a few qualities that make mtDNA sequencing in an especially attractive marker for evolutionary research. DNA sequences of mtDNA have higher replication and mutation rates than nuclear DNA. All alleles are haploid, which is a great benefit for DNA sequencing, though sometimes nuclear copies of mtDNA may cause problems (Gellissen et al. 1983). In most animal species, cytoplasm and mtDNA are 11 maternally inherited, which prevents a conflict of two parental cytoplasts in offspring (Hurst & Hamilton 1992). Selection acting on mtDNA is thought to be restricted, even though recent studies have cast some doubt on the neutrality of mtDNA variation (Nigro 1994, Hurst and Jiggins 2005, Dowling et al. 2007). Present-day knowledge of nuclear DNA evolution predicts that so called “junk” or non-coding DNA is mostly functionally neutral and thus no selection pressures act upon it (but see Cheng et al. 2005, Rigoutsos et al. 2006). Most microsatellites are located in non-coding DNA, while microsatellites found close to the promoter area or within the coding region of the genes can cause drastic changes in the phenotype (Yu et al. 1992). Microsatellites consist of 2-6 bp motifs that are arranged as repetitive structures scattered randomly all over a nuclear genome. They are codominant markers and are therefore a preferred marker type in studies of recent history in population demography and in the analysis of population genetic structures. The mutation rate of microsatellites is elevated compared to the rest of the nuclear DNA due to a slipped-strand mispairing which acts as the chief mutational force (Tautz 1989). The mechanism ensures the continuous inflow of polymorphic alleles to a population, but the drawback of the system is the downward direction of mutations in long repeats (Harr & Schlötterer 2000). The result of this is that probability of homoplasy (i.e. convergent evolution of the size of the alleles) increases when the evolutionary distance between the alleles grows longer. A combination of different molecular marker types offers an excellent way to estimate the demographic history and phylogeography of populations. Microsatellites have limitations, which can at least partly be bypassed with mtDNA and vice versa. The two markers also give information on population history on a different time scale. For example, in the D. virilis group, the existing estimates of population phylogeographic history based on molecular markers fit well into the overall picture of genetic changes induced by inter- and post- glacial periods of the Pleistocene (Spicer & Bell 2002, Hewitt 2004).

1.2 Phenotypic divergence

Behavioural mating signals change fast in early stages of speciation when populations that have genetically diverged come into secondary contact with each other (Dobzhansky 1951). Also, promiscuity combined with sexual selection can lead to rapid changes in behaviour and morphology, like in the cichlids of Lake Victoria (Seehausen & van Alphen 1999). During the courtship, the males and females emit various kinds of species-specific stimuli (e.g. courtship songs) to each other. The songs may play an important role in sexual selection and/or in species recognition, and they may simultaneously be affected by directional, diversifying and balancing selection. The signals important in species recognition may not, however, vary too much if they are to retain species-specificity (Lambert & Henderson 1986). Lande (1982) has shown that the evolution of directional female mating preferences for male secondary 12 sexual characters can greatly amplify large-scale geographic variation in male characters. This coevolutionary process can be enhanced by variation in the strength of direct or indirect selection on female preferences through the species distribution area (see Houde 1993). For example in fish, a female sensory bias for yellow prey items may cause an indirect selection of female preference for yellow male colouration (Garcia & Ramirez 2005). At the species level, diversifying selection may speed up the evolution of species-specific courtship songs and might increase the effectiveness of prezygotic sexual isolation between sympatric species (Etges et al. 2006). Depending on the nature of a behavioural signal, the signal and morphology related to it may co-evolve and thus cause concordant patterns. Drosophila flies use their wings for flying and the production of species-specific male courtship songs. Even though these are not exclusive functions, some characters of the wing may be affected mainly by sexual selection and other characters by natural selection. Variation in wing shape can also lead to functionally identical outcomes with various internal structural rearrangements due to drift or neutral diversification. For example, the convergence of clinal variation in wing size of D. subobscura on different continents has been achieved through analogous changes in the relative lengths of different parts of the wing (Huey et al. 2000). Additionally, wing traits have been found to evolve quite rapidly in response to geographic clines, e.g. in Drosophila subobscura (Huey et al. 2000, Gilchrist et al. 2000), and they have also been found to respond well to artificial selection (Houle et al. 2003, Kennington et al. 2003). These rapid changes are not surprising because of the abundant genetic variation related to wing shape (Mezey & Houle 2005, Weber et al. 2005). The size and shape of male genitalia are rapidly evolving species-specific characters and they are often used for species identification, e.g. in Drosophila species (Grimaldi 1990). Genitalia have been suggested to evolve via lock-key mechanics (Dufour 1844), pleiotropy (Mayr 1963) or cryptic female choice (Eberhard 1996). The striking morphological diversity of genitalia may also have arisen through sperm competition (Parker 1970) or sexual conflict (Hosken & Stockley 2004, Arnqvist & Rowe 2005). The highest diversity in genitalia would be expected to evolve in promiscuous species which have weak sexual selection on other traits or whose genitalic behaviour maintains a significant role in mating (Peretti et al. 2006). Genital morphology is likely to be influenced by many of the above-mentioned mechanisms at the same time with varying intensities.

1.3 Why to study Drosophila virilis group species?

The annotated nucleotide sequence of the genetic model organism (Adams et al. 2000) is complete and also for several other Drosophila species, e.g. for D. virilis, total genomic sequences are available. The new genetic tools developed for D. melanogaster, as well as the availability of genetic 13 information, also enables comparative studies in genetically less well-known Drosophila species with ecologically and behaviourally interesting characters. The D. virilis group is comprised of 12 species or subspecies that originated somewhere in the ancient deciduous forests of China or in the arid regions of Iran or Afghanistan (Throckmorton 1982). D. virilis, the name species of the group, is a domestic species distributed mainly south of latitude 35oN. D. montana, on the contrary, is found in natural boreal forest habitats close to water. It is distributed mainly north of latitude 40oN, with southern populations at higher elevations (Fig. 1). D. virilis possesses a primitive karyotype of the D. virilis group and in contrast to other species of the group, it shows no inversion polymorphism (Throckmorton 1982). However, D. montana has a large amount of inversion polymorphisms within and between populations (e.g. Moorhead, 1954, Morales-Hojas et al. 2007). This contrast between our study species is interesting in terms of the possible role of chromosomal inversions in speciation (Butlin 2005). The males of all species of the D. virilis group produce acoustic signals, courtship songs, by vibrating their wings during the courtship rituals. These songs play an important role both in species recognition (Liimatainen & Hoikkala 1998) and in sexual selection (Aspi & Hoikkala 1995). Song characteristics vary much more among the montana phylad species than among the virilis phylad species (Hoikkala & Lumme 1987). Also, the importance of the courtship song varies between the species: D. virilis females accept the courting male even without hearing his song (e.g. Saarikettu et al. 2006), while D. montana females do not accept the courtship of a “mute” (wingless) male (Hoikkala 1988, Liimatainen et al. 1992). 14

FIGURE 1 Proposed distribution areas of D. montana (light grey) and D. virilis (dark grey) in the wild and their overlapping distribution areas in East Asia (black). In addition, D. virilis is found in breweries and market places around the northern hemisphere. Areas with only few observations are marked with a question mark (chart modified from Throckmorton (1982)). Collection sites of fresh D. montana strains (article IV) are marked as black squares in Canada (Vancouver), USA (Colorado) and Finland (Oulanka).

1.4 Objectives of the project

The main objective of this thesis was to study phenotypic divergence of D. virilis and D. montana populations in a phylogeographic framework. The extent of variations in male courtship song, wing and genital traits were compared to the genetic divergence of laboratory strains and wild populations of the species to find out whether the studied traits had evolved neutrally or whether their evolution had been enhanced by selection. The first objective of the project was to analyse neutral genetic differentiation among the populations of D. virilis and D. montana. The data on variation in mtDNA haplotypes and microsatellite loci of these species were used primarily to estimate their phylogeography and demographic history (articles I, II & III). The second objective of the project was to analyse the phenotypic divergence of male courtship song traits in D. virilis populations and male song traits as well as the size and shape of wings and genitalia in D. montana populations (II &IV). The data on neutral genetic divergence of conspecific 15 populations (I & III) was used as a reference for comparing the levels of genotypic and phenotypic divergence between populations. The last part of the project concentrated on finding a fast and accurate molecular identification method for the North European D. virilis group species, as these species are very difficult to distinguish morphologically (V). So far, the species identification has been done on the basis of male genitalia, female spermathecae, RAPD fingerprinting, protein gel electrophoresis, and by mating wild caught females to laboratory males or by analysing the male courtship songs. All of these methods are very time consuming, not accurate enough and/or require special expertise. Recently there has been also interest in DNA barcoding of all species which based on sequencing mtDNA COI gene. However, this method does not have potential for population genetic or paternity studies like microsatellites. 2 MATERIALS AND METHODS

2.1 Study species

D. virilis strains used in this study originated from the whole distribution area of the species, from Europe to North America. The 52 isofemale strains have been collected during a time period covering almost 90 years. The most recent strains are from Chinese breweries in 2002 and from a Japanese lumberyard in 2003 (I & II). D. montana laboratory strains used in article III have been collected from various parts of the species distribution area during a time period of 60 years. Fresh isofemale strains of this species were established from mated females collected from three locations (Oulanka, Colorado and Vancouver) in 2003 (IV). Detailed lists of the species and strains used in this study are attached at the end of each article and manuscript.

2.2 Data analysis

2.2.1 Molecular data

Two mitochondrial genes, cytochrome oxidase I and cytochrome oxidase II, were sequenced for the subsequent analyses of D. virilis and D. montana. ARLEQUIN 2.0 (Schneider et al. 2000) was used to calculate pairwise distances between the haplotypes and the mismatch distributions. It was also used to perform Tajima’s D (1989) and Fu’s F (Fu & Li 1993) tests of the standard neutral model for a demographically stable population. The program FLUCTUATE (Kuhner et al. 1998) was used to make simultaneous estimates of present day lj and population growth rate g. The parameters used for the simulations were obtained by running a hierarchy of likelihood-ratio tests in Modeltest 3.0 (Posada & Crandall 1998) to choose the model of evolution with the best fit to the data. Phylogenetic trees were obtained using maximum likelihood with the molecular clock enforced in PAUP 4.0b10 (Swofford 1996), while skyline plots 17 were constructed using GENIE v. 3.0 (Pybus et al. 2000). GENIE was also used to calculate the fit to different models of population growth using the corrected Akaike Information Criterion. Networks of haplotypes were constructed based on statistical parsimony using the program TCS 1.06 (Clement et al. 2000). We used for D. virilis (paper II) 48 microsatellites distributed along the chromosomes (see the linkage map of Huttunen et al. 2004). A linkage disequilibrium test for pairs of these loci in ARLEQUIN 2.0 (Schneider et al. 2000) showed no linkage disequilibrium in European and American groups and only small fraction of loci had significant linkage disequilibrium in Asian and Japanese groups after Bonferroni correction. Also the 16 microsatellites of D. montana were chosen partly on the basis of their location on different chromosomes (Schäfer et al. in preparation, III). In both species, microsatellites were used to estimate recent neutral genetic differentiation of populations. Genetic differentiation between populations was calculated using F statistics according to Weir & Cockerham (1984). Basic measures of microsatellite variability were calculated using the Microsatellite Analyser Program (MSA; Dieringer & Schlötterer 2003). The amount of genetic variation resulting from differentiation between continents relative the genetic variation from geographical separation within continents was estimated with AMOVA in ARLEQUIN 2.0. Population substructure was studied with BAPS 2.0 (Corander et al. 2003) and STRUCTURE (Pritchard et al. 2000) programs depending on which suited better for the data in question. Using individuals of old laboratory strains in phylogeographic analysis caused some troubles in data analysis. Accordingly, we have used “group” in stead of “population” when clustering the strains from different geographic areas.

2.2.2 Phenotypic data

Male courtship songs were recorded on a tape recorder at a temperature of 20±1oC and analysed with the Signal 4.0 sound analysis system (Engineering Design, Belmont, MA, USA). The analysed song traits were the number of pulses in a pulse train (PN), the length of a pulse train (PTL), the length of a sound pulse (PL), the interpulse interval (IPI, the length of the time from the beginning of one pulse to the beginning of the next one), the number of cycles in a sound pulse (CN) and the carrier frequency of the song (FRE). The position of the wing veins was extracted from the image of a wing. The wing landmark data were aligned using the Generalized Procrustes Analysis (GPA). Centroid size was retained as a scaling variable so that size- dependent changes in shape could be explored. Two types of data were extracted from the wings: 15 (pseudo)landmarks describing the outlines of the wing (OL) and 12 landmarks (LM) describing the junctions between wing veins or between the veins and the outline of the wing (each having x and y coordinates). Digital images of the genitalia were analysed with the SHAPE 1.2 program (Iwata & Ukai 2002), which is a based in Principal Component Analysis (PCA) performed on elliptical Fourier descriptors (EFD) of an enclosed contour (Kuhl 18

& Giardina 1982). The resulting normalised Principal Component (PC) scores can be used as measured trait values that include only the allometric variation. The structure of the phenotypic data was nested; individuals within isofemale strains and strains within populations. Subsequent analyses of variance were done using nested ANOVAs as described in Sokal and Rohlf (1997). The direction of population differentiation was estimated with linear discriminant analysis (DA). The calculated strain means for different traits were used in DA to avoid pseudoreplication.

2.3 D. virilis group species identification

The species identification method was based on 14 microsatellite loci. A genetic distance tree was constructed with GenAlEx (Peakall & Smouse 2006) and visualised with MEGA (Kumar et al. 2004). Clustering of species and populations in groups was done using a Bayesian maximum likelihood based program, STRUCTURE (Pritchard et al. 2000). Identification of North European D. virilis group species was based on one locus. 3 RESULTS AND DISCUSSION

Global populations of D. virilis and D. montana showed moderate to high genetic and phenotypic differentiation. The estimated divergence time of D. montana populations on different continents appeared to be vastly longer than that of D. virilis populations. High genetic divergence of D. montana populations on different continents has not, however, led to high phenotypic divergence compared to that of populations on the same continent.

3.1 Phylogeographic patterns

3.1.1 Mitochondrial haplotypes

Mitochondrial haplotypes of D. virilis strains from different parts of the species distribution area did not show any geographic structure (I). A low number of substitutions shared among locations suggests a rapid population expansion over a relatively short time-scale (~50,000 years with a mutation rate of 10-8 per year). This time-scale is consistent with an expansion of the human population and may either reflect a shift of the flies into domestic environments or postglacial colonization. The central haplotype of the network can be found around the species distribution area and thus it gives no information on the geographic origin of the species. In their study on sequence variation for six X- linked genes in 21 D. virilis strains from different continents, Vieira and Charlesworth (1999) found no fixed differences between the Asian strains and strains originating from Europe, North- and South-America. The fact that all the variants found outside Asia were also present in Asia, but not vice versa, refers to a large population centered in Asia with smaller migrant populations elsewhere (Vieira & Charlesworth, 1999). According to Throckmorton (1982), D. virilis group species originate from an ancestral form in Asia, with the most primitive species of the virilis-repleta section observed in South-East Asia. Unlike D. virilis, D. montana possessed some geographic structure. mtDNA analysis indicates the presence of at least two distinct populations, one in 20

Eurasia and the other one representing the expansion of the species to the New World (III). The mtDNA haplotypes found in North America formed a clade nesting within the more diverse set of haplotypes present in Finland, supporting the idea that the species has spread from Eurasia to North-America (Throckmorton 1982). Genetic distances between the two major mtDNA clades range from 0.9% to 1.8%, which, assuming a mitochondrial divergence rate of 2 % per million years, implies a separation of 450,000 to 900,000 years. Päällysaho et al. (2005) obtained congruent divergence times between Finnish and American D. montana populations based on the silent substitutions on three X- linked genes (fused, elav and su(s)).

3.1.2 Microsatellites

Microsatellite loci, contrary to the mtDNA haplotype data, gave support to the population structure in D. virilis (II). The assignment test using a priori information about the geographical origin of the strains gave high posterior probabilities for the correct assignment of American, Asian, European and Japanese strains. Also, genetic variation measured with FST statistics suggested a significant population differentiation following the isolation-by-distance model. Variation in microsatellite loci can have a more recent origin than variation in mtDNA haplotypes. The lack of population differentiation in mtDNA indicates a shared ancestral polymorphism at the mitochondrial level, while variation in microsatellite loci indicates that differentiation among populations is still ongoing. Microsatellite analysis distinguished Finnish and North American D. montana populations like the mtDNA analysis, but it also suggested genetic differentiation between North American populations (IV). Although mtDNA haplotypes from Canada and USA were not phylogenetically distinct, the populations were significantly differentiated as judged by mtDNA genetic variation. Microsatellite variation also showed evidence for different historical demographic patterns of populations. This might reflect partial or complete isolation into distinct northern (Beringian) and southern (Rocky Mountains) refugia (Hewitt 2004) during the last glaciation, or conversely to different colonization times of populations. Both of the Finnish and the Canadian samples suggest a very rapid population expansion around 35,000 years ago, somewhat before the end of the last glaciation period. The USA sample suggests an earlier gradual expansion consistent with the more southern location of the Rocky Mountains refuge. 21

3.2 Phenotypic differentiation

3.2.1 Male courtship songs in D. virilis and D. montana

In D. virilis, the male courtship songs showed significant interstrain and geographic variation (IV) in the number of pulses in a pulse train (PN) and the pulse train length (PTL), as well as in two sound pulse characters: the number of cycles in a pulse (CN) and the carrier frequency of the song (FRE). The strains from continental Asia and Europe expressed higher values than the American and Japanese strains for all of these traits. The genetic distances and the song divergence of the strains did not show significant association, which suggests that the songs have not diverged solely as a side-effect of genetic divergence. The length of time the strains had been kept in the laboratory prior to song recording decreased the values of CN and FRE. Also, the songs of laboratory strains showed a slight, but significant, decrease in CN and FRE and higher increases in pulse length (PL) and interpulse interval (IPI) when compared to the songs of the progenies of wild-caught females. Laboratory rearing had no effect on PN or PTL, and it did not eradicate interstrain and geographic variation in CN and FRE. However, parallel changes in the songs of the laboratory strains could explain the lack of geographic variation in PL and IPI. Lack of intra- and interspecific variation in IPI has also been observed in other Drosophila species (e.g. in D. ananassae and D. pallidosa), where the phenomenon has been suggested to be due to selection on other species-specific song characters requiring constant IPI (Yamada et al. 2002). The male courtship songs of the laboratory strains of D. montana originating from a wide geographic area showed the highest divergence in the pulse characters of the song, especially in FRE (III). Also, the songs of males from wild populations (F1 progenies of wild-caught females) in Oulanka (Finland), Colorado (USA) and Vancouver (Canada) showed divergence in pulse characters PL and CN (IV). The fact that the songs of the males from wild populations showed no significant divergence in FRE was mainly due to high variation in this trait within the isofemale strains. The song frequency is quite sensitive to changes in environmental factors (Hoikkala & Suvanto 1999) and has a low heritability (Aspi & Hoikkala 1993, Suvanto et al. 1998), and thus one would expect any change under selection in this character to be slow. In addition to PL and CN, IPI also varied significantly among D. montana populations (IV). This trait has been found to play an important role in species recognition in D. montana (Saarikettu et al. 2005) as well as in several other Drosophila species (e.g. Cowling & Burnet 1981). The songs of all the species of the montana phylad have a species-specific IPI (Hoikkala & Lumme 1987). Variation between D. montana populations in the male courtship song could have been enhanced by character displacement if the flies of different populations are sympatric with different species of the D. virilis group and if they interact/hybridize with them in the wild. In Finland, interspecific courtships are quite common between D. montana, D. ezoana, D. littoralis and D. 22 lummei (Aspi et al. 1993, Liimatainen & Hoikkala 1998). In Colorado, D. montana occurs sympatrically with D. borealis and D. flavomontana (Hoikkala & Mazzi personal observation), while in Vancouver it is probably the only representative of the D. virilis group (Klappert & Orsini unpublished observation). It is not possible to trace the historical patterns of sympatry between different species, but it is intriguing that a species-specific song trait (IPI) shows such a high variation between conspecific populations (see Hoikkala & Lumme 1987). It is important for the species-recognition signals to differ from those of other sympatric species to effectively prevent hybridization.

3.2.2 Wing size and shape in D. montana

Drosophila species are known to evolve latitudinal wing morphology clines when introduced into novel environments (Gilchrist et al. 2000, Santos et al. 2004), reflecting increasing body size in a colder climate (Huey et al. 2000). Similarly, altitude has been shown to have an effect on wing morphology by increasing wing load due to colonisation or the scarcity of food resources (Norry et al. 2001). In the present study (IV), variation between the wing measures of males from the Oulanka, Colorado and Vancouver populations was mainly due to changes on the tip as well as on the front and back edges of the wings; the differences were highest between the two latter populations. The males from the Vancouver population also diverged by the size and the internal landmarks of the wings from the other two populations. The differentiation, which is likely to have aerodynamic consequences, could be due to natural selection on the wing form to adjust to the aerodynamic optimum of a local climate. Also, a wing landmark pinpointing the endpoint of the second long vein on the outline of the wing showed co-variation with the combined effect of three pulse characters of the song.

3.2.3 Genitalia size and shape in D. montana

Male genitalia size and shape showed divergence between Vancouver and Oulanka as well as the Vancouver and Colorado populations in the discriminant analysis (IV). The cross-validated classification test showed an overlap between the Colorado and Oulanka populations. Nested ANOVA revealed a strong within-strain variance for all genitalia shape traits suggesting that they are sensitive to environmental factors and/or to effects of sample preparation. Size and shape of male genitalia could be affected in the wild by cryptic female choice or by variation of a male’s ability to remove the previous male’s sperm with the genitalia. These factors might also be important in D. montana as the females of this species mate repeatedly in the wild where the last male sires most of the offspring (Aspi & Lankinen 1991). Also, sexual conflict over the duration of copulation (Klappert & Mazzi unpublished results) may influence the curvature of the genitalia and the genital hook angle to increase the male’s ability to prolong copulation (Arnqvist & Rowe 2005). 23

3.3 Genetic species identification in the D. virilis group

Species identification was based on 14 microsatellite loci that were amplified in all of the D. virilis group species. This set of markers was divergent enough to separate the species according to their species status with a few exceptions. D. montana clustered to three separate groups according to their geographic origin (Fig. 2). Subspecies D. a. americana and D. a. texana were grouped together, thereby supporting the view that these subspecies are just chromosomal forms of one single species of D. americana (Schlötterer 2000, Schäfer et al. 2006) Additionally, microsatellite data did not support species status for D. canadiana.

FIGURE 2 Unrooted neighbor joining tree based on 14 microsatellite loci. 4 CONCLUSIONS

D. virilis and D. montana, two species with different ecological adaptations, show remarkably contrasting phylogeographic patterns and phenotypic divergence. The expansion of D. virilis on different continents has probably been associated with evolutionarily recent human population expansion. The environmental niches of this species are rich in resources and relatively free of predators, which may have lead to low levels of genetic and phenotypic differentiation. D. montana, on the other hand, lives in highly heterogeneous environments, which has led to high genetic and phenotypic divergence between populations. The studied D. montana populations have unique combinations of phenotypic characters, which have not followed convergent evolutionary pathways. The study on geno- and phenotypic variation in D. virilis (II) reveals possible pitfalls in evolutionary studies. The male courtship songs of all laboratory strains of this species have evolved in the same direction compared to wild populations, which may have led to diminished geographic variation in the lengths of sound pulses and interpulse intervals. While the laboratory strains provide good material for genetic crosses and studies on gene structure and function, the studies on the genetic and geographic variation in traits like the male courtship song should be done within the framework of a wild population. D. montana populations have intriguing divergence patterns that are demonstrated in article IV. The males of the laboratory strains from USA (III) as well as the males from the Colorado population (IV) have a higher song frequency than the males from Canada and Finland. The Finnish D. montana females have previously been shown to prefer the males that produce a high frequency song as their mating partner (e.g. Aspi & Hoikkala 1995, Ritchie et al.1998). Surprisingly, the females from the Colorado population, where the males sing with a high frequency, do not show this preference or show a preference for lower song frequency (Klappert et al. 2007). Does this mean that all of the males from this population sing with a high frequency so that this signal has lost its value in mate choice? This question requires further studies 25 on male song / female preference covariation in the Colorado population, similar to the one made for the Finnish population (Ritchie et al. 2005). Also, mate choice experiments between the flies from different populations could give more information on the genetic basis and evolution of male songs and female song preferences. Another interesting discovery worth studying further is the coevolution of male song traits and wing shape. The ease of extracting data from the wing morphology opens new possibilities also for evo-devo studies. Due to a small number of study populations, the genetic and phenotypic variance was not compared to each other directly (for example, using FST and QST statistics; Merilä & Crnokrak 2001). Instead, we used a multivariate analysis and Mantel test to trace the direction of phenotypic divergence between populations as well as its correlation with genetic divergence. Differences between populations in male song, wing and genital traits were not compared with each other because of different time and scale measurement units (e.g. between song and wing traits) and also because of scale differences between wing and genital traits. The two species, D. montana and D. virilis, together form a powerful model system in evolutionary ecology and speciation studies of wild populations. The fully sequenced D. virilis genome can be used for designing experiments in D. montana, the species more interesting in evolutionary and ecological aspects. Also, behavioural mutations of the genetic model species, D. melanogaster, could be traced down in the D. virilis genome and then eventually in D. montana. Studies that require massive sequencing operations of previously unknown sequences in D. montana (e.g. like the construction of microarrays), benefit greatly from the known genetics of the model species. The species concept is a controversial topic. It depends on how the species is defined and what kinds of methods are used to distinguish species and conspecific populations. The results of our species identification analyses using microsatellite markers were not fully identical with the previously known species and subspecies boundaries. The genetic method may help clarify inconsistencies and mistakes in previous cases of species identification, and even more importantly, this method will enable large-scale studies on the flies of wild populations among North European D. virilis group species. More studies with more microsatellite loci and fly samples are necessary to reliably identify all of the species in the D. virilis group. 26

Acknowledgements

Thesis work was carried out during years 2003-2007 at the Department of Biological and Environmental Science in University of Jyväskylä, Finland, at the School of Biology in University of Leeds, England and at the Institut für Tierzucht und Genetik, Veterinärmedizinische Universität Wien, Austria. The Section of Ecology in Jyväskylä provided great facilities and a creative atmosphere. The Graduate School in Evolutionary Ecology also had a strong positive impact. The Centre of Excellence in Evolutionary Research and EU Research Training Network ‘Co-evolved Traits’ offered the facilities and the framework for the whole project. Greatest thanks go to Anneli Hoikkala, who provided me the opportunity to start a PhD project. She especially helped during the writing period. Maaria Kankare created a friendly and dynamic atmosphere. Dominique Mazzi helped a lot with collecting data and was good company when I arrived in Jyväskylä. Anneli’s new PhD students are acknowledged for freeing me from fly stock maintenance. Roger Butlin is acknowledged for giving me a chance to work in his lab and improving my English. Patricia Mirol introduced me to phylogenetic inference. Pedro Pedro upgraded my bioinformatics skills. Anahi Espindola was an excellent person to supervise! Bunch of thanks for the large and dynamic community of PhD students in Leeds during the summer of 2004! Christian Schlötterer is acknowledged for letting me work in his lab. Martin Schäfer introduced me to radioactive genotyping. Thanks also go to the whole group in the CS lab during September 2005! Kirsten Klappert, Ramiro Morales-Hojas, Losia Nakagawa-Lagisz, Jorge Vieira and Mike Ritchie gave helpful comments and were otherwise cool in the EU network meetings. Kim van der Linde analysed the wing morphology. Many thanks go to valuable co-authors Susanna Huttunen, Luisa Orsini and Jouni Aspi. Thanks also go to the reviewers, Helena Korpelainen and Perttu Seppä, whose comments greatly improved this thesis. Mika Mökkönen and Emily Knott improved the English. Over the years, a number of individuals have helped me with the project somehow and/or have caused a good atmosphere in situ; billiards master Tuomo Pihlaja, Marjo Pihlaja, Christophe Lebigre, Teppo Hiltunen, Carita Lindstedt, Tarmo Ketola, Nina Pekkala, Mikael Puurtinen and Anssi Rantakari from the old school. Maarit Kokkonen contributed greatly to the last manuscript. Members of the Aikido groups, Jigotai in Jyväskylä and Yorkshire Aikikai in Leeds, helped me to reset my brain at the end of the day and kept me physically in condition. My sister Johanna Routtu is appreciated for the help in TAYS. Last but not least, my parents are acknowledged for a huge amount of help, especially my mother Anja Routtu who always encouraged me to find my own way. 27

This work was financially supported by Marie Curie Scholarship, Centre of Excellence in Evolutionary Research, Finnish Cultural Foundation and Ellen and Artturi Nyyssönen Foundation. 28

YHTEENVETO (RÉSUMÉ IN FINNISH)

Geneettinen ja fenotyyppinen erilaistuminen Drosophila virilis ja D. montana lajien mahlakärpäsillä

Drosophila virilis ja D. montana lajien mahlakärpästen geneettinen ja fenotyyppinen erilaistuminen on edennyt evolutiivisesti eri reittejä. Ihmisen seuralaislajin, D. virilis -lajin, eri maantieteellisiltä alueilta peräisin olevat kannat eivät erotu toisistaan mtDNA haplotyyppien osalta. Lajin leviäminen ja populaatioiden koon kasvu on voinut tapahtua ihmispopulaatioiden kasvun ja uudelle mantereelle levittäytymisen myötä. Mikrosatelliitti-analyysi ryhmittelee D. virilis -lajin laboratoriokannat maantieteellisen alkuperän mukaan neljään ryhmään, mikä viittaa populaatioiden viimeaikoina tapahtuneeseen erilaistumiseen. Luonnossa vesistöjen äärellä elävän D. montana -lajin Euroopassa ja Pohjois-Amerikassa elävät populaatiot ovat eriytyneet toisistaan sekä mitokondriaalisten haplotyyppien että mikrosatelliittialleelien osalta. Tällä lajilla mtDNA haplotyypit eivät eroa kahden pohjois-amerikkalaisen populaation välillä, kun taas mikrosatelliitti-analyysi erottelee kaikki populaatiot omiksi ryhmikseen. Fenotyyppistä eriytymistä tutkittiin D. virilis -lajilla koiraan kosintalaulun ja D. montana -lajilla koiraan kosintalaulun sekä siipien ja genitaalien koon ja muodon perusteella. D. virilis -lajin laboratoriokantojen laulut olivat erilaistuneet kantojen maantieteellisen alkuperän mukaan. Geneettinen erilaistuminen ja laulujen eriytyminen eivät olleet edenneet yhdenmukaisesti, mikä viittaa siihen, että laulut eivät ole erilaistuneet pelkästään neutraalin geneettisen erilaistumisen sivutuotteena. D. montana -lajilla laboratoriokantojen koiraiden laulut poikkesivat toisistaan selkeimmin laulun frekvenssin suhteen, kun taas luonnon populaatioiden koiraiden laulut erosivat toisistaan enemmän äänipulssien muissa ominaisuuksissa. D. montana –lajin populaatiot poikkesivat toisistaan myös siipien ja genitaalien koon ja muodon suhteen. Populaatioiden välinen fenotyyppinen eriytyminen ei ole kulkenut samaan tahtiin genotyyppisen eriytymisen kanssa, mikä viittaa siihen, että ominaisuuksissa syntyneet erot eivät olleet kehittyneet neutraalisti. Viimeisessä osatutkimuksessa kehitimme pohjois-eurooppalaisille D. virilis -ryhmän lajeille lajintunnistusmenetelmän helpottamaan luonnosta kerättyjen yksilöiden tunnistusta ja mahdollisesti väärin tunnistettujen laboratoriokantojen identifioimista. Menetelmä perustuu kaikissa D. virilis - ryhmän lajeissa monistuvien polymorfisten mikrosatelliittien käyttöön. Tasaisesti koko genomiin jakautuneet markkerit selkeyttävät D. virilis -ryhmän lajien keskinäisiä sukulaisuussuhteita. Näitä molekyylimarkkereita voi myös käyttää mm. isyystutkimuksiin ja geenien paikallistamiseen samoin kuin luonnonpopulaatioiden laajamittaiseen tutkimiseen. 29

REFERENCES

Adams, M.D., Celniker, S.E., Holt, R.A., Evans, C.A., Gocayne, J.D., Amanatides, P.G., Scherer, S.E., Li, P.W., Hoskins, R.A., Galle, R.F. et al. 2000. The genome sequence of Drosophila melanogaster. Science 287: 2185- 2195. Arnqvist, G. & Rowe, L. 2005. Sexual Conflict. Princeton University Press, Princeton and Oxford. Arnqvist, G., Edvardsson, M., Friberg, U. & Nilsson, T. 2000. Sexual conflict promotes speciation in . Proc. Natl. Acad. Sci. 97: 10460-10464. Aspi, J. & Hoikkala, A. 1993. Laboratory and natural heritabilities of male courtship song characters in Drosophila montana and D. littoralis. Heredity 70: 400-406. Aspi, J. & Hoikkala, A. 1995. Male mating success and survival in the field with respect to size and courtship song characters in Drosophila littoralis and D. montana (Diptera: ). J. Behav. 8: 67-87. Aspi, J. & Lankinen, P. 1991. Frequency of multiple insemination in a natural population of Drosophila montana. Hereditas 117: 169-177. Aspi, J., Lumme, J., Hoikkala, A. & Heikkinen, E. 1993. Reproductive ecology of the boreal riparian guild of Drosophila. Ecography 16: 65-72. Bush, G.L. 1969. Sympatric host race formation and speciation in frugivorous flies of the genus Rhagoletis (Diptera: Tephritidae). Evolution 23:237-251. Butlin, R.K. 2005. Recombination and speciation. Mol. Ecol. 14: 2621-2635. Butlin, R.K. & Ritchie, M.G. 1991. Variation in female mate preference across a grasshopper hybrid zone. J. Evol. Biol. 4: 227-240. Butlin, R.K., Ritchie, M.G. & Hewitt, G.M. 1991. Comparisons among morphological characters and between localities in the Corthippus parallelus hybrid zone (Orthoptera, Acrididae). Philos. Trans. R. Soc. Lond. B. 334: 297-308. Cheng, J., Kapranov, P., Drenkow, J., Dike, S., Brubaker, S., Patel, S., Long, J., Stern, D., Tammana, H., Helt, G. et al. 2005. Transcriptional maps of 10 human chromosomes at 5-nucleotide resolution. Science 308: 1149-1154. Clement, M., Posada, D. & Crandall, K.A. 2000. TCS: a computer program to estimate gene genealogies. Mol. Ecol. 9: 1657-1659. Corander, J., Waldmann, P. & Sillanpää, M.J. 2003. Bayesian analysis of genetic differentiation between populations. Genetics 163:367-374. Coyne, J.A. & Orr, H.A. 1997. “Patterns of speciation in Drosophila” revisited. Evolution 51: 295-303. Coyne, J.A. & Orr, H.A. 2004. Speciation, Sinauer Associates, Inc., Sunderland, Massachusetts. Cowling, D. E., & Burnet, B. 1981. Courtship songs and genetic control of their acoustic characteristics in sibling species of the Drosophila melanogaster subgroup. Anim. Behav. 29: 924–935. 30

Darwin, C. 1859. The Origin of Species by Means of Natural Selection, first edition, John Murray, London. Dieringer, D. & Schlötterer, C. 2003. Microsatellite analyser (MSA): a platform independent analysis tool for large microsatellite data sets. Mol. Ecol. Notes 3: 167-169. Dobzhansky, T. 1951. Genetics and the Origin of Species. Third edition, Columbia Univ. Press, New York. Dowling, D.K., Friberg, U., Hailer, F. & Arnqvist, G. 2007. Intergenomic epistasis for fitness: Within-population interactions between cytoplasmic and nuclear genes in Drosophila melanogaster. Genetics 175: 235-244. Dufour, L. 1844. Anatomie generale des Dipteres. Ann. Sci. Nat. 1: 244-64. Eberhard, W. G. 1985. Sexual Selection and Animal Genitalia. Harvard Univ. Press, Cambridge. Etges, W.J., Over, K.F., Cardoso de Oliveira, C & Ritchie M.G. 2006. Inheritance of courtship song variation among geographically isolated populations of Drosophila mojavensis. Anim. Behav. 71: 1205–1214. Fisher, R.A. 1930. The Genetical Theory of Natural Selection. Clarendon Press, Oxford. Fu, Y.X. & Li, W.H. 1993. Statistical tests of neutrality of mutations. Genetics 133: 693-709. Garcia, C.M. & Ramirez, E. 2005. Evidence that sensory traps can evolve into honest signals. Nature 434: 501-505. Gibson, G. 2002. Microarrays in ecology and evolution: a preview. Mol. Ecol., 11: 17–24 Gilchrist, G.W., Huey, R.B. & Serra, L. 2000. Rapid evolution of wing size clines in Drosophila subobscura. Genetica 112-113: 273-86. Gellissen, G., Bradfield, J.Y., White, B.N. & Wyatt, G.R. 1983. Mitochondrial DNA sequences in the nuclear genome of a locust. Nature 301: 631-634. Grimaldi, D.A. 1990. A phylogenetic, revised classification of genera in the Drosophilidae (Diptera). Bull. Am. Mus. Nat. Hist. 197: 1–139. Harr, B. & Schlötterer, C. 2000. Long microsatellite alleles in Drosophila melanogaster have a downward mutation bias and shorter persistence times, which cause their genome-wide underrepresentation. Genetics 155: 1213–1220. Hewitt, G.M. 1999 Post-glacial re-colonization of European biota. Biol. J. Linn. Soc. 68: 87-112. Hewitt, G. M. 2004. Genetic consequences of climatic oscillations in the Quaternary. Philos. Trans. R. Soc. Lond. B. 359: 183–195. Hoikkala, A. 1988. The importance of different courtship stimuli in the mating behavior of European species of the Drosophila virilis group. 1988. Ann. Zool. Fenn. 25: 257-263. Hoikkala, A. & Lumme, J. 1987. The genetic basis of evolution of the male courtship sounds in the Drosophila virilis group. Evolution 41: 827-845. Hoikkala, A. & Suvanto, L. 1999. Male courtship song frequency as an indicator of male mating success in Drosophila montana. J. Insect Behav. 12: 599- 609. 31

Hosken, D.J. & Stockley, P. 2004. Sexual selection and genital evolution. Trends Ecol. Evol. 19: 87–93. Houde, A. E. 1993. Evolution by sexual selection: What can population comparisons tell us? Am. Nat. 141: 796-803. Houle, D., J.G. Mezey, P. Galpern & Carter, A. 2003. Automated measurement of Drosophila wings. BMC Evol. Biol. 3: 25. Huey, R. B., Gilchrist, G. W., Carlson, M. L., Berrigan, D. & Serra, L. 2000. Rapid evolution of a geographic cline in size in an introduced fly. Science 287: 308-309. Hurst, G.D.D. & Jiggins, F.M. 2005. Problems with mitochondrial DNA as a marker in population, phylogeographic and phylogenetic studies: the effects of inherited symbionts Proc. R. Soc. Lond. B. 272: 1525-1534. Hurst, L.D. & Hamilton, W.D. 1992. Cytoplasmic fusion and the nature of sexes. Proc. R. Soc. Lond. B. 247: 189-194. Iwata, H. & Ukai, Y. 2002. SHAPE: A computer program package for quantitative evaluation of biological shapes based on elliptic Fourier descriptors. J. Hered. 93: 384-385. Huttunen, S., Aspi, J., Hoikkala, A. & Schlötterer, C. 2004. QTL analysis of variation in male courtship song characters in Drosophila virilis. Heredity 92: 263-269. Kennington, W.J., Killeen, J.R., Goldstein, D.B. & Partridge, L. 2003. Rapid laboratory evolution of adult wing area in Drosophila melanogaster in response to humidity. Evolution 57: 932–936. Klappert, K., Mazzi, D., Hoikkala, A. & Ritchie, M.G. 2007. Male courtship song and female preference variation between phylogeographically distinct populations of Drosophila montana. Evolution 61-6: 1481–1488. Kuhl, F.P. & Giardina, C.R. 1982. Elliptic Fourier features of a closed contour. Comp. Graph. Imag. Proc. 18: 236-258. Kuhner, M.K., Yamato, J. & Felsenstein, J. 1998. Maximum likelihood estimation of population growth rates based on coalescent. Genetics 149: 429-434. Kumar, S., Tamura, K & Nei, M. 2004. MEGA3: Integrated software for Molecular Evolutionary Genetics Analysis and sequence alignment. Brief. Bioinf. 5:150-163. Lambert, D.M. & Henderson, N.R. 1986. The stability of the specific-mate recognition system of Drosophila melanogaster. Behav. Genet. 16: 369–373. Lande, R. 1982. Rapid origin of sexual isolation and character divergence in a cline. Evolution 36: 213–223. Liimatainen, J. & Hoikkala, A. 1998. Interactions of the males and females of three sympatric Drosophila virilis group species, D. montana, D. littoralis and D. lummei, in intra- and interspecific courtships in the wild and in the laboratory. J. Insect Behav. 11: 399-417. Liimatainen, J., Hoikkala, A., Aspi, J. & Welbergen, Ph.1992. Courtship in Drosophila montana: the effects of male auditory signals on the behaviour of the flies. Anim. Behav. 43: 35-48. Maynard Smith, J. 1966. Sympatric speciation. Am. Nat. 100: 637–650. 32

Mayr, E. 1954. Change of genetic environment and evolution. In Huxley, J., ed., Evolution as a Process, Allen and Unwin, London. Mayr, E. 1963. Animal Species and Evolution. Harvard Univ. Press, Cambridge. Merilä, J. & Crnokrak, P. 2001. Comparison of genetic differentiation at marker loci and quantitative traits. J. Evol. Biol. 14: 892-903. Mezey, J. G. & Houle, D. 2005. The dimensionality of genetic variation for wing shape in Drosophila melanogaster. Evolution 59: 1027-1038. Moorhead, P.S. 1954. Chromosome variation in giant forms of Drosophila montana. Univ. Texas Publ. 5422: 106–129. Morales-Hojas, R., Päällysaho, S., Vieira, C.P., Hoikkala, A. & Vieira, J. 2007. Comparative polytene chromosome maps of D. montana and D. virilis. Chromosoma 116: 21-27. Nigro, L. 1994. Nuclear background affects frequency dynamics of mitochondrial DNA variants in Drosophila simulans. Heredity 72: 582-586. Norry, F. M. Bubliy, O. A. & Loeschcke, V. 2001. Developmental time, body size and wing loading in Drosophila buzzatii from lowland and highland populations in Argentina. Hereditas 135: 35-40. Päällysaho, S., Vieira, C.P., Hoikkala, A. & Vieira, J. 2005. Evidence for introgression in differentiated North-American and Finnish Drosophila montana populations. Genetica 123: 285-293. Parker, G.A. 1970. Sperm competition and its evolutionary consequences in insects. Biol. Rev. Camb. Phil. Soc. 45: 525-567. Parker, G.A. 1979. Sexual selection and sexual conflict. In M.S.Blum and N.A.Blum(Eds), Sexual Selection and Reproductive Competition in Insects, Academic, New York, pp. 123-166. Peakall, R. & Smouse P.E. 2006. GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research. Mol. Ecol. Notes. 6: 288-295. Peretti, A., Eberhard, W.G. & Briceno, R.D. 2006. Copulatory dialogue: female spiders sing during copulation to influence male genitalic movements. Anim. Behav. 72: 413-421. Posada, D. & Crandall, K.A. 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817-818. Pritchard, J.K., Stephens, M. & Donnelly, P. 2000. Inference of population structure using multilocus genotype data. Genetics 155: 945–959. Pybus O.G., Rambaut, A. & Harvey, P.H. 2000. An integrated framework for the inference of viral population history from reconstructed genealogies. Genetics 155: 1429-1437. Rigoutsos, I., Huynh, T., Miranda, K., Tsirigos, A., McHardy, A. and Platt, D. 2006. Short blocks from the noncoding parts of the human genome have instances within nearly all known genes and relate to biological processes. Proc. Nat. Acad. Sci. 103: 6605-6610. Ritchie, M.G. Butlin, R.K. & Hewitt, G.M. 1989. Assortative mating across a hybrid zone in Corthippus parallelus (Orthoptera, Acrididae) J. Evol. Biol. 2: 339-352. 33

Ritchie, M.G., Townhill, R.M. & Hoikkala, A. 1998. Female preference for fly song: playbacks confirm correlational evidence of the targets of sexual selection. Anim. Behav. 56: 713–717. Ritchie, M.G., Saarikettu, M. & Hoikkala, A. 2005. Variation, but no covariance, in female preference functions and male song in a natural population of Drosophila montana. Anim. Behav. 70: 849-854. Saarikettu, M., Liimatainen, J.O. & Hoikkala, A. 2005. The role of male courtship song in species recognition in Drosophila montana. Behav. Genet. 35: 257- 263. Salo, J., Kalliola, R., Häkkinen, I., Mäkinen, Y., Niemelä, P., Puhakka, M. & Coley, P.D. 1986. River dynamics and the diversity of Amazon lowland forest. Nature 332: 254-258. Santos, M., Iriarte, P.F., Céspedes, W., Balanyà, J., Fontdevila, A. & Serra, L. 2004. Swift laboratory thermal evolution of wing shape (but not size) in Drosophila subobscura and its relationship with chromosomal inversion polymorphism. J. Evol. Biol. 17: 841-855. Schäfer, M.A., Orsini, L., McAllister, B-F. & Schlötterer, C. 2006. Patterns of microsatellite variation through a transition zone of a chromosomal cline in Drosophila americana. Heredity 97: 291-295. Schlötterer, C. 2000. Microsatellite analysis indicates genetic differentiation of the neo-sex chromosomes in Drosophila americana americana. Heredity 85: 610-616. Schneider, S., Roessli, D. & Excoffier, L. 2000. Arlequin ver. 2.000: A software for population genetics data analysis. Genetics and Biometry Laboratory, University of Geneva, Switzerland. Seehausen, O. & van Alphen, J.M. 1999. Can sympatric speciation by disruptive sexual selection explain rapid evolution of cichlid diversity in Lake Victoria? Ecol. Lett. 2: 262-271. Sokal, R. R. & Rohlf, F. J. 1997. Biometry. The Principles and Practice of Statistics in Biological Research. Third edition. New York: Freeman. Spicer, G.S. & Bell, C.D. 2002. Molecular phylogeny of the Drosophila virilis species group (Diptera: Drosophilidae) inferred from mitochondrial 12S and 16S ribosomal RNA genes. Ann. Entomol. Soc. Am. 95: 156-161. Suvanto, L., Liimatainen, J. & Hoikkala, A. 1998. Variability and evolvability of male song characters in Drosophila montana populations. Hereditas 130: 13-18. Swofford, D.L. 1996. PAUP*: Phylogenetic analysis using parsimony (and other methods), version 4.0b10. Sinauer Associates, Sunderland, Massachusetts. Tajima, F. 1989. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123: 585–595. Tautz, D. 1989. Hypervariability of simple sequences as a general source for polymorphic DNA markers. Nucleic Acids Res. 17: 6463–6471. Throckmorton, L.H. 1982. The Genetics and Biology of Drosophila Vol. 3b. Academic Press, New York. Wagner, W.L. & Funk, V.A. 1995. Hawaiian Biogeography: Evolution on A Hot- Spot Archipelago. Smithsonian Institution Press, Washington DC. 34

Weber, K., Johnson, N., Champlin, D. & Patty, A. 2005. Many P-element insertions affect wing shape in Drosophila melanogaster. Genetics 169:1461-1475. Weir, B.S. & Cockerham, C.C. 1984. Estimating F-statistics for the analysis of population structure. Evolution 38: 1358-1370. Vieira, J. & Charlesworth, B. 1999. X chromosome DNA variation in Drosophila virilis. Proc. R. Soc. Lond. B. 266: 1905–1912. Yamada, H., Sakai, T., Tomaru, M., Doi, M., Muneo, M. & Oguma, Y. 2002. Search for species-specific mating signal in courtship songs of sympatric sibling species Drosophila ananassae and D. pallidosa. Genes Genet. Syst. 77: 97-106. Yu, S., Mulley, J., Loesch, D., Turner, G., Donnelly, A., Gedeon, A., Hillen, D., Kremer, E., Lynch, M., Pritchard, M. et al. 1992. Fragile-X syndrome – Unique genetics of the heritable unstable element. Amer. J. Hum. Genet. 50: 968-980. Zahavi, A. & Zahavi, A. 1997. The Handicap Principle: A Missing Piece of Darwin’s Puzzle. Oxford University Press, Oxford, UK. ORIGINAL PAPERS

I

SIGNALS OF DEMOGRAPHIC EXPANSION IN DROSOPHILA VIRILIS

by

Patricia M. Mirol, Jarkko Routtu, Anneli Hoikkala and Roger K. Butlin

Submitted Manuscript

Signals of demographic expansion in Drosophila virilis

Patricia M. Mirol1, Jarkko Routtu2, Anneli Hoikkala2 and Roger K. Butlin3

1 School of Biology, The University of Leeds, Leeds LS2 9JT, UK. 2 Department of Biological and Environmental Sciences, P.O. Box 35, 40014 University of Jyväskylä, Finland. 3 Animal and Plant Sciences, The University of Sheffield, Sheffield S10 2TN, UK.

ABSTRACT

The pattern of genetic variation within and among populations of a species is strongly affected by its phylogeographic history. Analyses based on putatively neutral markers provide data from which past events, such as population expansions and colonizations, can be inferred. Drosophila virilis is a cosmopolitan species belonging to the virilis group, where divergence times between different phylads go back to the early Miocene. We analysed 35 Drosophila virilis strains covering the species’ ranges in order to detect demographic events that could be used to understand the present characteristics of the species, as well as its differences from other members of the group. D. virilis showed very low nucleotide diversity with haplotypes distributed in a star-like network, consistent with a recent world-wide exponential expansion possibly associated either with domestication or post- glacial colonization. All analyses point towards a rapid population expansion over approximately 50,000 years. Coalescence models support this interpretation. The species showed no geographic structure in the distribution of mitochondrial haplotypes in contrast to results of a recent microsatellite- based analysis. The central haplotype in the network, which could be interpreted as ancestral, is widely distributed and gives no information about the geographical origin of the population expansion. 2

INTRODUCTION

The development of methods to analyse intraspecific phylogenies has provided very valuable tools to understand how populations have been influenced by historical and contemporary processes (Emerson et al., 2001; Posada & Crandall, 2001). In particular, the pattern of variation of neutral molecular markers such as mitochondrial DNA sequences, permits inferences about a species‘ demographic history, including events such as population expansions and colonizations. It has been shown that demographic expansions lead to star shaped genealogies (Slatkin & Hudson, 1991) and comparison of sequences shows an excess of rare mutations (Harpending & Rogers, 2000) and unimodal mismatch distributions (Rogers & Harpending, 1992). Spatial or range expansions lead basically to the same molecular signature if the migration rate between demes was large (Ray et al., 2003; Excoffier et al., 2004). During the Pleistocene, global climate oscillations and the associated cycles of glaciation had a profound influence on the population distributions of organisms (Hewitt, 1996; Taberlet et al., 1998; Hewitt, 2000; 2001). Range expansions and colonizations followed by demographic expansions often happened after these events, and it is possible to find the signals of those processes in contemporary populations. Drosophila virilis, a member of D. virilis group, is a domestic species of nearly cosmopolitan distribution. It has been proposed to originate somewhere in the ancient deciduous forests of China or in arid regions such as Iran or Afghanistan (Throckmorton, 1982). D. virilis group species have been used as study objects in evolutionary studies for decades, and their role in this kind of research will become even more important now that the whole genome of D. virilis has been sequenced (http://insects.eugenes.org/species/data/dvir/). D. virilis differs from other species of the group for example by its association with man-made habitats (Throckmorton, 1982), by showing no chromosomal polymorphism (Throckmorton, 1982) and by the low species-recognition and mate choice exercised by the females (e.g. Throckmorton, 1982; Saarikettu et al., 2005). Adaptation of D. virilis in different kinds of environmental conditions has been shown to have induced changes, e.g. in chill-coma tolerance (Gibert et al., 2001) and variation in alcohol dehydrogenase (Alahiotis, 1982) and heat-shock proteins (Garbush et al. 2003). Also the mobilization of transposable elements in this species has been under extensive study (e.g. Evgenev et al., 2000). In this paper, we report an analysis of the demographic history of D. virilis using mitochondrial DNA sequence data, complementing a recent analysis using microsatellites (Huttunen et al., submitted). Together, these analyses provide a historical framework for evolutionary studies on life history and behavioural traits of the species and comparisons with other species in the group, especially in D. montana. 3

MATERIALS AND METHODS

Drosophila stocks and sampling

In total, 35 Drosophila virilis strains, covering the species range, were selected for analysis (Table 1). The stocks were collected during a time period covering almost 90 years, from 1913 to flies sampled in China in 2001/2002. A single individual from each strain, either from laboratory stocks or freshly caught, was used to extract DNA and for PCR amplification of the COI and COII mitochondrial genes.

Amplification and sequencing of mitochondrial DNA

DNA was extracted from ethanol-preserved flies following a standard protocol (Sambrook et al., 1989), where the samples were homogenised in buffer and proteinase k, and DNA was extracted with chloroform-isoamyl alcohol and precipitated with isopropanol. The amplification of mitochondrial DNA was carried out with primers flanking the COII gene in the tRNALYS and tRNALEU (Liu & Beckenbach, 1992; TL2: 5’-ATGGCAGATTAGTGCAATGG-3’, TKN: 5’- GTTTAAGAGACCAGTACTTG-3’), which amplify an 850 bp fragment that includes the 688bp COII gene, and COI-1460-F: 5’- ATCTATCGCCTAAACTTCAGCC-3’ and COI-2195-R: 5’- ACTTCAGGGTGACCAAAAAATC–3’ (Simon et al,. 1994; de Brito et al. 2002) which amplify the complete 670 bp corresponding to the COI gene. PCR reactions were performed in 50Ǎl volumes including 0.5ǍM of each primer, 200ǍM dNTPs, 1.5mM MgCl2 and 1U Taq polymerase (Bioline) in reaction buffer. Initial denaturation was for 7 minutes at 94°C followed by 35 cycles of 1 minute at 94°C, 1 minute at the annealing temperature (54°C for COI and 56°C for COII) and 1 minute at 72°C, and a final incubation of 5 minutes at 72°C. The products were purified using QIAquick columns (QIAGEN) and sequenced using the forward primer. Sequences (GenBank accession nos. DQ426800 to DQ426823) were aligned with CLUSTAL-V (Higgins et al. 1992).

Mitochondrial DNA analysis of population history and phylogeography

The partition homogeneity test, PHT (Farris et al., 1995), as implemented in PAUP 4.0 was used to test for incongruence between the COI and COII data sets. There was congruence between the data-sets and therefore the two fragments were combined for all subsequent analyses. ARLEQUIN 2.0 (Schneider et al. 2000) was used to calculate pairwise distances between 4 haplotypes, the mismatch distributions and tests of the standard neutral model for a demographically stable population (Tajima’s D (1989) and Fu’s F (Fu & Li, 1993)). The program FLUCTUATE (Kuhner et al., 1998) was used to make simultaneous estimates of present day lj and the population growth rate g, assuming an exponential model of growth and using a maximum likelihood approach. The parameters used for the simulations were obtained by running a hierarchy of likelihood-ratio tests in Modeltest 3.0 (Posada & Crandall, 1998) to choose the model of evolution with the best fit to the data. Skyline plots were constructed using GENIE v. 3.0 (Pybus et al., 2000). The starting trees were obtained using maximum likelihood with molecular clock enforced. GENIE was also used to calculate the fit to different models of population growth, with fit assessed using the corrected Akaike Information Criterion. Networks of haplotypes were constructed based on statistical parsimony using the program TCS 1.06 (Clement et al., 2000).

RESULTS AND DISCUSSION

We examined a total of 35 lines covering the species range: 9 from Japan, 9 from Eastern Europe and Central Asia, 5 from Western Europe, 2 from the United States and 10 from China (Table 1). As most strains are isofemale lines possessing a single mtDNA haplotype and as it is highly unlikely that the studied mtDNA regions would have acquired new mutations during the laboratory maintenance of the strains, we sequenced only a single representative from each strain. The analyses based on haplotypes frequencies of laboratory strains (e.g. mismatch distribution) may have been slightly influenced by the scattered information from populations. However, the results suggest that this is unlikely since there is so little evidence of population structure. Laboratory strains provide the best available information on population structure of D. virilis because it has proved very difficult to collect new strains in recent years. This seems to be at least in part because the breeding sites this species (breweries, market places and timber yards) have raised hygienic standards. D virilis has rarely been collected in non-domestic situations (Throckmorton, 1982). The analysis was based on a total of 1358 base pairs, 670 corresponding to the cytochrome oxidase I and 688 to the cytochrome oxidase II gene. There was a total of 19 haplotypes and, among them, 24 nucleotide substitutions, of which 20 were transitions and 4 transversions. Nucleotide diversity (per base) was 0.00185 ± 0.00114. The haplotype network (Fig. 1), estimated using statistical parsimony, included one common haplotype, represented in six lines and present in all geographical regions except the USA. The second most frequent haplotype was found in four lines from China. The difference between these haplotypes was only two substitutions: in fact, the maximum number of 5 differences between any pair of haplotypes was only 6 substitutions. There was no apparent geographical structure in the relationships among the haplotypes: the network was star-like with low levels of mitochondrial divergence and a high frequency of unique mutations, indicating either a rapid population expansion or that selection has caused the rapid spread of a mitochondrial lineage carrying beneficial mutations. The mismatch distribution was smooth and unimodal (Figure 2a), but it departed from the expectations of the stepwise population expansion model (Harpending’s raggedness index= 0.0721, p= 0.11, fit to the stepwise growth model, SSD= 0.0157, p= 0.03). There is an excess of unique mutations compared with what is expected for a constant-size population (Table 2). We used the program FLUCTUATE to fit a model of exponential population expansion and estimate lj0, the present-day scaled population size (Table 2). The site categories and the rate of change were obtained using a maximum likelihood heuristic search in PAUP, using HKY85 as preferred model, following MODELTEST, with 5 rate categories, transition/transversion ratio of 2.3 and shape parameter of 0.105. Model comparisons using GENIE confirm that the exponential growth model was a good fit to the data and this can be seen in the skyline plot (Fig. 2b) which indicates a rapid and recent expansion (Table 2). The Drosophila virilis lines analysed in this study did not show any geographic structure of mitochondrial haplotypes. There were a very low number of substitutions differentiating haplotypes, which were commonly shared among locations. All analyses point towards a rapid population expansion over a short time-scale (~50,000 years using a mutation rate of 10-8 per year and assuming a stepwise growth model). Given the uncertainties surrounding this estimate, it is consistent with population growth following the end of the last glaciation and/or a shift into domestic environments. These are the most obvious possible causes of expansion, although they suggest population growth would have initiated less than 10,000 years ago. The central haplotype in the network appears to be ancestral (Fig. 1) but is widely distributed and so this gives no information on the geographic origin of the population expansion. In their study on sequence variation in six X-linked genes of 21 D. virilis strains from different continents, Vieira and Charlesworth (1999) found no fixed differences between the Asian strains and strains originating from Europe or North- or South-America, all the variants found outside Asia being also present in Asia, but not vice versa. These data were consistent with either a large population centred in Asia and a smaller migrant population elsewhere, or a large migrant population that went through a bottleneck (Vieira & Charlesworth, 1999). Throckmorton (1982) has previously suggested that D. virilis originated from an ancestral form in Asia, since the most primitive species of the virilis-repleta section of the genus Drosophila have been observed in Southeast Asia. In a recent study, Huttunen et al. (submitted) analysed 48 microsatellite loci in 30 D. virilis strains, many of which were analysed here for mitochondrial DNA. Although a phylogenetic tree and STRUCTURE analysis showed only moderate clustering of the strains originating from Continental Asia, Europe, 6

America and Japan, an assignment test using a priori information about the geographical origin of the strains gave high posterior probabilities for their correct assignment. Genetic variation also showed significant population differentiation, as measured by FST, with evidence for isolation by distance. Variability detected by microsatellites can have a more recent origin than variation measured by mitochondrial DNA. Consequently, the lack of population differentiation in mitochondrial DNA could indicate shared ancestry while microsatellite variation between populations indicates that differentiation is in progress because current gene flow is restricted. Alternatively, the homogeneity of mitochondrial DNA could be the result of a recent selective sweep which did not disturb the pre-existing population structure for nuclear loci. Unfortunately, the use of laboratory strains prohibits testing for a demographic expansion using the microsatellite data. The pattern of geographic variation in haplotypes of D. virilis is in sharp contrast with the pattern recently described for another species of the group, D. montana (Mirol et al., 2007). In this case there was a clear differentiation based on mitochondrial DNA and microsatellites, between lines from populations in Finland, Canada and USA. Both markers indicated the presence of at least two distinct populations, one in Eurasia and the other one representing the expansion of the species to the New World, with a divergence time estimated between them from 450,000 to 900,000 years ago, within the Pleistocene. Although D. virilis and D. montana belong to the same Drosophila group, they represent different phylads and differ in many characteristics, including chromosomal variation, habitat preferences, and courtship behaviour. Differences in biogeographic history, reflected in the pattern of mitochondrial variation in the two species, could have been important in the origin of these characteristics and constitute the basis for the interpretation of their evolution.

Acknowledgements

We are grateful to the members of the ‘Co-evolved Traits’ Research Training Network for their valuable input to the work presented here and to the European Commission for funding the network (HPRN-CT-2002-00266). Special thanks are due to Zhang Wenxia for help in collecting D. virilis flies in China. Laboratory strains of the species were obtained also from Jorge Vieira, Michael Evgen’ev and Bowling Green stock center. 7

REFERENCES

Alahiotis, S. N. 1982. Adaptation of Drosophila enzymes to temperature IV. Natural selection at the Adh locus. Genetica 59: 81-87 Clement, M., Posada, D. & Crandall, K.A. 2000. TCS: a computer programn to estimate gene genealogies. Mol. Ecol. 9: 1657-1659. de Brito, R.A., Manfrin, M.H. & Sene, F.M. 2002. Mitochondrial DNA phylogeography of Brazilian populations of Drosophila buzzatii. Genet. Mol.. Biol. 25: 161-171. Emerson, B.C., Paradis, E. & Thŝbaud, C. 2001. Revealing the demographic histories of species using DNA sequences. Trends. Ecol. Evol. 16: 707-716. Evgen'ev, M. B., Zelentsova, H., Poluectova, H., Lyozin, G. T., Veleikodvorskaja, V., Pyatkov, K. I., Zhivotovsky, L. A. & Kidwell, M. G. 2000. Mobile elements and chromosomal evolution in the virilis group of Drosophila. Proc. Natl. Acad. Sci. USA 97:11337-11342 Excoffier, L. 2004. Patterns of DNA sequence diversity and genetic structure after a range expansion: lessons from the infinite-island model. Mol. Ecol. 13: 853-864. Farris, J.S., Kallersjo, M., Kluge, A.G. & Bult, C. 1995. Testing significance of incongruence. Cladistics 10: 315-319. Fu, Y.X. & Li, W.H. 1993. Statistical tests of neutrality of mutations. Genetics 133: 693-709. Garbuz, D., Evgenev, M.B. Feder, M.F. and Zatsepina, O.G. 2003. Evolution of thermotolerance and the heat-shock response: evidence from inter/intraspecific comparison and interspecific hybridization in the virilis species group of Drosophila. I. Thermal phenotype. J. Exp. Biol. 206: 2399- 2408. Gibert, P., Moreteau, B., Petavy, G., Karan, D. & David, J.R. 2001. Chill-coma tolerance, a major climatic adaptation among Drosophila species Evolution 55: 1063-1068 Harpending, H. & Rogers, A. 2000. Genetic perspectives on human origins and differentiation. Proc. Nat. Acad. Sci. USA 96: 10597-10602. Hewitt, G.M. 1996. Some genetic consequences of ice ages, and their role in divergence and speciation. Biol. J. Linn. Soc. 58: 247-276. Hewitt, G.M. 2000. The genetic legacy of the Quaternary ice ages. Nature 405: 907-913. Hewitt, G.M. 2001. Speciation, hybrid zones and phylogeography or seeing genes in space and time. Mol. Ecol. 10: 537-549. Higgins, D., Bleasby, A., & Fuchs, R. .1992. ClustalV: Improved software for multiple sequence alignment. Comp. Appl. Biosc.. 8: 189-191. Huttunen,S., Aspi, J., Hoikkala, A:, Routtu, J. & Schlötterer, C. Variation in male courtship song traits in Drosophila virilis: the effects of selection and drift on song divergence at intraspecific level. Submitted in Behav. Genet. 8

Kuhner, M.K., Yamato, J. & Felsenstein, J. 1998. Maximum likelihood estimation of population growth rates based on coalescent. Genetics 149: 429-434 Liu, H. & Beckenbach, A.T. 1992. Evolution of the mitochondrial cytochrome oxidase II gene among ten orders of insects. Mol. Phylog. Evol. 41: 31–52 Mirol, P.M., Schäfer, M.A., Orsini, L., Routtu, J. Schlötterer, C., Hoikkala, A. & Butlin, R. 2006. Phylogeographic patterns in Drosophila montana. Mol. Ecol. 16: 1085-1097. Posada, D. & Crandall, K.A. 1998. Modeltest: testing the model of DNA substitution. Bioinformatics 14: 817-818. Posada, D. & Crandall, K.A. 2001. Intraspecific gene genealogies: trees grafting into networks. Trends. Ecol. Evol. 16: 37-45. Pybus, O.G., Rambaut, A. & Harvey, P.H. 2000. An integrated framework for the inference of viral population history from reconstructed genealogies. Genetics 155: 1429-1437. Ray, N., Currat, M. & Excoffier, L. 2003. Intra-deme molecular diversity in spatially expanding populations. Mol. Biol.. Evol. 20: 76-86. Rogers, A.R. & Harpending, H. 1992. Population growth makes waves in the distribution of pairwise genetic divergences. Mol. Biol.. Evol. 9: 552-569. Saarikettu, M., Liimatainen. J.O. & Hoikkala, A. 2005. Intraspecific variation in mating behaviour does not cause sexual isolation between Drosophila virilis strains. Anim. Behav. 70: 417-426. Sambrook, J., Fritsch, E.F. & Maniatis, T. 1989. Molecular cloning: a laboratory manual, 2nd. Ed. Cold spring Harbor Laboratory Press, New York. Schneider, S., Roessli, D. & Excoffier, L. 2000. Arlequin ver. 2.000: A software for population genetics data analysis. Genetics and Biometry Laboratory, University of Geneva, Switzerland. Simon, C., Frati, F., Beckenbach, A., Crespi, B., Liu, H. & Flook, P. 1994. Evolution, weighting and phylogenetic utility of mitochondrial gene sequences and a compilation of conserved Polymerase Chain Reaction primers. Ann. Entom. Soc. Am. 87: 651-701. Slatkin, M. & Hudson, R.R. 1991. Pairwise comparisons of mitochondrial DNA sequences in stable and exponentially growing populations. Genetics 129, 555-562. Taberlet, P., Fumagalli, L., Wust-Saucy, A.G. & Cosson J.F. 1998. Comparative phylogeography and postglacial colonization routes in Europe. Mol. Ecol. 7: 453-464. Tajima, F. 1989. Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123: 585-595. Throckmorton, L.H. 1982. The Genetics and Biology of Drosophila Vol. 3b. Academic Press, New York. Vieira, J. & Charlesworth, B. 1999. X chromosome DNA variation in Drosophila virilis. Proc. R.. Soc. Lond. B. 2666: 1905-1912. 9

TABLE 1 Lines of D. virilis used in the study, indicating, when it was available, year of collection and coordinates of the site from where the line originates.

Site Line Year Coordinates Matsuyama, Japan A11 1973 23° 50'N , 132° 46'E Matsuyama, Japan A12 1973 23° 50'N , 132° 46'E Matsuyama, Japan B15 1973 23° 50'N , 132° 46'E Matsuyama, Japan B31 1973 23° 50'N , 132° 46'E Matsuyama, Japan B42 1973 23° 50'N , 132° 46'E Sapporo, Japan SBB 1986 43° 3'N , 141° 21'E Sakata, Japan SKT 1987 38° 54'N , 139° 50'E Japan Jap. Inv. Mishima, Japan W158 35° 7'N , 138° 55'E Hangzhow, China 15010-1051.47 1948 30° 14'N , 120° 10'E Hunan, China V-Hunan 27° 23'N , 111° 30'E Zeziping, China V-ZZP-01 2001 Wuwei, China V-WW-03 2002 37° 58'N , 102° 47'E Wuwei, China V-WW-05 2002 37° 58'N , 102° 47'E Wuwei, China V-WW-08 2002 37° 58'N , 102° 47'E Lanzhou, China V-EH-01 2002 36° 6'N , 103° 39'E Dunghuang, China V-DNH Nanjing, China V-NANJING 32° 3'N , 118° 46'E Qufu, China V-QUFU 35° 36'N , 116° 57'E Russia 15010-1051.52 1976 Baku, Azerbaijan 1413 1974 40° 22'N , 49° 48'E Jalta, Ukraine 1415 1973 46° 57'N , 37° 16'E Batumi, Georgia A 41° 39'N , 41° 39'E Batumi, Georgia S9 1970 41° 39'N , 41° 39'E Yerevan, Armenia 1 40° 12'N , 44° 31'E Mzheta, Caucasus 25 Caucasus 1411 1973 Tashkent, 41° 18'N , 69° 16'E Uzbekistan 12 Holland W159 52° N, 5° E Leeds, UK LeedsA 1995 53° 47'N , 1° 32'W Leeds, UK LeedsB 1995 53° 47'N , 1° 32'W Leeds, UK 1430 1981 53° 47'N , 1° 32'W Leeds, UK 1433 1982 53° 47'N , 1° 32'W Pasadena, USA 15010-1051.0 1913 34° 8'N , 118° 8'W Truckee, USA 15010-1051.8 39° 19'N , 120° 12'W 10

TABLE 2 Summary statistics of mitochondrial variation and results from the FLUCTUATE and GENIE analyses.

FLUCTUATE GENIE n H ljS ljǑ Tajima’s D Fu’s F lj0 g AICc lj0 (95% CI) g (95% CI) Best model (second best model) 35 19 0.0044 0.0019 -2.0055 -14.311 0.0499 3532.6 Exp 149.83 0.091 3951 (0.0015) (0.0011) p<0.01 p<0.0001 (0.0103) (358.8) (Log 148.38) (0.027-0.410) (2610-5433) Estimate (standard errors): n - number of lines surveyed, H - number of haplotypes, ljS - diversity (Waterson 1975), ljǑ - diversity (Tajima 1989), Tajima’s D – neutrality test (Tajima 1989), Fu’s F - neutrality test (Fu & Li, 1993) lj0 current estimate of 2NǍ and g - scaled population growth parameter. GENIE model abbreviations: AICc – Akaike Information Criterion with a second order correction for small sample sizes, Exp – exponential growth, Log – logistic growth and CI – confidence interval. 11

FIGURE 1 Network obtained for the D. virilis haplotypes using statistical parsimony. Haplotypes are represented by ellipses, the area of the ellipse is proportional to the frequency of the haplotype, points in the lines connecting circles indicate substitutions. Shading denotes region: white - North America, pale grey – Japan, dark grey – Western Europe, black – Asia and Eastern Europe. 12

a.

180 160 140 120 100 80

Frequency 60 40 20 0 12345678 Number of pairwise differences b.

0.1

0.08

0.06

0.04

0.02

0 0 0.0002 0.0004 0.0006 0.0008 0.001 0.0012 0.0014 0.0016 0.0018 0.002

FIGURE 2 (a) Mismatch distribution. Expectations from the stepwise growth model, fitted in ARLEQUIN, are superimposed. (b) Generalised skyline plot among haplotypes of D. virilis. Observed values (solid line) and fitted values from the best model (broken line) – see Table 2. Smoothing parameter (epsilon) was 4E-5 (maximum likelihood value from option ‘maxepsilon’ in GENIE). II

VARIATION IN MALE COURTSHIP SONG TRAITS IN DROSOPHILA VIRILIS: THE EFFECTS OF SELECTION AND DRIFT ON SONG DIVERGENCE AT THE INTRASPECIFIC LEVEL

by

Susanna Huttunen, Jouni Aspi, Christian Schlötterer, Jarkko Routtu and Anneli Hoikkala

Submitted Manuscript

Variation in male courtship song traits in Drosophila virilis: the effects of selection and drift on song divergence at the intraspecific level

Susanna Huttunen*, Jouni Aspi*, Christian Schlötterer**, Jarkko Routtu*** and Anneli Hoikkala***

* Department of Biology, University of Oulu, Oulu, Finland ** Institut für Tierzucht und Genetik, Veterinärmedizinische Universität, Wien, Austria ***Department of Biological and Environmental Science, University of Jyväskylä, Jyväskylä, Finland

ABSTRACT

Genetic and phenotypic divergence of Drosophila virilis laboratory strains originating from different parts of the species range were studied with the aid of microsatellite markers and by analysing male courtship songs. The strains from America, Europe, continental Asia and Japan showed moderate geographic clustering both at the genetic level and in several traits of the male song. The genetic distances and the song divergence of the strains did not show significant association, which suggests that the songs have not diverged solely as a side-effect of genetic divergence. Comparison of the songs of the laboratory strains to those of freshly collected strains showed that pulse characters of the song are quite sensitive to culture conditions. While laboratory rearing of the flies had no effect on the number of pulses in a pulse train or the pulse train length, the tendency of the sound pulses to become longer during laboratory maintenance could explain the lack of geographic variation in pulse length and interpulse interval. Sensitivity of songs to culturing conditions should be taken in account in studies on song divergence. 2

INTRODUCTION

Evolution of species-specific mating rituals and male ornaments has long been a target of intense theoretical and empirical research. Coevolution of female preferences and male secondary sexual traits may cause rapid divergence between geographically isolated populations if females exercise selection on male traits and if both the male trait and the female preferences for it possess genetic variation (Lande 1981). Also, random genetic drift (Lande 1976) and adaptation to variable environmental conditions (Hoffmann and Merilä 1999) may give rise to population differentiation in mating signals, but their effects have been less studied than those of sexual selection. Stabilising selection, on the other hand, should lead to evolutionary stability of traits involved in the mate recognition system throughout the distribution of a species (Lambert and Henderson 1986). Population comparisons provide a powerful way to study the first steps of divergence in species-specific phenotypic characters because genetic differentiation between populations is expected to be smaller than that between the species. Intraspecific genetic and phenotypic variation has been studied e.g. in Drosophila melanogaster, in which geographically isolated populations exhibit large variation at the molecular level (Begun and Aquadro 1993; Veuille et al. 1998; Andolfatto 2001; Kauer et al. 2002) and in several morphological traits (Capy et al.1993; Long and Singh 1995; Prout and Baker 1993), as well as in traits important in the mate recognition system (Wu et al. 1995; Ferveur et al. 1996; Hollocher 1997a,b; Capy et al. 2000). Since the path breaking work of Dobzhansky and his colleques (e.g. Dobzhansky and Levene 1948), geographic variation also in species with extensive inversion polymorphisms (e.g. like D. pseudoobscura) has been under intensive study. Most studies describing the songs of Drosophila species are based on the songs of a few laboratory strains, which do not give much information on geographical variation in songs (review on studies in Hoikkala 2005). Male courtship song traits have been shown to express genetic variation in several Drosophila species (e.g. Hoikkala 1985; Ritchie et al. 1994), but geographic variation in these traits has so far been demonstrated only in D. melanogaster (Colegrave et al. 2000), D. mojavensis (Etges et al. 2006) and D. montana (Mirol et al. 2007, Routtu et al. 2007). However, many studies dealing with geographic variation in songs (e.g. Ritchie et al. 1994, Colegrave et al. 2000, Mirol et al. 2007) have been performed using fly strains maintained in laboratory for a shorter or longer period before song recording, which may have affected the results. For example, Hoikkala (1985) found the songs of old D. littoralis laboratory strains from Eurasia to differ from each other more than the songs of the fresh isofemale strains. Also, while the songs of the old laboratory strains of D. montana studied by Mirol et al. (2007) showed significant geographic variation in only the carrier frequency of the song, the songs of freshly collected strains from Finland, Canada and USA showed variation in several song traits (Routtu et al. 2007). 3

The male courtship song of D. virilis represents the conservative song type shared by other species of the virilis subgroup of the D. virilis group (Hoikkala and Lumme 1987). D. virilis females are able to recognize species-specific characters of the male song (Isoherranen et al. 1999), but contrary to the females of some other species of the group (e.g. Aspi and Hoikkala 1995) they do not require hearing the song before mating (Hoikkala 1988). Saarikettu et al. (2005) found that five D. virilis strains originating from different parts of the species distribution exhibit high inter-strain variation in mating rituals, including the male courtship song. Behavioural differences between the strains did not, however, give rise to sexual selection/isolation, as neither the males nor the females favoured the flies of their own strain over others (the males of one strain were incapable of producing song, but they had only a slight disadvantage in mating). We studied genetic and phenotypic divergence of D. virilis laboratory strains originating from different parts of the species range with the aid of 48 microsatellite markers and by analysing the male courtship songs of the strains. We assume that if the interstrain and geographic variation in male song traits has evolved solely as a side effect of the genetic divergence of the strains, the songs should reflect the geographic origin and the genetic relatedness of the strains. Higher interstrain and geographic variation in song traits compared to neutral genetic variation could be due to sexual selection for different optimums, drift and/or adaptation to varying environmental conditions. We also compared the songs of the laboratory strains to those of F3 progenies of females collected from the wild in Japan and China 2002-2003, to find out how the songs are liable to change during laboratory maintenance, and whether such changes could increase/decrease geographic variation in songs.

MATERIAL AND METHODS

Genetic variation among D. virilis strains

Genetic variation in 48 microsatellite loci (Orsini et al. 2004) was studied among 30 D. virilis strains originating from different parts of the species distribution area. The details for six of the loci (v68-74, v71-38, v68-4, v71-6, v68-86-1 and v93-93) are given in Schlötterer and Harr (2000). The remaining 42 loci are described in Huttunen and Schlötterer (2002). DNA from one individual of each strain was extracted by the method described in Miller et al. (1988). Radioactive genotyping of microsatellites followed standard protocols described in Schlötterer (1998). PCR was performed using J32P-ATP labelled primer in a 10 Pl volume with 50-100 ng of DNA, 1 PM of each primer, 200 PM of dNTP, 1.5 mM of MgCl2 and 1 U of Taq DNA Polymerase. The PCR profile was 5 min at 95qC, followed by 30 cycles of 1 4 min at 95qC, 30 s at 45-59 qC (depending on the locus), and 30 s at 72qC, and finally one cycle of 30 min at 72qC. PCR products were separated in 7% denaturing polyacrylamide gels (32 % formamide, 5.6 M urea) at 90 W and visualised by autoradiography after 12-24 hours. Allele sizes were determined by running a ‘PCR slippage ladder’ and a known size standard adjacent to the samples (Schlötterer and Zangerl 1999). Basic measures of microsatellite variability, the expected heterozygosity and variance in the repeat number (calculated by averaging over 200 randomly discarded data sets) were calculated using Microsatellite Analyzer (MSA; Dieringer and Schlötterer 2003). On the basis of multilocus microsatellite data we constructed a neighbor-joining tree using the proportion of shared alleles (Bowcock et al. 1994) as a distance measure. The tree was constructed using the Neighbor program in Phylip v. 3.6 (Felsenstein 1993) and visualised using TreeView (Page 1996). Population structure of D. virilis, as well as the patterns of possible admixture between genetically divergent groups, was investigated using the BAPS 3.2 program (Bayesian Analysis of Population Structure; Corander and Marttinen 2005). The mixture model of this program clustered the strains into four groups, America, Europe, continental Asia (Asia from hereon) and Japan, on the basis of their multilocus genotypes (see Corander et al. 2004; Corander et al. 2006). We then used the admixture analysis (see Corander and Marttinen 2006) to estimate individual strains’ coefficients of admixture (q) with regard to the previously detected groups (proportion of alleles showing ancestry with each of these groups). In this analysis, we used 100 iterations to estimate the admixture coefficients for the individuals representing different strains, 200 simulated reference individuals from each group and 20 iterations to estimate the admixture coefficients for the reference individuals, following the advice in the software manual (Corander and Marttinen 2005). The Bayesian clustering in BAPS assumes that the marker loci are unlinked and that the source populations contributing to the observed sample are in Hardy–Weinberg equilibrium (e.g. Corander et al. 2003). Because these assumptions are not necessarily appropriate in our laboratory populations, the results of this analysis should be interpreted with caution. Finally, we estimated the proportion of total variance within and between the four geographic groups based on hierarchical variance of allele frequencies (Excoffier et al.1992) using AMOVA (Arlequin 2.000; Schneider et al. 2000). FST values (Weir and Cockerham 1984) were estimated for each locus separately and for pair-wise group comparisons with MSA. The significances of these values were determined by permuting genotypes among groups. 5

Recording and analysing the male courtship songs

The male courtship songs were recorded for the above-mentioned 30 strains originating from different parts of the species distribution area, for five additional strains from China and for F3 progenies of females collected from China (four progenies) and Japan (13 progenies). All flies were maintained in the culture room for at least one generation before recording their song (continuous light, 70 % humidity and 19°C temperature). The flies were sexed at the age of two days (or less) and the males and the females were stored separately in vials containing malt medium (Lakovaara, 1969) until they were sexually mature (14 ± 2 days). The songs of five males per strain were recorded in a single-pair courtship at 20 r 1°C between 9 and 12 am, and analysed for six song characters. Details of the recording methods are described e.g. in Saarikettu et al. (2005). Data for the songs of 12 of the 35 laboratory strains have been published earlier (Huttunen et al. 2002; Saarikettu et al. 2005; see Appendix I). The male courtship song of D. virilis consists of dense pulse trains with short pauses between successive sound pulses normally only at the beginning of the pulse train (see Figure 1). Due to the irregularity of the first sound pulses of the pulse trains and the fact they usually were of lower intensity and included less sound cycles than the rest of the pulses (see e.g. Campesan et al. 2001), they were not included in pulse structure analysis. The songs of different fly strains were analysed with the SIGNAL Sound Analysis System (©Engineering Design) by measuring the lengths of the pulse trains (PTL) and by counting the number of pulses in each train (PN) from the oscillograms. The traits for single sound pulses (i.e. CN = number of cycles in a pulse, PL = length of a pulse and IPI = length of an inter pulse interval) were measured for the fourth sound pulse of each pulse train. Due to the dense structure of pulse trains the IPIs of most strains were of the same length as PLs (see Figure 1 and Table II). The carrier frequency (FRE) was measured from the Fourier spectra of the analysed pulse trains. For each strain the means of different song traits were calculated over the means of five males (the mean for each male was first calculated over the traits measured for three pulse trains). Geographical variation in the songs was studied with a multivariate MANCOVA test using the geographic origin of the strain (i.e., the four groups: America, Europe, Asia and Japan) as a factor and the age of the strain as a covariate. Variation in male song traits was partitioned to differences within and among the four geographic groups with nested ANOVA. We also performed a Mantel (1967) test with 10000 permutations (Genepop; Raymond and Rousset 1995) to examine the relationship between genetic and song distances between the strains. Song divergence between each pair of strains was estimated using an Euclidean distance measure based on the four song characters showing geographic variation (CN, FRE, PN and PTL). 6

The effect of laboratory maintenance on male song traits was studied by comparing the songs of the 35 strains maintained in the laboratory for 10-58 years with those of the F3 progenies of females collected in Japan and China 2002-2003, using a discriminant analysis.

RESULTS

Patterns of genetic variation among the strains

Genetic variation in 48 microsatellite loci was studied in 30 laboratory strains of D. virilis. All loci, except tra, were polymorphic among the studied strains, the mean number of alleles per locus ranging from 3 to 16 (average 9). The average heterozygosity of polymorphic loci was 0.69 and the mean variance in repeat number was 13.32. Geographic structuring of D. virilis was examined by constructing a Neighbor-joining tree of individual strains of this species and using seven D. lummei strains (Orsini et al. 2004) as an outgroup. This tree showed moderate geographic clustering among the D. virilis strains originating from America, Europe, Asia and Japan (Figure 2). The Bayesian mixture analysis suggested the same grouping and the admixture analysis confirmed the levels of admixture between the genomes of the strains of these groups to be very low. Only three strains showed significant admixture to a different group: one Asian strain (1051.47) had admixture coefficients of 0.31 to the American group (i.e. 0.31 of its microsatellite alleles showed common ancestry with the American strains) and 0.07 to the European group and two Japanese strains showed admixture to the American group (strain B22, admixture coefficient of 0.24) or to the European group (strain W158, admixture coefficient of 0.18). FST-statistics calculated from the microsatellite data showed significant differentiation between American, European, Asian and Japanese strains, varying from 0.058 to 0.140 (Table I). The lowest differentiation was detected between American and Japanese strains and the highest was between European and Asian strains. The global FST value over all polymorphic loci was 0.097 and highly significant (p < 0.001). Significant differences between the strains from the four localities were detected at 22 out of 47 polymorphic loci (p < 0.05). Variance components were highly significant both between and within the geographic groups (AMOVA; 1.169 and 9.034, respectively, p < 0.001; 1000 permutations). Microsatellite variation between the groups explained 11.5 % of the total variation, whereas most of the variation (88.5%) was within the geographic groups (see Table III). 7

Variation in male song traits among the D. virilis strains

The 35 laboratory strains of D. virilis (30 strains with microsatellite information plus five additional strains from China) showed high variation in male song traits, especially in the pulse train characters, PN and PTL (see Appendix I). The mean number of pulses per train (PN) varied from 6.2 to 12.6 and mean pulse train length (PTL) varied from 142 ms to 270 ms. Among the pulse characters, the mean number of sound cycles per pulse (CN) varied from 4.9 to 6.4, mean carrier frequency (FRE) varied from 242 Hz (A13) to 311 Hz (strain 59) and mean pulse length (PL) and interpulse interval (IPI) varied from 18.4 ms to 21.6 ms. Variation among the songs of the laboratory strains was significant after sequential Bonferroni correction in PN (F34=10.17, p<0.001), PTL (F34=8.94, p<0.001), CN (F34=3.70, p<0.001) and FRE (F34=7.49, p<0.001), but not in PL (F34=1.38, p=0.10) and IPI (F34=1.29, p=0.15).

Divergence of D. virilis strains at genetic vs. phenotypic (song) levels

Geographical variation in the songs of the 30 laboratory strains for which genetic data were available was studied with a multivariate MANCOVA –test using the geographic origin (America, Europe, Asia and Japan) of the strain as a factor and the age of a strain (10 to 58 years, see Appendix I) as a covariate. The song characters CN, FRE, PN and PTL showed significant variation between the geographic groups (MANCOVA; F-transformed Wilks lambda; F18, 255=5.69, p<0.001), and the age of the strain had a significant negative effect on FRE and CN (F-transformed Wilks lambda; F6,90= 4.106, p= 0.01). The means of the strains from America, Europe, Asia and Japan showed significant variation in CN (F3=5.63, p<0.001), FRE (F3=17.08, p<0.001), PN (F3=9.06, p<0.001) and PTL (F3=8.23, p<0.001) even when the effect of the age was controlled. The American and Japanese strains showed lower values in all the above-mentioned song traits than the strains from Asia and Europe (Table II). In IPI the proportion of variation between groups was of the same level as in PTL (see Table III), but it remained nonsignificant, since it was mainly caused by three strains with longer IPIs than PLs (1051.8 from America, A12 from Japan and 1430 from Europe). Variation in male song traits of D. virilis laboratory strains was partitioned into differences within and among the four geographic groups with nested ANOVA to compare the magnitudes of the two variance components with those measured for genetic variance with AMOVA. PL showed no between- group variation, whereas most of the other song characters showed two- to nearly four-fold higher variance between the groups compared to genetic variance in microsatellite markers (Table III). Variance among the four 8 geographic groups was highest in the two negatively correlated song traits, FRE and CN. The divergence in male song traits of the D. virilis strains was compared to the genetic distances of the respective strains to investigate if song evolution could have occurred as a side-effect of genetic divergence. The genetic distances (the proportion of shared alleles) and the song distances between the strains did not show significant association (Mantel test; r = 0.105, p = 0.714).

Changes in male song traits due to laboratory maintenance

Conditions in culture bottles are quite different from those in market places, breweries and timber yards, where D. virilis is found in the wild. To study the effects of laboratory rearing on male song traits, we compared the songs of the 35 strains maintained in the laboratory for 10-58 years (see Appendix I, ) with those of the 17 F3 progenies of females collected in Japan and China 2002-2003. Figure 3 shows a scattergram of a discriminant analysis of the overall song variation for the mean values of each song trait of all 52 D. virilis strains (laboratory strains and the F3 progenies). The first and second discriminant axes (DA) accounted for 72.4% and 20% of the variance, respectively (Table IV). Pulse length (PL) and inter pulse interval (IPI) were highly associated with the 1st DA, whereas the number of cycles (CN) showed highest correlation with the 2nd DA. The old laboratory strains showed neither interstrain nor geographic divergence in IPI and PL, but the songs of the F3 progenies from Japan and China had clearly shorter IPIs and PLs than the laboratory strains. The lack of significant interstrain/geographic divergence in PL and IPI in laboratory strains may, indeed, have been caused by the tendency of sound pulses to get longer during laboratory maintenance. The second discriminant factor showed that the Japanese (both recently established and old strains) and American strains have in general lower CNs and FREs than the strains from Asia and Europe (see also Table II).

DISCUSSION

Evolution of male secondary sexual characters playing a role in sexual selection within the species and/or in species recognition may have a major impact on speciation. The present study shows that variation in male song traits in D. virilis has not arisen purely as a side-effect of genetic divergence of populations, even though the females of this species show only weak selection on male song (Saarikettu et al. 2005). Another important finding is that the pulse characters of the song are quite sensitive to culture conditions, which has to be taken into account in studies on song divergence. 9

Throckmorton (1982) suggested that D. virilis originated in Asia, where the most primitive species of the virilis-repleta section have been found. Vieira and Charlesworth (1999) found evidence for this in their study of sequence variation in six X-linked genes of 21 D. virilis strains from different continents; all the genetic variants found outside Asia were also present in Asia, but not vice versa. Mirol et al. (submitted) analysed mitochondrial DNA sequence data (COI and COII genes) of 35 D. virilis strains and found no geographic structure in haplotype distribution. The species showed very low nucleotide diversity with haplotypes distributed in a star-like network, consistent with a recent world- wide exponential expansion possibly associated with either domestication or post-glacial colonisation. However, although our phylogenetic analysis based on microsatellites did not reveal the ancestry of the strains, it did show moderate geographic clustering. The fact that D. virilis strains showed population structure in microsatellite but not in mtDNA analysis is not surprising, since microsatellite variation usually has a more recent origin than mitochondrial DNA variation. The relationship between genetic and quantitative phenotypic variation is frequently analysed by comparing FST values for genetic variation with an analogous measure, QST (Spitze 1993), for phenotypic traits. This method enables one to infer the effects of selection on phenotypic characters assuming that allelic effects at each locus are additive (reviewed by Merilä and Crnokrak 2001; Mackay and Latta 2002). Our D. virilis data are not applicable for this kind of analysis because the data consist mainly of old laboratory strains and the number of groups to be compared is quite small (four). Instead, we used a multivariate analysis to test the divergence of the strains at the genetic and phenotypic (song) level, and Mantel tests to study the relationship between genetic and song distances. The main finding in these comparisons was that geographic variation in most male song traits exceeds the amount of genetic variation between the strains. Furthermore, the genetic distances and the song distances of D. virilis strains did not show significant association, i.e. the songs had not diverged solely as a side-effect of the genetic divergence of the strains. This kind of trend has been found in several animal species (see e.g. Butlin and Tregenza 1998; Tregenza et al. 2000), with phenotypic traits being sensitive especially to sexual selection. For example, the songs of laboratory strains of D. montana, in which the females exercise strong selection on male song (e.g. Aspi and Hoikkala 1995), show high interstrain and geographic variation in several traits (Routtu et al. 2007). During laboratory maintenance, the quality of male song traits could have been affected by drift, sexual selection, inbreeding and changed environmental conditions. In crowded culture bottles the females have a variety of mating partners to choose from, which in principle could lead to increased sexual selection. The songs of the males of laboratory strains could have longer sound pulses due to changes in muscle physiology taking place during adaptation to laboratory conditions and/or due to inbreeding depression (see Aspi 2000). The fact that changes in song traits during laboratory maintenance had occurred in 10 all D. virilis laboratory strains in the same direction, i.e. towards longer PLs and IPIs, suggests that the changes have not been caused by random genetic drift. It is also worth noting that laboratory maintenance had a significant effect only on the pulse characters of the song, and not on the pulse train characters PN and PTL. Studies of interspecific hybrids between D. virilis species with unique courtship songs have shown that X chromosomal genes have an important role in the evolution of species-specific song (e.g. in allowing the IPIs to become longer than PLs in most species of the montana phylad; Hoikkala and Lumme 1987; Päällysaho et al. 2003). At the intraspecific level, song variation in D. virilis is determined mainly by autosomal genes (Huttunen and Aspi 2003; Huttunen et al. 2004). In a recent biometric analysis of song differences between two D. virilis strains diverging in PN and PTL (strains 1431 and B22) Huttunen and Aspi (2003) suggested significant additive and dominance components and a significant additive interaction between maternal and progeny genotypes. The direction of dominance in PN was towards a lower number of pulses and in PTL it was towards shorter pulse trains. This direction is different from that observed at the species level in the virilis phylad species (Hoikkala and Lumme 1987), among which D. virilis has the shortest pulse trains. In the present study the lowest PNs and the shortest PTLs were found in American and Japanese strains, which might represent derived populations, if D. virilis originated in continental Asia as Throckmorton (1982) suggested. In our study PL and IPI of the D. virilis song (which had the same mean value in most study strains) appeared not to vary significantly among the geographic groups. Lack of intra- and interspecific variation in the length of the interpulse interval (IPI) has been observed also in other Drosophila species, e.g. in D. ananassae and D. pallidosa, and the phenomenon was suggested to be due to selection on other species-specific song parameters requiring constant IPI (Yamada et al. 2002). The fact that the pulse characters of the song may change during laboratory rearing should be taken into account in song studies. While the songs of the laboratory strains give important information on factors affecting male song evolution and also provide good material for genetic crosses, their use in studies on the genetic and geographic variation in male songs may lead to false interpretations (see introduction). These studies should always be done within a natural population framework. 11

Acknowledgements

Many thanks to Zhang Wenxia, Hisaki Takamori, Michael Evgen’ev and Jorge Vieira and Bowling Green stock center for providing the D. virilis strains. Special thanks to Mari Saarikettu for help in rearing and sexing the flies for song recording and Max Kauer, Jaana Liimatainen, Emily Knott and Maaria Kankare for discussions and critiques. This work was supported by grants from the Finnish Cultural Foundation, Emil Aaltonen Foundation and University of Oulu to S. H. and from the Academy of Finland to A.H (project 44960 and the Centre of Excellence in Evolutionary Research). Co-operation between laboratories has been supported by the EU Research Training Network ‘Genetic analysis of complex co-evolved behavioural traits’ (HPRN-CT-2002-00266). 12

REFERENCES

Andolfatto P (2001) Contrasting patterns of X-linked and autosomal nucleotide variation in Drosophila melanogaster and Drosophila simulans. Mol Biol Evol 18:279-290 Aspi J (2000) Inbreeding and outbreeding depression in male courtship song characters in Drosophila montana. Heredity 84:273-282 Aspi J, Hoikkala A (1995) Male mating success and survival in the field with respect to size and courtship song characters in Drosophila littoralis and D. montana (Diptera: Drosophilidae). J Insect Behav 8:67-87 Begun D, Aquadro CF (1993) African and North American populations of Drosophila melanogaster are very different at the DNA level. Nature 365:548- 550 Bowcock AM, Ruiz-Lineares A, Tonfohrde J, Minch E, Kidd JR, Cavall-Sforza LL (1994) High resolution of human evolutionary trees with polymorphic microsatellites. Nature 368:55-457 Butlin RK, Tregenza T (1998) Levels of genetic polymorphism: marker loci versus quantitative traits. Phil Trans R Soc Lond B 353:187-198 Campesan S, Dubrova Y, Hall J, Kyriacou CP (2001) The nonA gene in Drosophila conveys species-specific behavioral characteristics. Genetics 158:1535-1543. Capy P, Pla E, David JR (1993) Phenotypic and genetic variability of morphometrical traits in natural populations of Drosophila melanogaster and D. simulans. I. Geographic variation. Genet Sel Evol 25:517-536 Capy P, Veuille M, Paillette M, Jallon J-M, Vouibidio J, David JR (2000) Sexual isolation of genetically differentiated sympatric populations of Drosophila melanogaster in Brazaville, Congo: the first step towards speciation? Heredity 84:468-475 Colegrave N, Hollocher H, Hinton K, Ritchie MG (2000) The courtship song of African Drosophila melanogaster. J Evol Biol 13:143-150 Corander J, Marttinen, B (2005) BAPS: Bayesian Analysis of Population Structure. Manual v. 3.2. Department of Mathematics, University of Helsinki Corander J, Waldmann P, Marttinen P, Sillanpää M J (2004) BAPS 2: enhanced possibilities for the analysis of genetic population structure. Bioinformatics 20:2363-2369 Corander J, Marttinen, P (2006) Bayesian identification of admixture events using multilocus molecular markers. Mol Ecol 15:2833-2843 Corander J, Marttinen P, Mäntyniemi S (2006) A Bayesian method for identification of stock mixtures from molecular marker data. Fish Bull 104:550-558 Dieringer D, Schlötterer C (2003) Microsatellite analyser (MSA): a platform independent analysis tool for large microsatellite data sets. Mol Ecol Notes 3:167-169 13

Dobzhansky Th, Levene H. (1948) Genetics of Natural Populations. Xvii. Proof of Operation of Natural Selection in Wild Populations of Drosophila Pseudoobscura. Genetics 33:537–547. Etges WJ, Over KF, Cardoso de Oliveira C, Ritchie, M (2006) Inheritance of courtship song variation among geographically isolated populations of Drosophila mojavensis. Anim Behav 71:1205-1214 Excoffier L, Smouse P, Quattro J (1992) Analysis of molecular variance inferred from metric distances among DNA haplotypes: Applications to human mitochondrial DNA restriction data. Genetics 131:479-491 Felsenstein J (1993) PHYLIP phylogeny inference package, version 3.5c. Technical report. University of Washington, Seattle Ferveur JF, Cobb M, Boukella H, Jallon J-M (1996) World-wide variation in Drosophila melanogaster sex pheromone: behavioural effects, genetic bases and potential evolutionary consequences. Genetica 97:73-80 Hoffmann AA, Merilä J (1999) Heritable variation and evolution under favourable and unfavourable conditions. Trends Ecol Evol 14:96-101 Hoikkala A (1985) Genetic variation in the male courtship sound of Drosophila littoralis. Behav Genet 15:135-142 Hoikkala A (1988) The importance of different courtship stimuli in the mating behaviour of European species of the Drosophila virilis group. Ann Zool Fennici 25:257-263 Hoikkala A (2005) Inheritance of male sound characteristics in Drosophila species. In: Drosopoulos S, Claridge MF (eds) Insect sounds and communication: physiology, behaviour, ecology and evolution, Taylor & Francis, pp 167-177. Hoikkala A, Lumme J (1987) The genetic basis of evolution of the male courtship sounds in the Drosophila virilis group. Evolution 41:827-845 Hollocher H, Ting C-I, Pollack F, Wu C-I (1997a) Incipient speciation by sexual isolation in Drosophila melanogaster: variation in mating preference and correlation between sexes. Evolution 51:1175-1181 Hollocher H, Ting C-I, Wu M-L, Wu C-I (1997b) Incipient speciation by sexual isolation in Drosophila melanogaster: extensive genetic divergence without reinforcement. Genetics 147:1191-1201 Huttunen S, Aspi J (2003) Complex inheritance of male courtship song characters in Drosophila virilise. Behav Genet 33: 17-24 Huttunen S, Schlötterer C (2002) Isolation and characterisation of microsatellites in Drosophila virilis and their cross species amplification in members of the D virilis group. Mol Ecol Notes 2:593-597 Huttunen S, Aspi J, Hoikkala A, Schlötterer C (2004) QTL analysis of variation in male courtship song characters in Drosophila virilis. Heredity 92:263-269 Huttunen S, Vieira J, Hoikkala A (2002) Nucleotide repeat variation at the nonA gene of the Drosophila virilis group species and its effects on male courtship song. Genetica 15:159-167 14

Isoherranen E, Aspi J, Hoikkala A (1999) Variation and consistency of female preferences for simulated courtship songs in Drosophila virilis. Anim Behav 57:619-625 Kauer M, Zangerl B, Dieringer D, Schlötterer C (2002) Chromosomal patterns of microsatellite variability contrast sharply in African and non-African population of Drosophila melanogaster. Genetics 160:247-256 Lakovaara S (1969) Malt as a culture medium for Drosophila species. Dros Inf Serv 44:128. Lambert DM, Henderson NR (1986) The stability of the specific-mate recognition system of Drosophila melanogaster. Behav Genet 16:369-373 Lande R (1976) Natural selection and random genetic drift in phenotypic evolution. Evolution 30:214-334 Lande R (1981) Models of speciation by sexual selection on polygenic traits. Proc Natl Acad Sci USA 78:3721-3725 Long AD, Singh RS (1995) Molecules versus morphology: the detection of selection acting on morphological characters along a cline in Drosophila melanogaster Heredity 74:569-581 Mantel N (1967) The detection of disease clustering and a generalized regression approach. Cancer Res 27:209-220 McKay JK, Latta RG (2002) Adaptive population divergence: markers, QTL and traits. Trends Ecol Evol 17:285-291 Merilä J, Crnokrak P (2001) Comparison of genetic differentiation at marker loci and quantitative traits. J Evol Biol 14:892-903 Miller SA, Dykes DD, Polesky HF (1988) A simple salting out procedure for extracting DNA from human nucleated cells. Nucleic Acids Res 16:1215 Mirol PM, Schäfer MA, Orsini L, Routtu J, Schlötterer C, Hoikkala A, Butlin R (2007) Phylogeographic patterns in Drosophila montana. Mol Ecol 16:1085- 1097 Mirol, PM, Routtu, J, Hoikkala, A, Butlin, R Signals of demographic expansion in Drosophila. Submitted Orsini L, Huttunen S, Schlötterer C (2004) A multilocus microsatellite phylogeny of the Drosophila virilis group. Heredity 93:161-165 Page RDM. 1996. TREEVIEW: An application to display phylogenetic trees on personal computers Comput Appl Biosci 12:357-358 Prout T, Baker JSF (1993) F statistics in Drosophila buzzattii: selection, population size and inbreeding. Genetics 134:369-375 Päällysaho S, Aspi J, Liimatainen JO, Hoikkala A (2003) The role of X chromosomal song genes in the evolution of species-specific courtship songs in Drosophila virilis group species. Behav Genet 33:25-32 Raymond M, Rousset F (1995) GENEPOP (version1.2): population genetics software for exact tests and ecumenicism. J Heredity 86:248-249 (http://wbiomed.curtin.edu.au/genepop) Ritchie MG, Yate VH, Kyriacou CP (1994) Genetic variability of the interpulse interval of courtship song among some European populations of Drosophila melanogaster. Heredity 71:59-464 15

Routtu J, Mazzi D, Van der Linde K, Mirol P, Butlin R, Hoikkala A (2007) The extent of variation in male song, wing and genital characters among allopatric Drosophila montana populations. J EvolBiol, in press (Doi: 10.1111/j.1420-9101.2007.01323.x) Saarikettu M, Liimatainen JO, Hoikkala A (2005) Intraspecific variation in mating behaviour does not cause sexual isolation between Drosophila virilis strains. Anim Behav 70:417-426 Schlötterer C (1998) Microsatellites. In: Hoelzel AR (ed) Molecular Genetic Analysis of Populations: a Practical Approach, Oxford University Press, pp. 237-261 Schlötterer C, Zangerl B (1999) The use of imperfect microsatellites for DNA fingerprinting and population genetics. In: Epplen JT, Lubjuhn T (eds) DNA profiling and DNA fingerprinting. Birkhäuser, Basel, pp153-165 Schlötterer C, Harr B (2000) Drosophila virilis has long and highly polymorphic microsatellites. Mol Biol Evol 17:1641-1645 Schneider S, Roessli D, Excoffier L (2000) Arlequin ver. 2.000: A software for population genetics data analysis. Genetics and Biometry Laboratory University of Geneva Switzerland Spitze K (1993) Population structure in Daphnia obtusa – quantitative genetic and allozymic variation. Genetics 135:367-374 Throckmorton LH (1982) The virilis species group. In: Ashburner M, Carson HL, Thompson JN Jr (eds) The Genetics and Biology of Drosophila, vol. 3b. Academic Press, London, pp. 227-296 Tregenza T, Pritchard VL, Butlin RK (2000) Patterns of trait divergence between populations of the meadow grasshopper Chorthippus parallelus. Evolution 54:574-585 Veuille M, Bènassi V, Aulard S, Depaulis F (1998) Allele-specific population structure of Drosophila melanogaster Alcohol dehydrogenase at the molecular level. Genetics 149:971-981 Vieira J, Charlesworth B (1999) X chromosome DNA variation in Drosophila virilis. Proc R Soc Lond B 266:1905-1912 Weir BS, Cockerham CC (1984) Estimating F-statistics for the analysis of population structure. Evolution 38:1358-1370 Wu CI, Hollocher H, Begun DJ, Aquadro CF, Xu Y, Wu ML (1995) Sexual isolation in Drosophila melanogaster: a possible case of incipient speciation. Proc Natl Acad Sci USA 92:2519-2523 Yamada H, Sakai T, Tomaru M, Doi M, Muneo M, Oguma Y (2002) Search for species-specific mating signal in courtship songs of sympatric sibling species Drosophila ananassae and D. pallidosa. Genes Genet Syst 77: 97-106 16

TABLE 1 Pairwise population differentiation (FST values obtained by 1000 permutations; Weir & Cockerham, 1984) among D. virilis strains grouped according to their geographic origin.

America Asia Europe Asia 0.101NS Europe 0.118** 0.140*** Japan 0.058NS 0.087*** 0.069** *** = p < 0.001; ** = p < 0.01; NS= not significant

TABLE 2 The means ± standard deviations within geographic groups in six male song characters.

Geographic N PN PTL PL IPI CN FRE group America 4 8.9 ± 0.8 191.3 ± 14.5 20.2 ± 1.3 20.8 ± 0.6 5.3 ± 0.1 262 ± 10.1 Europe 7 10.6 ± 1.6 228.1 ± 33.9 19.9 ± 0.8 20.1 ± 0.8 5.9 ± 0.4 295 ± 8.4 Asia 9 9.8 ± 1.0 207.8 ± 23.4 19.6 ± 0.5 19.7 ± 0.5 5.7 ± 0.4 292 ± 13.5 Japan 10 8.7 ± 1.1 191.5 ± 26.0 20.2 ± 0.8 20.3 ± 0.7 5.4 ± 0.3 271 ± 14.1 N = the number of strains originating from each geographic area. PN = number of pulses in a train, PTL = length of the pulse train, PL = length of a pulse, IPI = length of an inter pulse interval, CN = number of cycles in a pulse and FRE = carrier frequency of the song.

TABLE 3 The proportion of variance between and within the geographic groups of D. virilis (America, Europe, Asia and Japan) in six male courtship song traits from nested ANOVA and in variance components derived from 48 microsatellite loci using AMOVA (Excoffier et al. 1992).

PN PTL PL IPI CN FRE Microsatellites Between 30.6 23.9 0 25.0 40.3 41.0 11.5 groups Within 69.4 76.1 100.0 75.0 59.7 59.0 88.5 groups PN = number of pulses in a train, PTL = length of the pulse train, PL = length of a pulse, IPI = length of an inter pulse interval, CN = number of cycles in a pulse and FRE = carrier frequency of the song. 17

TABLE 4 Pooled within-groups correlations between discriminating variables and standardized canonical discriminant functions for each song character for the 1st and 2nd discriminant axis (DA).

PN PTL PL IPI CN FRE DA1 -0.436 -0.237 0.625 0.747 0.042 -0.502 DA2 0.241 0.354 0.507 0.523 0.956 0.524 PN = number of pulses in a train, PTL = length of the pulse train, PL = length of a pulse, IPI = length of an inter pulse interval, CN = number of cycles in a pulse and FRE = carrier frequency of the song.

PL IPI PTL 600 ms

FIGURE 1 Oscillogram of the male courtship song of D. virilis. PL = the length of a pulse, IPI = the length of an interpulse interval and PTL = the length of a pulse train. In the songs of most strains PLs and IPIs were of equal length. 18

D. lummei Ŷ B42 Ŷ A11 Ŷ B31 Ŷ A12 Ŷ A13

Ŷ 1051.9

1051.48 Ÿ ł 25 W157 Ÿ ł 12 ł 59 LeedsA ź ł S9 1431 ź ł A 1433 ź ł 1 1430 ź

ł 1415 1432 ź LeedsB SKT SBB ź Ŷ Ŷ W158 1051.8 W159 Hunan B22 Ŷ Ÿ 1051.47 1051.49 ź ł Ŷ ł Ÿ

FIGURE 2 Neighbor-joining tree of D. virilis strains from America (ʆ), Asia (ŏ), Europe (ʈ) and Japan (ʄ) using D. lummei (7 strains, Orsini et al. 2004) as an outgroup species. Analysis was based on 48 microsatellite loci, with distances measured as 1 – (the proportion of shared alleles) (Bowcock et al. 1994). 19

4

3

2

1

0

discriminant axis (20%) discriminant -1

nd + New strains (23) 2 Europe (7) -2 America (4) Japan (10) -3 Asia (14) -6-4-20246

1st discriminant axis (72.4%)

FIGURE 3 Discriminant analysis on male courtship song characters of D. virilis. The number of strains in each group is shown in parenthesis. New strains refer to the progenies of females recently collected in Japan and China; others are old laboratory strains from different geographic areas. 20

Appendix I Geographical origin and collection year (if known) of D. virilis strains and the means ± standard deviations of their song traits. Number of males analysed for each strain is five. Abbreviations: PN = number of pulses in a pulse train, PTL = length of a pulse train, PL = length of a pulse, IPI = length of an inter pulse interval, CN = number of cycles in a pulse and FRE = carrier frequency of the song.

Strain Strain Origin Collection PN PTL PL IPI CN FRE Number Year Strains with microsatellite data 1 15010-1051.8a, b Truckee, California, USA - 9.1 ± 0.8 188 ± 26 18.4 ± 1.3 21.0 ± 0.7 5.1 ± 0.2 261 ± 20 2 15010-1051.9a, b Sendai, Japan - 9.1 ± 0.3 212 ± 9 19.6 ± 2.0 19.6 ± 2.0 5.2 ± 0.5 261 ±19 3 15010-1051.47a, b Hangchow, China 1948 8.8 ± 0.6 192 ± 18 20.2 ± 2.3 20.2 ± 2.3 5.9 ± 0.9 285 ± 9 4 15010-1051.48a, b Texmelucan, Puebla, Mexico 1947 8.6 ± 0.4 194 ± 9 21.4 ± 1.1 21.4 ± 1.1 5.3 ± 0.2 259 ± 8 5 15010-1051.49a, b Chaco, Argentina 1950 9.9 ± 0.3 209 ± 4 20.0 ± 1.0 20.0 ± 1.0 5.4 ± 0.4 276 ± 11 6 Ac Batumi, Georgia - 9.0 ± 0.7 187 ± 24 19.6 ± 0.9 19.6 ± 0.9 5.6 ± 0.4 289 ± 11 7 A11c Matsuyama, Ehime, Japan 1973 9.9 ± 0.8 205 ± 24 20.4 ± 1.1 20.4 ± 1.1 5.4 ± 0.4 266 ± 14 8 A12c Matsuyama, Ehime, Japan 1973 8.3 ± 0.8 168 ± 17 18.8 ± 1.6 19.2 ± 0.8 5.0 ± 0.5 264 ± 24 9 A13c Matsuyama, Ehime, Japan 1973 7.5 ± 1.0 168 ± 16 21.6 ± 1.1 21.6 ± 1.1 5.1 ± 0.4 242 ± 16 10 B22c Matsuyama, Ehime, Japan 1986 6.2 ± 0.2 142 ± 13 20.4 ± 1.5 20.4 ± 1.5 5.9 ± 0.7 294 ± 10 11 B31c Matsuyama, Ehime, Japan 1986 8.9 ± 0.7 189 ± 15 19.4 ± 1.1 19.4 ± 1.1 5.5 ± 0.5 285 ± 17 12 B42c Matsuyama, Ehime, Japan 1986 9.1 ± 0.8 190 ± 11 20.2 ± 1.8 20.2 ± 1.8 5.1 ± 0.5 271 ± 12 13 SBBc Sapporo, Hokkaido, Japan 1986 10.1 ± 0.5 228 ± 13 20.4 ± 1.8 20.4 ± 1.8 5.4 ± 0.4 273 ± 14. 14 SKTc Sakata, Yamagata, Japan 1987 9.1 ± 0.8 199 ± 21 20.6 ± 1.8 20.6 ± 1.8 5.7 ± 0.3 274 ± 16 15 S9b Batumi, Georgia 1970 10.0 ± 0.5 222 ± 17 20.0 ± 1.2 20.0 ± 1.2 5.9 ± 0.2 299 ± 13 16 Hunand Hunan, China - 9.0 ± 0.4 180 ± 10 19.4 ± 0.9 19.4 ± 0.9 5.9 ± 0.2 304 ± 12 17 LeedsAb Leeds, England 1995 11.2 ± 1.3 240 ± 32 20.4 ± 1.5 20.8 ± 1.6 6.3 ± 0.4 305 ± 9 18 LeedsBb Leeds, England 1995 8.0 ± 0.8 175 ± 23 20.0 ± 1.6 20.0 ± 1.6 5.8 ± 0.7 289 ± 15 19 W157b Mexico - 8.0 ± 1.1 174 ± 27 20.8 ± 1.3 20.8 ± 1.3 5.2 ± 0.2 252 ± 10 20 W158b Mishima, Japan - 8.9 ± 1.8 214 ± 32 20.8 ± 2.4 21.0 ± 2.1 5.5 ± 0.7 278 ± 2 21 W159b Holland - 10.7 ± 1.6 245 ± 31 21.2 ± 0.8 21.2 ± 0.8 6.4 ± 0.3 289 ± 3 22 1b Erevan, Caucasus, Armenia - 9.5 ± 0.5 195 ± 8 19.5 ± 1.1 19.5 ± 1.0 5.4 ± 0.4 280 ± 9 23 12b Tashkent, Middle Asia - 10.9 ± 0.5 234 ± 12 19.4 ± 1.1 19.4 ± 1.1 5.5 ± 0.5 287 ± 26 24 25b Mzheta, Caucasus - 8.9 ± 0.5 184 ± 14 19.4 ± 1.1 19.4 ± 1.1 5.3 ± 0.4 270 ± 15 25 59b Seishel Islands 1986 10.2 ± 1.6 219 ± 32 20.2 ± 1.3 20.2 ± 1.3 6.3 ± 0.4 311 ± 14 26 1415b Jalta, Russia 1973 10.3 ± 0.7 222 ± 18 19.2 ± 2.8 19.2 ± 2.8 5.7 ± 0.7 303 ± 11 27 1430b England (53q N, 1qE) 1981 12.6 ± 0.6 270 ± 20 19.4 ± 0.6 20.4 ± 1.5 5.4 ± 0.6 285 ± 11 28 1431b England (53q N, 1qE) 1981 12.0 ± 0.7 253 ± 16 19.4 ± 0.9 19.4 ± 0.9 6.0 ± 0.6 297 ± 12 29 1432b England (53q N, 1qE) 1981 9.3 ± 0.8 193 ± 18 18.8 ± 0.5 18.8 ± 0.5 5.5 ± 0.2 293 ± 7 30 1433b England (53q N, 1qE) 1982 10.4 ± 1.7 221 ± 28 19.8 ± 1.8 19.8 ± 1.8 5.8 ± 0.4 307 ± 17 Strains without microsatellite data 1 TOY3F2 Toyama, Japan 2003 10.3 ± 1.3 209 ±31 17.4 ±1.3 17.4 ±1.3 5.1 ±0.4 290 ±23 2 TOY3F3 Toyama, Japan 2003 9.7 ± 0.6 198 ± 14 16.8 ± 1.1 16.8 ± 1.1 5.0 ± 0.4 302 ± 20 3 TOY3F6 Toyama, Japan 2003 12.2 ± 0.8 244 ± 19 17.2 ± 1.3 17.2 ± 1.3 5.4 ± 0.4 306 ± 12 4 TOY3F7 Toyama, Japan 2003 10.7 ± 0.6 208 ± 14 17.8 ± 1.1 17.9 ± 1.1 5.4 ± 0.5 305 ± 18 5 TOY3F9 Toyama, Japan 2003 10.5 ± 2.0 207 ± 40 17.6 ± 0.6 17.6 ± 0.6 5.3 ± 0.2 298 ± 9 6 TOY3F10 Toyama, Japan 2003 10.3 ± 1.2 209 ± 31 17.4 ± 1.3 17.4 ± 1.3 5.1 ±0.4 290 ± 23 7 TOY3F11 Toyama, Japan 2003 10.3 ± 1.0 201 ± 23 17.4 ± 1.1 17.4 ± 1.1 5.1 ± 0.6 292 ± 30 8 TOY3F12 Toyama, Japan 2003 10.6 ± 0.8 219 ± 14 17.0 ± 1.0 17.0 ± 1.0 4.8 ± 0.3 278 ± 25 9 TOY3F15 Toyama, Japan 2003 9.5 ± 0.4 185 ± 13 16.8 ± 0.8 16.8 ± 0.8 5.1 ± 0.1 298 ± 14 10 TOY3F16 Toyama, Japan 2003 9.1 ± 1.2 180 ± 23 18.2 ± 1.1 18.2 ± 1.1 5.4 ± 0.4 298 ± 10 11 TOY3F17 Toyama, Japan 2003 11.6 ± 1.3 242 ± 27 19.8 ± 1.3 19.8 ± 1.3 5.4 ± 0.3 275 ± 19 12 TOY3F19 Toyama, Japan 2003 11.6 ± 0.7 219 ± 24 17.6 ± 0.9 17.6 ± 0.9 5.5 ± 0.4 316 ± 31 13 TOY3F20 Toyama, Japan 2003 9.8 ± 0.3 188 ± 6 17.6 ± 0.9 17.6 ± 0.9 5.4 ± 0.5 303 ± 24 14 V-EH-01 Lanzhou, China 2002 11.7 ± 1.3 249 ± 31 19.8 ± 1.5 19.8 ± 1.5 5.4 ±0.4 273 ± 23 15 V-WW-03 Wuwei, China 2002 10.7 ± 1.0 222 ± 30 19.4 ± 0.9 19.4 ± 0.9 5.3 ± 0.3 278 ± 12 16 V-WW-05 Wuwei., China 2002 11.2 ± 1.7 233 ± 2 18.4 ± 1.5 18.4 ± 1.5 5.5 ± 0.3 290 ± 20 17 V-WW-08 Wuwei, China 2002 10.5 ± 1.1 226 ± 30 20.2 ± 1.1 20.2 ± 1.1 5.7 ± 0.5 279 ± 12 18 V-DNH Dunghuang, , China - 10.3 ± 0.5 216 ± 22 19.6 ± 1.1 19.6 ± 1.1 5.9 ± 0.3 300 ± 18 19 V-HUNAN Hunan Prov., China - 9.9 ± 0.9 217 ± 18 20.6 ±0.6 20.6 ± 0.6 6.4 ± 0.3 310 ± 12 20 V-NANJIN Nanjing, China - 10.9 ±1.3 218 ± 32 18.7 ± 1.3 18.8 ± 1.1 4.9 ± 0.6 268 ± 17 21 V-QUFU Qufu, China - 7.9 ± 1.3 170 ± 29 19.8 ± 1.1 19.8 ± 1.1 5.9 ± 0.4 292 ± 19 22 V-ZZP-01 Zeziping, China - 11.7 ± 1.3 253 ± 27 19.4 ± 2.0 19.4 ± 2.0 5.7 ± 0.4 290 ± 16 a Strains obtained from Bowling Green Drosophila Stock Center, b Song data published in the present study, c Song data published in Huttunen 21 et al. (2002), d Song data published in Saarikettu et al. (2005).

III

PHYLOGEOGRAPHIC PATTERNS IN DROSOPHILA MONTANA

by

Paricia M. Mirol, Martin A. Schäfer, Luisa Orsini, Jarkko Routtu, Christian Schlötterer, Anneli Hoikkala and Roger K. Butlin 2007

Molecular Ecology 16: 1085-1097

Reprinted with kind permission of Blackwell Publishing

Molecular Ecology (2007) 16, 1085–1097 doi: 10.1111/j.1365-294X.2006.03215.x

PhylogeographicBlackwell Publishing Ltd patterns in Drosophila montana

P. M. MIROL,* M. A. SCHÄFER,† L. ORSINI,†** J. ROUTTU,‡ C. SCHLÖTTERER,† A. HOIKKALA‡ and R. K. BUTLIN§ *School of Biology, The University of Leeds, Leeds LS2 9JT, UK, †Institut für Tierzucht und Genetik, Josef Baumann Gasse 1, Universitat Innsbruck, Institut für Okologie, 1210 Wien, Austria, ‡Department of Biological and Environmental Sciences, PO Box 35, 40014 University of Jyväskylä, Finland, §Animal and Plant Sciences, The University of Sheffield, Sheffield S10 2TN, UK

Abstract The Drosophila virilis species group offers valuable opportunities for studying the roles of chromosomal re-arrangements and mating signals in speciation. The 13 species are divided into two subgroups, the montana and virilis ‘phylads’. There is greater differentiation among species within the montana phylad in both karyotype and acoustic signals than exists among members of the virilis phylad. Drosophila montana is a divergent species which is included in the montana phylad. Here, we analyse the phylogeography of D. montana to provide a framework for understanding divergence of acoustic signals among populations. We analysed mitochondrial sequences corresponding to the cytochrome oxidase I and cytochrome oxidase II genes, as well as 16 microsatellite loci, from 108 lines of D. montana covering most of the species’ range. The species shows a clear genetic differentiation between North American and Scandinavian populations. Microsatellite allele frequencies F and mitochondrial DNA haplotypes gave significant ST values between populations from Canada, USA and Finland. A Bayesian analysis of population structure based on the microsatellite frequencies showed four genetically distinct groups, corresponding to these three populations plus a small sample from Japan. A network based on mitochondrial haplotypes showed two Finnish clades of very different shape and variability, and another clade with all sequences from North America and Japan. All D. montana populations showed evidence of demographic expansion but the patterns inferred by coalescent analy- sis differed between populations. The divergence times between Scandinavian and North American clades were estimated to range from 450 000 to 900 000 years with populations in Canada and the USA possibly representing descendants of different refugial populations. Long-term separation of D. montana populations could have provided the opportunity for differentiation observed in male signal traits, especially carrier frequency of the song, but relaxation of sexual selection during population expansion may have been necessary. Keywords: demography, Drosophila montana, microsatellites, mitochondrial DNA, phylogeography Received 9 July 2006; revision received 26 September 2006; accepted 30 October 2006

three lineages, montana, littoralis and kanekoi (Spicer Introduction 1992; Spicer & Bell 2002). Although most studies published The Drosophila virilis group comprises 13 species and to date support the existence of the four groups, the subspecies divided into two clades, the virilis and montana phylogenetic relationships between species within the phylads. The montana phylad is further subdivided into groups are more contentious, varying with the marker used to infer the phylogeny (Spicer 1991, 1992; Nurminsky et al. 1996; Spicer & Bell 2002; Adrianov et al. 2003). Within Correspondence: Patricia Mirol, Museo Argentino de Ciencias the virilis phylad, the Palearctic endemics D. virilis and Naturales, CONICET, Angel Gallardo 470, C1405DJR, Buenos Aires, D. lummei are clearly defined. However, relationships among Argentina. Fax: 0054 11 49824494; E-mail: [email protected] **Current address: Metapopulation Research Group, Department North American members of the phylad are confounded of Biological and Environmental Sciences, PO Box 65, FIN-00014 by shared ancestral polymorphisms in mitochondrial DNA University of Helsinki, Finland. (mtDNA) sequences (Caletka & McAllister 2004), nuclear

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd 1086 P. M. MIROL ET AL. genes (Hilton & Hey 1996, 1997) and microsatellites (Orsini differences (e.g. Hoikkala & Lumme 1987), variation in et al. 2004). male song traits and female responses within the species Throckmorton (1982) suggested that the phylads diverged (e.g. Ritchie et al. 2005), and sexual selection exercised by in the Early Miocene, or not later than the Pliocene, when females on male song (e.g. Aspi & Hoikkala 1995; Ritchie both of them entered the New World by way of Beringia. et al. 1998). Divergence times between the phylads have been calcu- The pattern of genetic variation within and among lated using different markers, and vary from 7 million populations of a species is strongly affected by its phylo- years ago (Ma) using restriction fragments of mitochon- geographic history. Particularly strong signatures might be drial genes (Ostrega 1985), to 9 Ma according to sequences expected for species occupying formerly glaciated regions of the Adh gene (Nurminsky et al. 1996) and 11 Ma based (Hewitt 2001) and for domesticated species (Cymbron et al. on mitochondrial 12S and 16S ribosomal RNA genes (Spicer 2005; Larson et al. 2005; Pedrosa et al. 2005). It is important & Bell 2002). In general, species within the montana to know this history if one wishes to interpret patterns of phylad have evolved more in terms of chromosomal re- variation in traits such as mating signals and responses arrangements, and are also more variable regarding the (cf. Tregenza et al. 2000). Analyses based on putatively neutral number of inversions segregating within populations markers such as mtDNA sequences or microsatellites pro- than members of the virilis phylad (Throckmorton 1982), vide a baseline against which the effects of selection can be although species in this group show a higher number of tested as well as providing data from which past events, such fusions. The primitive karyotype is found in D. virilis which, as population expansions and colonizations, can be inferred. in contrast to other species of the group, shows no inversion These events may have been responsible for changes in selec- polymorphisms. This contrast is interesting in terms of the tion pressures that underlie variation in mating behaviour, possible role of chromosomal inversions in speciation for example. In this study, we propose phylogeographic (Butlin 2005). Also, species of the virilis phylad have higher and demographic scenarios for one of the species of the crossabilities among taxa within the phylad than species virilis group, D. montana, as a basis for studies of the evolu- of the montana phylad (Throckmorton 1982). tionary history of songs and preferences. We also provide a All species of the D. virilis group differ in the acoustic preliminary analysis of song variation within this species. mating signals (‘songs’) produced by males during court- ship. These male courtship songs, produced by wing vibra- Materials and methods tion, play an important role both in species recognition (Liimatainen & Hoikkala 1998) and in sexual selection Drosophila stocks and sampling within species in the wild (Aspi & Hoikkala 1995). Song characteristics vary much more among species of the In total, 108 Drosophila montana strains, covering the species’ montana phylad than they do among virilis phylad range, were selected for analysis (see Supplementary species. Also, the importance of courtship song varies material). The stocks were collected during a time period among species: Drosophila montana females rarely accept covering 50 years, from 1947 to flies sampled in Finland the courtship of a ‘mute’ (wingless) male, whereas D. virilis in 2001/2002. In addition, three new wild populations females readily accept this kind of courtship (Hoikkala of D. montana were sampled in 2003, in Colorado (USA), 1988; Hoikkala et al. 2005). Vancouver (Canada) and Oulanka (Finland). New isofemale Divergence in mating signals and responses between strains were established from wild caught individuals populations contribute to prezygotic reproductive isola- from these populations. A single individual from each tion. There is abundant evidence that this type of barrier to strain, either from laboratory stocks or freshly caught, was gene flow is critical for the existence of many animal and used to extract DNA and for polymerase chain reaction plant species and that it evolves early in the process of (PCR) amplification of the COI and COII mitochondrial speciation (Coyne & Orr 2004). Divergence in signals and genes. A different individual from a subset of strains was responses may be due to sexual selection (Panhuis et al. used to extract DNA and amplify microsatellite markers, 2001) or to reinforcement (Servedio & Noor 2003) but it under the assumption that individuals within each strain could also be an incidental by-product of divergence are genetically homogeneous. due to ecological selection pressures or to drift. Progress in understanding the evolution of signals and responses Amplification and sequencing of mitochondrial DNA requires documentation of patterns of variation both within and among species, and of the genetic basis of this variation. DNA was extracted from ethanol-preserved flies following The D. virilis group has been a productive model system a standard protocol (Sambrook et al. 1989), where the for the analysis of the contribution of acoustic mating signals samples were homogenized in buffer and proteinase k, and to reproductive isolation (Liimatainen & Hoikkala 1998; DNA was extracted with chloroform-isoamyl alcohol Saarikettu et al. 2005), the genetic basis of interspecific and precipitated with isopropanol. The amplification of

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd PHYLOGEOGRAPHY OF DROSOPHILA MONTANA 1087 mitochondrial DNA was carried out with primers flank- Mitochondrial DNA analysis ing the COII gene in the tRNALYS and tRNALEU (Liu & Beckenbach 1992; TL2: 5′-ATGGCAGATTAGTGCAATGG- The partition homogeneity test (PHT), as implemented 3′, TKN: 5′-GTTTAAGAGACCAGTACTTG-3′), which in paup 4.0 was used to test for incongruence between amplify an 850-bp fragment that includes the 688 bp COII the COI and COII data sets. The test is based in the gene. Three other fragments corresponding to the mito- incongruence-length difference test by Farris et al. (1995). chondrial genes ND5 and COI, and to the AT-rich control The null hypothesis is that the two loci are no more region were also amplified in a subsample of strains in incongruent than two randomly generated partitions of order to assess their variability and select one of them for equal size. One hundred replicates were generated and the inclusion in the analysis. The variability found in COI was P value obtained was 0.23, which indicates congruence considered suitable and this second fragment was amplified between the data sets. Therefore, the two fragments were and sequenced for the entire sample. The primers used combined for all subsequent analyses. were COI-1460-F: 5′-ATCTATCGCCTAAACTTCAGCC-3′ Pairwise distances between haplotypes were estimated and COI-2195-R: 5′-ACTTCAGGGTGACCAAAAAATC- in arlequin 2.0 (Schneider et al. 1997). Due to the low level 3′ (Simon et al. 1994; de Brito et al. 2002) which amplify the of diversity, no corrections were made for multiple sub- complete 670 bp corresponding to the COI gene. PCRs stitutions. Analysis of population genetic structure was were performed in 50 μL volumes including 0.5 μm of carried out using analysis of molecular variance (amova) μ each primer, 200 m dNTPs, 1.5 mm MgCl2 and 1 U Taq in arlequin. amova takes into account the number of polymerase (Bioline) in reaction buffer. Initial denaturation molecular differences between haplotypes in an analysis was for 7 min at 94 °C followed by 35 cycles of 1 min at of variance framework equivalent to F statistics, with 94 °C, 1 min at the annealing temperature (54 °C for COI significance tested by permutation. and 56 °C for COII) and 1 min at 72 °C, and a final The frequency distributions of the numbers of segregat- incubation of 5 min at 72 °C. The products were purified ing sites in all possible pairwise comparisons, known using QIAquick columns (QIAGEN) and sequenced using as mismatch distributions, were calculated in arlequin. the forward primer. Sequences (GenBank Accession nos Slatkin & Hudson (1991) demonstrated that the mismatch DQ 426717 to DQ 426799) were aligned with clustal v distributions of a stable population have a ragged profile (Higgins et al. 1992). due to stochastic lineage loss. In contrast, an exponentially growing population has a smooth unimodal distribution approaching a Poisson distribution. This reflects a star- Microsatellite typing like genealogy in which all of the coalescent events Individuals were genotyped for a total of 16 microsatellite occurred in a short period of time. arlequin allows markers. For PCR amplification, DNA was extracted from comparisons of observed mismatch distributions with a single individual from each strain using a high-salt those expected at equilibrium in a stable population or extraction protocol (Miller et al. 1988). PCR reactions were after a sudden expansion at scaled time T (= 2 μT genera- μ θ μ θ μ performed in a final volume of 10 L containing 50–100 ng tions) from scaled size 1 (= 2N1 females) to 0 (= 2N0 of genomic DNA, 1 μm of each primer (forward end- females). 32 μ labelled with P), 200 m dNTPs, 1.5 mm MgCl2 and 1 U arlequin was also used to conduct tests of the standard Taq polymerase following standard protocols (Schlötterer neutral model for a demographically stable population. 1998). The amplification consisted of 30 cycles with 50 s Tajima’s D-test (1989) compares two estimators of the ° ° θ μ θ θ θ at 94 C, 50 s at 45–62 C (depending on locus), and 50 s at population parameter = 2N , π and S. π is based on 72 °C. We applied an initial denaturing step of 3 min at the average pairwise number of differences between ° ° θ 94 C and a terminal extension of 45 min at 72 C, allowing sequences (Tajima 1989) and S is estimated from the for a quantitative terminal transferase reaction. PCR pro- number of segregating sites in a population (Waterson ducts were separated on a 7% denaturing polyacrilamide 1975). The F-test of selective neutrality by Fu and Li (1993) gel (32% formamide, 5.6 m urea) and visualized by auto- evaluates the probability of observing a random sample radiography after 12–24 h. Allele sizes were determined with a number of alleles similar to, or smaller than the running a ‘PCR slippage ladder’ and a known size standard observed value, given the observed number of pairwise adjacent to the samples (Schlötterer & Zangerl 1999). differences, taken as an estimator of θ. The F statistic, Detailed information on primer sequences, repeat motifs especially, is very sensitive to recent fluctuations in and PCR conditions for the loci Mon17a, 21, 23, 25, 26, 29, effective population sizes. In general, negative values of 30a, 30b and 31 can be found in Orsini et al. (2004). The Tajima’s D and Fu’s F significantly different from zero indi- remaining seven loci have been developed recently and are cate a population demographic expansion while positive not yet published (Mon17b, 33a, 34, 36, 37, 39). Detailed values indicate contraction. However, selection may lead information on these loci will be provided on request. to similar patterns.

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd 1088 P. M. MIROL ET AL.

for significance by permuting genotypes among populations Population history and phylogeography and among groups (FST), by permuting genotypes among The program fluctuate (Kuhner et al. 1998) was used to make populations but within groups (FSC) and by permuting θ simultaneous estimates of present-day and the population populations among groups (FCT). growth rate g, assuming an exponential model of growth We used baps 2.0 (Corander et al. 2003) to test for popu- and using a maximum-likelihood approach. The parameters lation substructure. baps 2.0 estimates hidden population used for the simulations were obtained by running a substructure based on multilocus genotypes and on the hierarchy of likelihood-ratio tests in modeltest 3.0 (Posada geographical sampling information in a Bayesian statistical & Crandall 1998) to choose the model of evolution with the framework. We ran 10 000 updates after a burn-in phase of best fit to the data. Skyline plots were constructed using 5000 iterations. Three independent runs were performed to genie version 3.0 (Pybus et al. 2000). The starting trees were test the robustness of our results. obtained using maximum likelihood with molecular clock enforced. genie was also used to calculate the fit to Song recording and analysis different models of population growth, with fit assessed using the corrected Akaike information criterion (AIC). For song recording, a virgin sexually mature male and female The phylogenetic relationships between species were (18–24 days old) of the same strain were transferred into a inferred in paup 4.0b10 using maximum parsimony, mating chamber (a Petri dish with diameter 5.5 cm, height distance-based methods and maximum likelihood. The 1.3 cm). The roof of the chamber was made of nylon mesh, best-fitting model of nucleotide substitution for the and the floor covered with a moistened filter paper. Male maximum-likelihood analyses was selected using model- songs were recorded with a JVC condenser microphone, test 3.0, as above. Maximum parsimony and maximum- which was kept above the courting flies. Recordings were likelihood heuristic searches were conducted with 1000 made with a Sony TC-FX33 cassette recorder between 0800 random sequence addition replicates. Because the classic and 1200 at a temperature of 20 ± 1 °C. phylogenetic methods are not directed toward analysis Song analysis was carried out with the SIGNAL Sound of intraspecific data, we constructed networks based on Analysis System. Male songs were analysed by measuring statistical parsimony using the program tcs 1.06 (Clement the lengths of the pulse trains (PTL) and by counting the et al. 2000). Phylogenetic methods assume that ancestral number of pulses per train (PN) from the oscillograms. haplotypes are no longer present; yet coalescent theory Pulse length (PL), interpulse interval (IPI; the time from the predicts that ancestral haplotypes may be the most frequent beginning of one pulse to the beginning of the next pulse) sequences sampled in a population level study. Statistical and the number of cycles in a pulse (CN) were measured parsimony is particularly useful to estimate robust networks for the fourth sound pulse of each pulse train. Carrier when few nucleotide differences exist among haplotypes frequency (FRE) was measured from the Fourier spectra and it assigns outgroup weights to haplotypes, allowing of the pulse trains. For each strain, the means of different hypothesis testing about geographical origin. song traits were calculated over the songs of five males per strain (three pulse trains per male). Mantel tests were used to investigate the relationship among genetic distances Microsatellite analysis between strains, as measured by mtDNA haplotype and Genetic differentiation between D. montana populations song characters. Analysis of variance was used to test was calculated using F statistics according to Weir & for song differentiation among geographic regions. We Cockerham (1984) where F estimates are weighted com- did not attempt to establish a correlation between micro- putations of the F coefficients of Wright (1978). Statistical satellite and song data because the number of samples for significance of FST values was tested by 10 000 permutations which we have both kind of information was small. of genotypes among populations. This conservative procedure does not assume Hardy–Weinberg equilibrium Results and allows for linkage among loci. We applied the sequential Bonferroni correction procedure to account for Species-level analysis multiple testing (Sokal & Rohlf 1995). These calculations were performed with version 3.12 of the microsatellite- A neighbour-joining tree (Fig. 1) based on the com- analyser software (Dieringer & Schlötterer 2003). bined COI and COII sequence data for Drosophila montana In order to quantify the amount of genetic variation and three other species of the virilis group, Drosophila resulting from differentiation between continents relative virilis (five strains from the UK, China, Japan and USA), to that resulting from geographical separation within Drosophila littoralis (isofemale lines established from continents, we performed an amova using arlequin 2.0 populations in Finland and in Portugal) and Drosophila (Schneider et al. 1997). Variance components were tested borealis (one isofemale line established from flies captured

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd PHYLOGEOGRAPHY OF DROSOPHILA MONTANA 1089

Fig. 1 Neighbour joining tree of mtDNA COI and COII haplotypes. F: Finland, P: Portugal, NA: North America, J: Japan. Numbers at the nodes are bootstrap values based on 1000 replicates. Only a random sample of sequences from Drosophila montana was included.

in Manitoba, Canada), is in agreement with topologies wild individuals captured in 2002 and 2003. The sam- described for these species previously (Throckmorton ples covered North America [Canada (C) and USA (U)], 1982; Nurminsky et al. 1996; Spicer & Bell 2002). D. borealis Finland (F) and Japan (J). The analysis is based on a total and Drosophila montana are sister species. Within D. littoralis, of 1358 base pairs, 670 corresponding to the COI and 688 to there is no detectable differentiation between the localities the COII. The number of different haplotypes obtained included, Portugal and Finland. The D. montana samples was 72. There were no significant differences between from Finland occupy an ancestral position within the laboratory strains established before 2002 and the recently species. Maximum likelihood and parsimony topologies established lines, or between years for regions where direct were similar to the neighbour-joining tree. comparisons were possible (F 2002 vs. F 2003: FST 0.02479, P = 0.592; F laboratory strains vs. F 2002/2003: FST = 0.00055, P = 0.356; U laboratory strains vs. U 2003: Mitochondrial variation in Drosophila montana FST = 0.00906, P = 0.195; C laboratory strains vs. C 2003: We examined a total of 108 D. montana lines, including FST = 0.03864, P = 0.085). Therefore, all individuals from many laboratory strains and new lines established from the same geographic region were pooled for subsequent

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd 1090 P. M. MIROL ET AL.

Table 1 amova results for the genetic differentiation among being highest in the lines from the USA and lowest in the Drosophila montana populations between and within continents lines from Canada. based on 16 microsatellite loci (upper lines) and mitochondrial Figure 2 shows the network created with statistical DNA (lower line) parsimony. The lines from Finland are clearly differentiated from the other populations, while lines from the remaining Source of Variance Percentage variation d.f. SS components of variation regions are mixed together. Despite the significant differ- entiation detected by amova, haplotypes from the USA Among C 2 95.262 0.848* 16.53 and Canada do not occupy separate clades in the network. 2 4.188 0.0612 12.51 The Finnish samples constitute two differentiated clades. Among P 3 38.866 0.450*** 8.78 One of them (Fig. 2) contains a sequence constituting the within C 3 1.606 0.011 2.17 central node, which is also the most frequent haplotype — Within P 138 528.469 3.829 74.69 present in 24 lines out of 54 examined — and the one with 87 36.604 0.421 85.33 the highest outgroup weight (0.12). We need to be careful about this result since there are many more samples from Fixation indices — microsatellites: FST = 0.248, FSC = 0.100; Φ Φ mitochondrial DNA: ST: 0.14672, P < 0.0001; SC: 0.02476, Oulanka than from any other single population. However, Φ P = 0.12805; CT: 0.12506, P = 0.02151. C, continents (Finland, the position of the Finnish samples in the species-level Japan, North America); P, populations; *P < 0.05; **P < 0.001. tree (Fig. 1) also confirms the basal placement of these haplotypes. This is the only clade with this type of topology, where most of the haplotypes are derived analyses. Note that three lines from Alaska were included from the root and differ in only one or two substitutions. within the Canadian sample following a geographic rather The Finnish haplotypes differ from all others by four than political criterion. shared substitutions. Genetic differentiation among lines from Europe There are three positions where all but six of the Finnish (Oulanka and Kemi), North America (Utah, Colorado and lines differ from the other regions. These six lines (five Vancouver) and Japan (Kawasaki) was tested using amova from Oulanka and one from Lappajarvi) are unusual since Φ (Table 1). The ST value was significant (0.147, P < 0.0001). they not only share six nucleotide substitutions with lines The differentiation between regions was also significant from Canada, USA and Japan, but also have eight positions Φ ( CT: 0.12506, P < 0.03). Pairwise comparisons (Table 2) where they have transitions not present in any of the other showed significant differentiation in samples from different lines. They constitute the second clade from Finland, populations within the North America sample, with the composed of highly divergent haplotypes. Some of the population from Vancouver being significantly different haplotypes in this group are separated from each other by from the populations in USA (Utah and Colorado), while up to 13 substitutions, and from haplotypes in the other no significant FST was obtained between these last two. Finnish clade by more than 20 substitutions. This can be Haplotype diversity varied depending on the region (Table 3) seen clearly in the network diagram (Fig. 2).

Table 2 Pairwise FST values (Weir & Cockerham 1984) among Drosophila montana populations (lower triangular matrix) and corresponding P values (upper triangular matrix) for microsatellites (upper line) and mitochondrial DNA (lower line)

Pairwise

FST n Oulanka Kemi Kawasaki Utah Colorado Vancouver

Oulanka 24 — 0.642 < 0.001B*** 0.002B** < 0.001B*** < 0.001B*** 45 0.662 < 0.02B* < 0.001B*** 0.002B** < 0.001B*** Kemi 14 −0.005 — < 0.001B*** 0.005 < 0.001B*** < 0.001B*** 5 −0.038 0.208 < 0.02B* < 0.04B* < 0.004B*** Kawasaki 4 0.366 0.317 — 0.057 0.018 < 0.001B*** 3 0.177 0.178 < 0.04B* 0.999 0.368 Utah 3 0.182 0.128 0.275 — 0.075 0.577 9 0.159 0.146 0.017 0.210 < 0.02B* Colorado 5 0.371 0.333 0.404 0.173 — < 0.001B*** 8 0.149 0.134 0.001 0.014 < 0.04B* Vancouver 22 0.162 0.135 0.255 −0.007 0.215 — 23 0.163 0.153 0.042 0.042 0.035

Significant P values after Bonferroni correction (B) are marked; n, sample size.

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd PHYLOGEOGRAPHY OF DROSOPHILA MONTANA 1091 250 − C (95% CI) , scaled population g

; α (95% CI) μ N

g (95% CI) , current estimate of 2 0 θ es.

0 (0.015–0.054) (151–564) θ (95% CI) model abbreviations: Exp, exponential growth; Log, logistic , diversity (Tajima 1989); π θ genie results AICc best model (second best model) genie (Log 140.26) analyses (b) genie SE) 22.27 Log 155.42 0.0105 48280 — 10 − G ( and results , diversity (Waterson 1975); S θ (SE) fluctuate 0 fluctuate θ

Fig. 2 Network obtained for the D. montana haplotypes using F statistical parsimony. The area of the figure representing the = 0.244 (0.0014) (42.67) (Exp 153.15) (—) (6820–72 400) (—) < 0.001 (0.0079) (90.49) < 0.001 (0.0279) (64.73) (Expan 191.42) (0.096–48.8) (3830–12 600) (—) 9.8345 0.0374 364.51 Exp 141.46 0.027 370 — — 2.4148 0.0103 Fu’s P − P P − haplotype is proportional to its frequency, points in the lines connecting circles indicate substitutions. The biggest circle

D represents the haplotype with the highest outgroup probability.

, number of haplotypes; Filled circles, Finland; open circles, Japan; open squares, Canada; H open triangles, USA. = 0.156 = 0.21 = 0.110 0.8429 1.2267 — 0.0986 1231.2 Pexp 193.23 0.534 6490 0.0073 — 14.7992 0.9848 Tajima’s P − P − − P −

There is one clade corresponding to the lines from the USA, Canada and Japan. It is a large clade with variable branch π θ lengths, where most of the haplotypes are unique, differing from each other by a few substitutions. The most common haplotype is represented by eight individuals out of 60. , number of lines surveyed; S θ n Although lines from Canada and USA did not constitute different clades in the network analysis, separate fluctu-

, scaled population size before growth as a proportion of current size; C, shape parameter. ate and mismatch distribution analyses were conducted α because of the evidence (above) for population structure 23 21 0.0096 0.0075 34 28 0.0065 0.0043 54 25 0.0086 0.0061 nH within North America. However, to increase statistical power, we pooled sequences within regions and added Summary statistics of mitochondrial variation (a) and results from the additional lines that could not be included in the amova because they were single representatives of their popu- Table 3 FinlandEstimate (standard errors): (0.0026) (0.0029) growth parameter; Pexp, piecewise expansion; Expan, expansion. Where confidence limits are not given, they were at the default parameter boundari D. montana USA (0.0034) (0.0037) D. montana Canada (0.0022) (0.0023) D. montana Region lations (lines from Lappajarvi, Kemi and Oulu were

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd 1092 P. M. MIROL ET AL.

Fig. 4 Generalized skyline plots for D. montana from the USA (a), Canada (b) and Finland (c). Observed values (solid line) and fitted Fig. 3 Mismatch distributions among haplotypes of Drosophila values from the best model (broken line) — see Table 3b. Smoothing montana from the USA (a), Canada (b) and Finland (c). Expectations parameters (epsilon) were: (a) 0.00102, (b) 0.00041, (c) 0.00036 from the stepwise growth model, fitted in arlequin, are superimposed. (maximum likelihood values from option ‘maxepsilon’ in genie).

included in the analysis of the Finnish sample, lines from 95% confidence interval from 6.765 to 13.250, Fig. 3a); Idaho, Wyoming, Yukon and Nevada were added to Utah it does not differ significantly from the expectations of and Colorado to form the USA sample and lines from the stepwise expansion model (SSD = 0.0063, P = 0.86) and Alaska, Ontario and Quebec were added to the Canadian the raggedness index is also very low (r = 0.00761, P = 0.97). sample with strains from Vancouver; see Supplementary The exponential model fitted by the fluctuate analysis material). modeltest favoured the HKY85 model, with also suggests growth and model comparison using genie a transition/transversion ratio of 7.76 and a shape param- finds the exponential model to be the best fit to the data, with eter (alpha) of 0.8434, and this was used in fluctuate. The similar estimates for growth rates and current population mismatch distribution for the USA sample is smooth and sizes (Table 3). The skyline plot (Fig. 4a) illustrates the good unimodal (mean number of pairwise differences 8.912, fit of the exponential expansion model for this population.

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd PHYLOGEOGRAPHY OF DROSOPHILA MONTANA 1093

The mismatch distribution for the Canadian sample Table 4 Mitochondrial FST (above the diagonal) and P values (mean number of pairwise differences 5.597, 95% con- (below the diagonal) for morphological subtypes of Drosophila fidence interval from 2.897 to 9.123, Fig. 3b) is more ragged montana. Significant values are shown in bold (r = 0.0134, P = 0.91) but is consistent with a stepwise 3-Alaska- expansion (SSD = 0.0106, P = 0.73). A recent stepwise 1-Finnish 2-Standard Canadian 4-Giant 5-Japanese expansion is also indicated by the skyline plot (Fig. 4b), with expansion around 35 000 years ago given a mutation 1— 0.518 0.517 0.392 0.485 rate of 10−8. The preferred model in genie analysis was the 2 0.000 — 0.032 0.342 0.136 piece-wise expansion model (i.e. exponential expansion 3 0.000 0.186 — 0.344 0.164 from a stable ancestral population) (Table 3). The lines 4 0.035 0.043 0.000 — 0.245 0.007 from Finland also have a ragged mismatch distribution 5 0.057 0.081 0.101 — (r = 0.0245, P = 0.98) and are consistent with stepwise expansion (SSD = 0.0210, P = 0.47) but with much higher maximum divergence (mean number of pairwise differ- ences 9.743, 95% confidence interval from 2.629 to 17.673, or completeness of isolation influences the extent of song Fig. 3c). The skyline plot (Fig. 4c) is very similar to the differentiation. We obtained song measurements for a Canadian sample, except that the ancestral population was subset of the strains: seven lines from Finland, nine from somewhat larger and with a longer history. It is not clear America and three from Japan (see Supplementary mater- why the logistic model has the highest AIC in the genie ial). There were significant Mantel correlations between analysis since it does not appear to be a good fit to the data. genetic distance and pulse train length (PTL) or carrier All three samples of D. montana yield negative estimates frequency (FRE), although did not remain significant of Tajima’s D and Fu’s F, which is consistent with the after Bonferroni correction. Only carrier frequency (FRE) inferences of population expansion. However, only the varied significantly among geographical regions (F3,32 = 5.43, F estimates for the Canadian and US samples differ signi- P = 0.0039) although variation in cycle number was also ficantly from zero (Table 3). nearly significant (F3,32 = 2.90, P = 0.050). The variation in FRE, which remains significant after Bonferroni correction, was mainly due to the low frequency of the song of Finnish Associations between phenotypic and mitochondrial strains (236.7[16.7]Hz, mean[among strain standard devi- variation ation]) relative to the other regions (USA, 261.7[22.6]Hz; North American D. montana has been traditionally divided Canada, 268.0[90]Hz; Japan, 252.9[17.4]Hz). into three subtypes, Giant, Standard and Alaskan–Canadian, according to inversion frequency data, body size and Microsatellite variation geographic location (see Throckmorton 1982). Finnish D. montana were originally described as Drosophila ovivororum For the microsatellite analyses, we used the same groupings by Lakovaara & Hackman (1973) and Vieira & Hoikkala as for the mtDNA population structure analysis: the (2001) showed them to differ genetically from the North- sample from Finland was divided into two geographical American D. montana populations. The status of the Japanese groupings (Oulanka and Kemi, separated by about population is less well known. We divided haplotypes 225 km) and the lines from Utah were separated from according to this classification (see Supplementary material). those collected in Colorado (both USA in the mtDNA Pairwise comparisons between different classes indicated analysis) giving six samples in total. Genotype data for 16 that the Finnish samples were significantly different from polymorphic microsatellite loci showed that D. montana all others at the 0.05 level, and that Giant montana differed populations are genetically highly differentiated over the from all other North American populations but not from geographical area studied (mean FST across loci = 0.208; lines from Japan (Table 4). However, only two Giant lines 95% confidence interval from 0.175 to 0.261). Significant were analysed, so this last result should be treated with differentiation was detected for each of the 16 loci (all caution. P < 0.05), suggesting that differentiation is a genome- Studies on song evolution require confronting the wide phenomenon rather than being concentrated at a few predicted pattern of song changes to a phylogeny as well individual loci. as tracing selective pressures affecting song traits at the To investigate the influence of geographical separa- population level. We tested whether divergence between tion on genetic differentiation, we calculated a matrix of strains for song characters was correlated with genetic dis- pairwise FST values between populations (averaged across tances, using partial Mantel tests, or whether they differed loci). Pairwise comparisons showed clear genetic differences between genetically distinct populations, using analysis between samples originating from different continents, of variance. Significant associations are expected if duration while comparisons within continents showed a much lower

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd 1094 P. M. MIROL ET AL.

Discussion Drosophila montana is an unusual species within the virilis group. It differs from other members of the group in several ways, including mating behaviour, chromosome variability, geographical range, specificity of larval substrate and association with human habitats (Throckmorton 1982). Here, we provide a phylogeographic analysis as a basis for understanding phenotypic variation within the species. Fig. 5 Admixture analysis based on six Drosophila montana Mitochondrial DNA and microsatellites indicate the populations. Genetically distinct groups are indicated by distinct presence of at least two distinct populations, one in Eurasia colours. Each individual is represented by a line, which is and the other one representing the expansion of the species partitioned into coloured segments that represent the individual’s estimated membership fraction to the corresponding cluster. to the New World. The mtDNA haplotypes found in North America form a clade that nests within the more diverse set of haplotypes present in Finland supporting this direction degree of differentiation (Table 2). The impact of large- of colonization. Genetic distances between the two major scale geographical separation on genetic differentiation mtDNA clades range from 0.9 to 1.8%, which, assuming among D. montana populations is best illustrated by the a mitochondrial divergence rate of 2% per million years, results obtained from amova. Approximately 67% of the implies a separation between them from 450 000 to 900 000 among-population genetic variation can be attributed years ago, within the Pleistocene. Päällysaho et al. (2005) to differentiation among continents (Table 1). Pairwise obtained congruent divergence times between Finnish

FST values indicate that genetic variation within continents and American populations of D. montana based on silent mainly resulted from the fact that the Vancouver sample substitutions in three X-linked genes, fused, elav and su(s). differs from the Colorado sample, while the Finnish popu- Microsatellite data revealed distinct Finnish and North lations did not differ genetically (Table 2). American populations, like mtDNA data, but they further Bayesian analysis supports the inferences of population suggested genetic differentiation within North America. structure from F statistics, indicating that the six samples Although samples of mtDNA haplotypes from Canada originate from four genetically distinct groups: a Finnish and USA are not phylogenetically distinct, the populations Φ group (Kemi, Oulanka), a Japanese group and two American are significantly differentiated as judged by ST. They also groups. One American group contained the populations show evidence for different historical demographic patterns from Utah and Vancouver and the second American group (Fig. 4). This might reflect partial or complete isolation contained the population from Colorado. This grouping into distinct northern (Beringian) and southern (Rocky was supported with very high probability (P > 0.999) Mountains) refugia (Hewitt 2004) during the last glaciation. while the probability for all possible alternative groupings The possibility of different colonization times could also be was very low (P < 0.001). To further address the role of recent considered. Both the Finnish and Canadian samples sug- gene flow, especially between the Northern American gest very rapid population expansion around 35 000 years populations, we performed a Bayesian admixture analysis. ago, somewhat older than the end of the last glaciation, The analysis showed that the inferred admixture between while the US sample suggests more gradual expansion individuals belonging to the four distinct genetic clusters starting earlier, which is consistent with the more southerly is very low (Fig. 5). Most notably, there were no signs location of the Rocky Mountains refuge. The mtDNA data of admixture between the Colorado and the Vancouver do not support a suggestion of recent gene exchange sample supporting their genetic distinctiveness. Note that between Finnish and North American populations based one individual of the Finnish group showed a high degree on X-linked genes (Päällysaho et al. 2005). However, this of admixture. Based on the current data set, it is difficult to suggestion was based on shared variation in short stretches judge whether this reflects a true admixture event or is of sequence that might represent remnants of ancestral due to the contamination of one line in the laboratory. polymorphisms that have yet to achieve reciprocal mono- Also, this analysis shows the individuals from Utah to be phyly (Päällysaho et al. 2005). closer to the individuals from Vancouver than to those in Acoustic signals used by males during courtship may Colorado, which contradicts the pattern obtained based on experience various types of selection. Some song traits, mtDNA. However, only three individuals were genotyped important for mate recognition, may be under stabilizing for microsatellites in the Utah population, and this could selection and others may be under directional sexual selec- be the cause of the result. A table including the number and tion while remaining traits have no signalling function the frequency of different microsatellite alleles for each locus and may evolve neutrally. Neutral traits should vary geo- and population is presented in the Supplementary material. graphically in a way that mirrors divergence at marker

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd PHYLOGEOGRAPHY OF DROSOPHILA MONTANA 1095 loci, such as mtDNA and microsatellites, while traits under Aknowledgements stabilizing or directional selection are expected to be less variable within and among populations. However, a We are grateful to the members of the ‘Co-evolved Traits’ change in the environment may result in rapid divergence Research Training Network for their valuable input to the work between populations in selected traits. It has also been sug- presented here and to the European Commission for funding the network (HPRN-CT-2002-00266). Special thanks are due to gested that rapid expansion of a population into a new, Oulanka Biological Station and LAPBIAT project, the Rocky vacant habitat might be associated with a reduction in Mountain Biological Laboratory, and Andrew Beckenbach for female discrimination among males and so a weakening of help in arranging fly-collecting trips in different parts of the selection on male mating signals (Kaneshiro 1980). This species’ distribution area, for all the people who have helped in might result in greater genetic variability in signals within collecting the flies and establishing of isofemale lines (especially populations and greater divergence among populations, Dominique Mazzi, Susanna Huttunen and Kirsten Klappert). but the effect may be weakened if the small initial population Laboratory strains of the flies were obtained also from Jorge Vieira, Michael Evgenev and Bowling Green stock centre. Very had low variability. special thanks to A.F.O. for helping with statistics. Our data suggest that D. montana populations in Europe and North America have been isolated for a long period, allowing the opportunity for accumulation of divergence Supplementary material in neutral song traits. They have also expanded in num- The supplementary material is available from bers, in some cases rapidly, which might have resulted in http://www.blackwellpublishing.com/products/journals/ relaxed selection or a change in the pattern of selection on suppmat/MEC/MEC3215/MEC3215sm.htm key song traits. Our initial analysis of song traits did not detect an association between genetic and phenotypic Lines of Drosophila montana used in the study, indicating, when it was available, year of collection and coordinates from which the distance, suggesting that song evolution is not neutral. line originates. Morphological subtypes, name of the line and However, the analysis did detect significant divergence mitochondrial (mt), microsatellite (ms) and song traits data are between the Finnish and North American populations, also indicated primarily in carrier frequency. In the Finnish D. montana population, females are known to show strong mate Allele sizes (size) number of alleles (count) and allele frequencies (frq) preferences based on song, specifically on carrier frequency for 16 microsatellite markers for six Drosophila montana populations (Aspi & Hoikkala 1995; Ritchie et al. 1998). It is particularly interesting that the trait that is known to be under sexual References selection is also the trait that distinguishes this population Adrianov BV, Sorokina SY, Gorelova TV, Mitrofanov VG (2003) from the other regions. Further work will be needed to Mitochondrial DNA polymorphism in natural populations of establish whether this results from relaxation of selection the Drosophila virilis species group. Russian Journal of Genetics, or a change in the direction of selection. However, the fact 39, 630–635. that carrier frequency is low in the Finnish population, Aspi J, Hoikkala A (1995) Male mating success and survival in the where it experiences directional sexual selection favouring field with respect to size and courtship song characters in high frequencies and where the effective population size is Drosophila montana and D. littoralis. Heredity, 70, 400–406. de Brito RA, Manfrin MH, Sene FM (2002) Mitochondrial DNA large, might suggest a relaxation of selective constraints in phylogeography of Brazilian populations of Drosophila buzzatii. North America. Future detailed studies on song variation Genetics and Molecular Biology, 25, 161–171. should concentrate on freshly collected strains because the Butlin RK (2005) Recombination and speciation. Molecular Ecology, songs, especially the pulse characters of the song, may be 14, 2621–2635. liable to change during laboratory maintenance. The Finnish Caletka BC, McAllister BF (2004) A genealogical view of chromo- strains used in the song analysis reported here were, on somal evolution and species delimitation in the Drosophila virilis average, younger than strains from other regions but there species subgroup. Molecular Phylogenetics and Evolution, 33, 664–670. were not sufficient strains to test for a systematic effect Clement M, Posada D, Crandall KA (2000) tcs: a computer program of age. to estimate gene genealogies. Molecular Ecology, 9, 1657–1659. We have presented here a possible scenario for the bio- Corander J, Waldmann P, Sillanpää MJ (2003) Bayesian analysis geographic history of D. montana, which will form the basis of genetic differentiation between populations. Genetics, 163, for the interpretation of evolution of their mating signals 367–374. and responses. Currently available information on male Coyne JA, Orr HA (2004) Speciation. Sinauer Associates, Sunderland, song shows divergence among populations that have a Massachusetts. Cymbron T, Freeman AR, Isabel Malheiro M, Vigne JD, Bradley long history of separation. Further analyses of these traits DG (2005) Microsatellite diversity suggests different histories will help to show how population demography interacts for Mediterranean and Northern European cattle populations. with selection to generate the divergence in mating behaviour Proceedings of the Royal Society of London. Series B, Biological Sciences, that might ultimately cause speciation. 272, 1837–1843.

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd 1096 P. M. MIROL ET AL.

Dieringer D, Schlötterer C (2003) Microsatellite analyser (msa): Ostrega MS (1985) Restriction endonuclease analysis of the relat- a platform independent analysis tool for large microsatellite edness of D. montana and D. virilis lines. Drosophila Inf. Serv, 61, data sets. Molecular Ecology Notes, 3, 167–169. 132–133. Farris JS, Kallersjo M, Kluge AG, Bult C (1995) Testing significance Päällysaho S, Vieira C, Hoikkala A, Vieira J (2005) Evidence for of incongruence. Cladistics, 10, 315–319. introgression in differentiated North-American and Finnish Felsenstein J (1991) Counting phylogenetic invariants in some Drosophila montana populations. Genetica, 123, 285–293. simple cases. Journal of Theoretical Biology, 152, 357–376. Page RDM (1996) TREEVIEW: an application to display phylo- Fu YX, Li WH (1993) Statistical tests of neutrality of mutations. genetic trees on personal computers. Computing Applications in Genetics, 133, 693–709. the Biosciences, 12, 357–358. Hewitt GM (2001) Speciation, hybrid zones and phylogeography Panhuis TM, Butlin R, Zuk M, Tregenza T (2001) Sexual selection or seeing genes in space and time. Molecular Ecology, 10, 537– and speciation. Trends in Ecology & Evolution, 16, 364–371. 549. Pedrosa S, Uzun M, Arranz JJ, Gutierrez-Gil B, San Primitivo F, Hewitt GM (2004) Genetic consequences of climatic oscillations Bayon Y (2005) Evidence of three maternal lineages in Near in the Quaternary. Philosophical Transactions of the Royal Society of Eastern sheep supporting multiple domestication events. London. Series B, Biological Sciences, 359, 183–195. Proceedings of the Royal Society of London. Series B, Biological Sciences, Higgins D, Bleasby A, Fuchs R (1992) clustal v: improved soft- 272, 2211–2217. ware for multiple sequence alignment. Computing Applications Posada D, Crandall KA (1998) modeltest: testing the model of in the Biosciences, 8, 189–191. DNA substitution. Bioinformatics, 14, 817–818. Hilton H, Hey J (1996) DNA sequence variation at the period locus Pybus OG, Rambaut A, Harvey PH (2000) An integrated frame- reveals the history of species and speciation events in the work for the inference of viral population history from recon- Drosophila virilis group. Genetics, 144, 1015–1025. structed genealogies. Genetics, 155, 1429–1437. Hilton H, Hey J (1997) A multilocus view of speciation in the Ritchie MG, Townhill RM, Hoikkala A (1998) Female preference Drosophila virilis species group reveals complex histories and for fly song: playbacks confirm correlational evidence of the taxonomic conflicts. Genetical Research, 70, 185–194. targets of sexual selection. Animal Behavior, 56, 713–717. Hoikkala A (1988) The importance of different courtship stimuli in Ritchie MG, Saarikettu M, Hoikkala A (2005) Variation, but no cov- the mating behaviour of European species of the Drosophila ariance, in female preference functions and male song in natural virilis group. Annales Zoologica Fennici, 25, 257–263. populations of Drosophila montana. Animal Behaviour, 70, 849–854. Hoikkala A, Lumme J (1987) The genetic basis of evolution of the Saarikettu M, Liimatainen JO, Hoikkala A (2005) The role of male male courtship sounds in the Drosophila virilis group. Evolution, courtship song in species recognition in Drosophila montana. 41, 827–845. Behavior Genetics, 35, 257–263. Hoikkala A, Klappert K, Mazzi D (2005) Factors affecting male Saarikettu M, Liimatainen JO, Hoikkala A (2005) Intraspecific song evolution in Drosophila montana. Current Topics in Develop- variation in mating behaviour does not cause sexual isolation mental Biology, 67, 225–250. between Drosophila virilis strains. Animal Behaviour, 70, 417–426. Kaneshiro KY (1980) Sexual isolation, speciation and the direction Sambrook J, Fritsch EF, Maniatis T (1989) Molecular Cloning: a of evolution. Evolution, 34, 437–444. Laboratory Manual, 2nd edn. Cold Spring Harbor Laboratory Kuhner MK, Yamato J, Felsenstein J (1998) Maximum likelihood Press, Cold Spring Harbor, New York. estimation of population growth rates based on coalescent. Schlötterer C (1998) Microsatellites. In: Molecular Genetic Analysis Genetics, 149, 429–434. of Populations: a Practical Approach (ed. Hoelzel AR), pp. 237–261. Lakovaara S, Hackman W (1973) Drosophila ovirorum sp. n., a new Oxford University Press, Oxford. species of the Drosophila virilis group (Diptera, Drosophilidae). Schlötterer C, Zangerl B (1999) The use of imperfect microsatellites Notulae Entomologicae LIII, 167–169. for DNA fingerprinting and population genetics. In: DNA Larson G, Dobney K, Albarella U et al. (2005) Worldwide phylo- Profiling and DNA Fingerprinting (ed. Lubjuhn T), pp. 153–165. geography of wild boar reveals multiple centers of pig domes- Birkhäuser, Basel. tication. Science, 307, 1618–1621. Schneider S, Kueffer JM, Roessli D, Excoffier L (1997) arlequin Liimatainen J, Hoikkala A (1998) Interactions of the males and 1.1. A software for population genetic data analysis. URL: females of three sympatric Drosophila virilis group species, D. http://anthropologie.unige.ch/arlequin. montana, D. littoralis and D. lummei, in intra- and interspecific Servedio MR, Noor MAF (2003) The role of reinforcement in courtships in the wild and in the laboratory. Journal of Insect speciation: theory and data. Annual Review of Ecology and Sys- Behavior, 11, 399–417. tematics, 34, 339–364. Liu H, Beckenbach AT (1992) Evolution of the mitochondrial cyto- Simon C, Frati F, Beckenbach A, Crespi B, Liu H, Flook P (1994) chrome oxidase II gene among ten orders of insects. Molecular Evolution, weighting and phylogenetic utility of mitochondrial Phylogenetics and Evolution, 41, 31–52. gene sequences and a compilation of conserved Polymerase Chain Miller SA, Dykes DD, Polesky HF (1988) A simple salting out Reaction primers. Annual Entomology Society of America, 87, 651–701. procedure for extracting DNA from human nucleated cells. Slatkin M, Hudson RR (1991) Pairwise comparisons of mito- Nucleic Acids Research, 16, 1215. chondrial DNA sequences in stable and exponentially growing Nurminsky DI, Moriyama EN, Lozovskaya ER, Hartl DL (1996) populations. Genetics, 129, 555–562. Molecular phylogeny and genome evolution in the Drosophila Sokal RR, Rohlf FJ (1995) Biometry. W.H. Freeman, New York. virilis species group: duplications of the alcohol dehydrogenase Spicer GS (1991) Molecular evolution and phylogeny of the gene. Molecular Biology and Evolution, 13, 132–149. Drosophila virilis species group as inferred by two-dimensional Orsini L, Huttunen S, Schlötterer C (2004) A multilocus micro- electrophoresis. Journal of Molecular Evolution, 33, 379–394. satellite phylogeny of the Drosophila virilis group. Heredity, 93, Spicer GS (1992) Reevaluation of the phylogeny of the Drosophila 161–165. virilis. Annual Entomology Society of America, 85, 11–25.

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd PHYLOGEOGRAPHY OF DROSOPHILA MONTANA 1097

Spicer GS, Bell CD (2002) Molecular phylogeny of the Drosophila Vieira J, Hoikkala A (2001) Variability levels, population size and virilis species group (Diptera: Drosophilidae) inferred from structure of American and European Drosophila montana popu- mitochondrial 12S and 16S ribosomal RNA genes. Annual lations. Heredity, 86, 506–511. Entomology Society of America, 95, 156–161. Waterson GA (1975) On the number of segregating sites in genetical Tajima F (1989) Statistical method for testing the neutral mutation models without recombination. Theoretical Population Biology, 7, hypothesis by DNA polymorphism. Genetics, 123, 585–595. 256–275. Throckmorton LH (1982) The Genetics and Biology of Drosophila, Weir BS, Cockerham CC (1984) Estimating F-statistics for the ana- Vol. 3b. Academic Press, New York. lysis of population structure. Evolution, 38, 1358–1370. Tregenza T, Pritchard VL, Butlin RK (2000) Patterns of trait diver- Wright S (1978) Evolution and Genetics of Populations, Vol. 4, gence between populations of the meadow grasshopper, Variability Within and Among Natural Populations. University of Chorthippus parallelus. Evolution, 54, 574–585. Chicago Press, Chicago.

© 2006 The Authors Journal compilation © 2006 Blackwell Publishing Ltd

IV

THE EXTENT OF VARIATION IN MALE SONG, WING AND GENITAL CHARACTERS AMONG ALLOPATRIC DROSOPHILA MONTANA POPULATIONS

by

Jarkko Routtu, Dominique Mazzi, Kim van der Linde, Patricia M. Mirol, Roger K. Butlin and Anneli Hoikkala 2007

Journal of Evolutionary Biology 20: 1591–1601

Reprinted with kind permission of Blackwell Publishing

doi:10.1111/j.1420-9101.2007.01323.x

The extent of variation in male song, wing and genital characters among allopatric Drosophila montana populations

J. ROUTTU,* D. MAZZI,*1 K. VAN DER LINDE, P. MIROL,à2 R. K. BUTLINà & A. HOIKKALA* *Department of Biological and Environmental Sciences, University of Jyva¨skyla¨, Jyva¨skyla¨, Finland Department of Biological Science, Florida State University, Tallahassee, FL, USA àAnimal and Plant Sciences, University of Sheffield, Sheffield, UK

Keywords: Abstract genital morphology; Drosophila montana, a species of the Drosophila virilis group, has distributed male courtship song; around the northern hemisphere. Phylogeographic analyses of two North multivariate statistics; American and one Eurasian population of this species offer a good background natural selection; for the studies on the extent of variation in phenotypic traits between sexual selection; populations as well as for tracing the selection pressures likely to play a role in speciation; character divergence. In the present paper, we studied variation in the male wing morphology. courtship song, wing and genital characters among flies from Colorado (USA), Vancouver (Canada) and Oulanka (Finland) populations. The phenotypic divergence among populations did not coincide with the extent of their genetic divergence, suggesting that the characters are not evolving neutrally. Divergence in phenotypic traits was especially high between the Colorado and Vancouver populations, which are closer to each other in terms of their mtDNA genotypes than they are to the Oulanka population. The males of the Colorado population showed high divergence especially in song traits and the males of the Vancouver population in wing characters. Among the male song traits, two characters known to be under sexual selection and a trait important in species recognition differed clearly between populations, implying a history of directional and/or diversifying rather than balancing selection. The population divergence in wing characters is likely to have been enhanced by natural selection associated with environmental factors, whereas the male genitalia traits may have been influenced by sexual selection and/or sexual conflict.

Introduction conditions (Fisher, 1930; Wright, 1931; Dobzhansky, 1951). Phylogenetic or phylogeographic analyses based Populations spreading into new environments may on neutral molecular markers provide a good back- diverge from the ancestral population both genetically ground for population level studies on phenotypic and phenotypically. Phenotypic divergence may reflect divergence but, as Grandcolas & D’Haese (2003) have genetic divergence or it can be enhanced by natural emphasized, the selective value of a character can best selection or by stochastic effects of mutation and be considered when examining patterns of phenotypic genetic drift. The role of selection is of special import- variation among wild populations. Phenotypic studies ance in populations adapting to novel environmental should, if possible, be targeted on several behavioural and morphological traits at the same time to distin- Correspondence: J. Routtu, Department of Biological and Environmental guish neutrally evolving traits from those whose Sciences, 40014 University of Jyva¨skyla¨, Jyva¨skyla¨, Finland. evolution has been affected by different kinds of Tel.: +358 142 604239; fax: +358 142 602321; selective pressures. e-mail: [email protected].fi During the courtship rituals, Drosophila males emit 1Present address: Institute of Plant Sciences, Applied Entomology, ETH (Swiss Federal Institute of Technology), CH-8092 Zurich, Switzerland. various kinds of species-specific stimuli, e.g. male court- 2Present address: Museo Argentino de Ciencias Naturales, C1405DJR, ship song. The songs may play an important role in Buenos Aires, Argentina. sexual selection and/or in species recognition, and they

ª 2007 THE AUTHORS 20 (2007) 1591–1601 JOURNAL COMPILATION ª 2007 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY 1591 1592 J. ROUTTU ET AL.

may simultaneously be affected by directional, diversify- of the Bering Sea between Eurasia and North America ing and balancing selection. However, signals important isolates populations on the two continents. Nevertheless, in species recognition may not vary too much if they are a study of nuclear sequences of three X chromosomal to retain species specificity (Lambert & Henderson, genes showed evidence for recurrent gene flow among 1986). Lande (1982) has shown that the evolution of North American populations and possibly also from directional female mating preferences for male secondary North America to Finland (Pa¨a¨llysaho et al., 2005). sexual characters can greatly amplify large-scale geo- A recent population genetic study (Mirol et al., 2007) graphic variation in male characters. This coevolutionary showed that the North American populations from process can be enhanced by variation in the strength of Vancouver and Colorado do not have distinct mitoch- direct or indirect selection on female preferences through ondrial haplotypes, whereas the haplotypes of the Fin- the species distribution area (see Houde, 1993). Diversi- nish Oulanka population differ from those of the North fying selection (where the direction of selection varies American populations and are more diverse. In the same between populations) may speed up the evolution of study, the microsatellites differentiated all of the study species-specific songs and might increase the effective- populations. Mirol et al. (2007) estimated population ness of prezygotic sexual isolation between sympatric divergence time for the Finnish and North American species (Etges et al., 2006). populations to be 0.45–0.90 Myr, with rapid population Among morphological traits, most variation (> 90%) expansion of Oulanka and Vancouver populations about in wing shape of Drosophila flies is correlated with the 35 000 years ago and a more gradual expansion of the phylogenetic history of the species (K. van der Linde & D. Colorado population starting earlier. These estimates are Houle, unpublished data), although shape differences in concordance with the expected genetic effects of likely between closely related species can vary considerably range changes during the Pleistocene glaciations (Hewitt, (Houle et al., 2003; K. van der Linde & D. Houle, 2004). unpublished data). Variation in wing shape can lead to The objective of the present study was to analyse the functionally identical outcomes with various internal phenotypic divergence of D. montana populations in male structural rearrangements, which suggests a large influ- courtship song and in wing and genital size and shape in ence of genetic drift and, therefore, a correlation with two North American and one Finnish population. The phylogenetic history. For example, the convergence of data on the genetic divergence of these populations clinal variation in wing size of Drosophila subobscura on (Mirol et al., 2007) is used as a reference to compare the different continents has been achieved through analo- levels of genotypic and phenotypic divergence between gous, not homologous, changes in the relative lengths of populations. We hypothesize that adaptation of D. mon- different parts of the wing (Huey et al., 2000). Addition- tana populations to different physical and biotic condi- ally, wing traits have been found to evolve quite rapidly tions (including different Drosophila communities) and in response to geographic clines, e.g. in D. subobscura the restricted gene flow between populations has led to (Gilchrist et al., 2000; Huey et al., 2000), and they the divergence of populations at several behavioural and respond well to artificial selection (Houle et al., 2003; morphological characters. The main issues we address Kennington et al., 2003). These rapid changes are not here are the extent of the phenotypic divergence of surprising because of the abundant genetic variation populations in the male song, wing and genital traits related to wing shape that is available (Mezey & Houle, when compared with their genetic divergence, and 2005; Weber et al., 2005). The flies use their wings also whether the evolution of behavioural and morphological for producing courtship songs, and so the wing can characters has occurred in concert. Moreover, we potentially be influenced by sexual selection. To date, attempted to pinpoint the selective pressures likely to little is known of the influence of wing morphology on be responsible for enhancing phenotypic divergence in male courtship song traits. the characters studied. The size and shape of male genitalia are rapidly evolving species-specific characters and they are often Materials and methods used for species identification, e.g. in Drosophila species (Grimaldi, 1990). Genitalia have been suggested to Isofemale strains evolve via lock–key mechanics (Dufour, 1844), pleiotro- py (Mayr, 1963) or cryptic female choice (Eberhard, Drosophila montana flies were collected from three loca- 1985). The striking morphological diversity of genitalia in tions: Oulanka (Finland; 6627¢N, 2900¢E, near sea the species of the Drosophila virilis group (Kulikov et al., level), Colorado (USA; 3852¢N, 10659¢W, 2700 m 2004) may, also, have arisen through sperm competition altitude) and Vancouver (Canada; 4915¢N, 12306¢W, (Parker, 1970) or sexual conflict (Hosken & Stockley, near sea level). Wild-caught females were transferred 2004; Arnqvist & Rowe, 2005). individually into malt-food vials to lay eggs. They were Our study species, D. montana, has a wide circumpolar moved into new vials after a few days to control for larval distribution in temperate forests around the northern density and food availability. The F1 generation flies hemisphere (Throckmorton, 1982). The physical barrier were sexed within 2 days of their emergence under light

ª 2007 THE AUTHORS 20 (2007) 1591–1601 JOURNAL COMPILATION ª 2007 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY Male song, wing and genital characters 1593

CO2 anaesthesia and the males and the females were kept spectrum of the studied pulse trains. For statistical in separate vials for 3 weeks, i.e. until they were sexually analyses, we calculated the mean values of the song mature. We established seven isofemale lines for the traits over three pulse trains for each male. Data for the Oulanka, six isofemale lines for the Colorado and male song frequencies have been used also in the study 20 isofemale lines for the Vancouver population in by K. Klappert et al. (unpublished data). 2003. Phenotypic characters (male courtship song, male wing size and shape, and male genitalia size and shape) Analysis of wing morphology were measured for three F1 males per isofemale line to allow the estimation of variation in these characters The original procedure for imaging the wings and for within and between populations. The same study lines estimating the vein locations and the corresponding were used for the mtDNA analysis and most of them also landmarks has been described in detail by Houle et al. for the microsatellite analysis in the phylogeographic (2003). Recently, the extraction procedure of the land- analysis of D. montana populations (Mirol et al., 2007). mark data has been modified, such that we can localize allometric variation more precisely (K. van der Linde & D. Houle, unpublished data). The wings of the flies, Recording and analyses of the male courtship songs stored in 70% ethanol after the song recording, were cut Drosophila montana males produce courtship song by wing with a sharp scalpel and attached to a microscope slide vibration. To record this song we transferred individual with transparent tape. A randomly chosen left or right males with a virgin, sexually mature female of the same wing of an individual was used for analysis. After a digital strain into a Petri dish with a moistened filter paper on image of the wing was obtained, the operator marked the bottom and a nylon net roof. The courting flies two start points at the basal end of the wing. The position walked upside down on the roof of the chamber allowing of the veins was described as clamped quadratic B-splines song to be recorded by holding the microphone (AKG C and extracted from the images using the splining proce- 1000 S) directly above the roof of the chamber. Male dure FindWing (Lu & Houle, 1995–1997). This pro- courtship songs were recorded on a Marantz Professional gramme is embedded in a wrapper programme Wings (model no. 74PDM 502/02B) tape recorder at the (van der Linde & Houle, 2004–2006), which automates temperature of 20 ± 1 C and analysed with Signal 4.0 its input and output. Wings also includes modules for (Engineering Design, Belmont, MA, USA) sound analysis outlier detection (e.g. faulty splined wings) using Mini- system. The analysed song traits were the number of mum Volume Ellipsoids (MVE) (Rousseeuw & Leroy, pulses in a pulse train (PN), the length of a pulse train 1987; Rousseeuw & van Zomeren, 1990; K. van der (PTL), the length of a sound pulse (PL), the interpulse Linde & D. Houle, unpublished data) and subsequent interval (IPI, the length of the time from the beginning of correction of the control points of those incorrectly one pulse to the beginning of the next one), the number positioned splines. of cycles in a sound pulse (CN) and the carrier frequency The wing landmark data were aligned using the of the song (FRE). The first five song traits were Generalized Procrustes Analysis (GPA). GPA seeks to measured manually in the oscillograms, PN and PTL for remove nonshape variation by centring, scaling and the whole pulse trains and PL, IPI and CN for the fourth rotating the landmark data to minimize the least-squares pulse of each train (see Fig. 1). The carrier frequency of deviations among them. Centroid size was retained as the the song (FRE) was measured from the frequency scaling variable so that size-dependent changes in shape

0.08

0.06

0.04

0.02

0.00

Fig. 1. Oscillogram of Drosophila montana –0.02 courtship song. The traits measured from the Amplitude (Volts) oscillogram are: the number of pulses in a –0.04 Sound cycle pulse train (PN), the length of a pulse train –0.06 Pulse length (PL) (PTL), the length of a sound pulse (PL), the Interpulse interval (IPI) interpulse interval (IPI) and the number of –0.08 cycles in a sound pulse (CN). Song frequency Pulsetrain length (PTL) (FRE) was measured from the Fourier spec- 0 50 100 150 200 250 300 350 400 450 tra of the pulse trains. Time (msec)

ª 2007 THE AUTHORS 20 (2007) 1591–1601 JOURNAL COMPILATION ª 2007 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY 1594 J. ROUTTU ET AL.

can be explored. Three degrees of freedom are lost in the (Fig. 3) was measured twice for 10 individuals from each estimation of the nuisance variables centring and rota- of the three populations giving a repeatability of 98.6% tion, and another degree of freedom is used to estimate averaged over all repeats. Pictures used in the PCA were centroid size. Two types of data were extracted from the edited to black and white to enable automated measure- wings: 15 pseudo-landmarks describing the outlines of ments with program SHAPE 1.2. Repeatability for the the wing and 12 landmarks describing the junctions PCA was measured for similar set of individuals as for the between wing veins and between the veins and the hook length and it varied from 100% to 94%, the last PCs outline of the wing. The outline data contain consider- being the least accurate ones. ably less allometric variation, and as such, fulfil better the assumption of isotropy that is required for the GPA Statistical analyses (Dryden & Mardia, 1998). The two data sets were aligned simultaneously (van der Linde, 2005–2006). The outline All the statistical analyses were performed with SPSS data were used to determine the nuisance parameters 12.0.1 for Windows (SPSS Inc., Chicago, Illinois, USA), and the nuisance parameters estimated for the outlines except for the PCA on genitalia shape described above. were used to determine the position of the landmarks. The data for the male courtship songs, wings and The outline shape data have 26 degrees of freedom genitalia consisted of the measurements for three males (15 pseudo-landmarks, each with two dimensions, minus of each isofemale line. Normal distribution and homo- the four degrees of freedom lost as described above), geneity of variance were tested with Kolmogorov–Smir- whereas the landmark data have 24 degrees of freedom nov and Levene’s tests respectively. Log10 transformation (12 landmarks times two dimensions, no loss of degrees was performed when necessary. of freedom as the nuisance parameters were estimated Principal component analysis was carried out to reduce for the outline data). The repeatability of wing landmarks the number of dependent variables to be entered in with this measurement technique is as high as 93% subsequent analyses for male genitalia. Only the traits (Houle et al., 2003). describing the shape were included in the PCA; genitalia size (area) and genital hook length were used in the subsequent analysis without modification. The PCA was Genital morphology based on variance–covariance matrices to reduce the The genitalia were removed from the male bodies, which effects of elliptical harmonics describing minor aspects of had been preserved in 70% ethanol after the song variation that are not likely to have biological import- recordings had been completed. They were transferred ance. The PCs that explained more than 1% of the shape singly into Eppendorf tubes with 0.1 M NaOH and kept at variation were included in the variance–covariance- 95 C for 8 min. Thereafter, NaOH was replaced with based PCA. water. The distiphallus, the distal part of the aedeagus, Variation between allopatric D. montana populations in was removed from the rest of the genital apparatus in male song, wing and genitalia traits was studied using the water with fine forceps and placed on its side on a glass F1 progenies of the females collected in Oulanka, slide. After removing the excess water, the distiphallus Vancouver and Colorado populations. The structure of was covered with EUKITT (Kindler, Freiburg, Germany). the data (three males per isofemale strain) allowed us to Slides were photographed with a SPOT Insight colour analyse the magnitude of variance components within digital camera installed on a Zeiss Axioscop 40 micro- and between populations in the measured behavioural scope and stored as bitmap files. Files were analysed with and morphological traits. The proportions of variation at the SHAPE 1.2 (Iwata & Ukai, 2002) programme, which different hierarchical levels were calculated with nested is based on principal component analysis (PCA) per- ANOVAs for unequal sample sizes (Sokal & Rohlf, 1997). formed on elliptical Fourier descriptors (EFD) of an The F-values and their significance for different hier- enclosed contour (Kuhl & Giardina, 1982; see Fig. 3). archical levels (between and within populations and The PCA was used on EFDs to describe morphological within strains) were tested for each trait group (courtship variation in the distiphalli, which, owing to their irregu- song, wings and genitalia), using a sequential Bonferroni lar shape, lack reliable landmarks. The resulting normal- (Holm, 1979; Rice, 1989) corrected significance value to ised Principal Component (PC) scores can be used as avoid type I error. Percentages of variation at different measured trait values that include only the allometric levels were calculated on the basis of the mean squares variation. The size component (area) was measured (MS) from the nested ANOVA. Significantly different traits separately by the programme. The distiphallus will be between populations in nested ANOVAs were used in a referred to as the ‘genitalia’ for simplicity in the rest of discriminant analysis (DA) to test for the influence of the paper. Genital hook length was included in the overall interactions of the combined dependent variables analysis as an additional measure because it has fre- and variance levels between populations. quently been used in species identification (Grimaldi, The extent of overall phenotypic divergence between 1990). The hook length from the endpoint of the hook to populations was studied using DA. The DA was carried the base of the hook at the head of the distiphallus out for the mean values calculated over the three males

ª 2007 THE AUTHORS 20 (2007) 1591–1601 JOURNAL COMPILATION ª 2007 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY Male song, wing and genital characters 1595

from each isofemale strain to avoid pseudoreplication. Table 1 The discriminant analysis structure matrices of the traits, The analysis defines axes that maximally separate the highest correlation of a trait with a discriminant axis is marked with populations and yields correlations of the trait values to an asterisk (*). The abbreviations are: CN (the number of cycles in a the separating axes. The reliability of the classification sound pulse), PL (the length of a sound pulse), IPI (an interpulse interval), PN (the number of pulses in a pulse train), LM (a obtained was confirmed with a cross-validation test. k landmark), OL (an outline (pseudo)landmark), and PC (a principal Wilks’ was used, in connection with the DA, to test the component). significance of the multivariate patterns and to determine the proportions of variance that the multivariate method Songs Wings Genitalia explained. Discriminant MtDNA data for COI and II genes from Mirol et al. axis Discriminant axis Discriminant axis (2007) for the populations used in this study (seven strains from Oulanka, 20 strains from Vancouver and six Trait 12Trait 12Trait 12 strains from Colorado) were plotted using multi-dimen- CN 0.844* 0.395 SIZE 0.295* 0.107 PC4 0.564* )0.008 sional scaling (MDS) analysis with the ’metric’ option PL 0.769* 0.363 LM Y1 )0.253* )0.167 PC7 )0.181 0.912* (performed in GenStat 8.0; VSN International Ltd, IPI 0.249 0.717* LM X9 )0.245* )0.017 PC3 0.523 0.527* Oxford, UK). We also performed AMOVA for the same PN )0.437 0.474* LM X10 )0.239* )0.036 data set. We did not have microsatellite data for the same LM X7 )0.230* 0.124 set of strains. Consequently, microsatellite and pheno- OL Y14 0.178 0.620* typic data for the Oulanka, Vancouver and Colorado OL Y13 0.086 0.488* populations have been compared only in the discussion. LM Y3 )0.037 )0.430* The courtship songs are produced by male wing OL Y9 )0.095 )0.401* ) vibration. Possible covariance between the wing and LM Y8 0.308 0.395* OL X10 )0.128 )0.395* song traits was studied using ANCOVA on the mean values OL X11 )0.139 )0.372* of these traits in isofemale strains. OL Y8 )0.095 )0.366* OL X2 0.134 0.362* Results OL X12 )0.135 )0.359* OL X13 )0.132 )0.348* OL Y15 0.119 0.341* Variation in male song, wing and genitalia traits OL X14 )0.125 )0.310* between populations OL X5 0.133 0.301* Male courtship songs OL X3 0.151 0.299* OL X4 0.157 0.292* Nested ANOVAs for the male song traits showed that LM X4 )0.243 0.286* the CN, PL and IPI, describing the quality of sound LM Y7 )0.191 )0.286* pulses and the PN, giving the number of pulses in a LM X1 )0.181 0.211* pulse train, vary significantly between populations after sequential Bonferroni correction (see Table S1 in Sup- plementary material). CN and PL had the highest between population variance components (66.3% and (Fig. 4). Cross-validated classification of the strains into 58.1% respectively) and IPI the lowest one (24.1%). their correct population of origin was high for the Song frequency did not vary significantly among strains from the Colorado and Vancouver populations populations (F2,30 ¼ 4.74, P ¼ 0.016) after sequential (83.3% and 95% respectively), whereas 71.4% of Bonferroni correction (pcrit ¼ 0.0125) and so it was not the strains from the Oulanka population were included in the subsequent DA. (mis)classified to the Vancouver population. In DA, the first and the second discriminant axes for the four song traits varying between populations Male wing size and shape accounted for 84.4% (eigenvalue 3.21) and 15.6% The traits describing variation in male wing size and (eigenvalue 0.59) of the total variation between pop- shape showed, on average, less variation between pop- ulations respectively. Wilks’ k of 0.15 for the two ulations than the male song traits. Nested ANOVAs discriminant axes confirmed significant divergence in resulted in 24 traits, from the original 55 traits, varying song among populations (P < 0.001). Divergence between populations after sequential Bonferroni correc- remained significant after the removal of the influence tion (pcrit ¼ 0.002). The among population variance of the first discriminant axis (Wilks’ k ¼ 0.63; P ¼ levels for these traits were between 59.9% and 16.5%. 0.004). The first discriminant axis was mainly affected Variation within the strains was quite high, whereas the by CN and PL and the second one by IPI and PN variation among the strains of the same population was (Table 1). The first axis separated the Colorado popu- generally low. lation from the Oulanka and Vancouver populations, The first and the second discriminant axes for the wing whereas the second axis had little explanatory power measures explained 83.7% (eigenvalue 17.64) and

ª 2007 THE AUTHORS 20 (2007) 1591–1601 JOURNAL COMPILATION ª 2007 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY 1596 J. ROUTTU ET AL.

Fig. 2 Wing landmarks. Each landmark has been portioned into an x and y variable. Landmarks (LM) (large numbers) and out- line (pseudo)landmarks (OL) (small num- bers) are shown in the picture. Only the first and the last number, for the outline (pseudo)landmarks, are illustrated.

k PC3 16.3% (eigenvalue 3.44) of total variation. Wilks’ had a combined effect of 0.012 for the two axes (P < 0.001) and of 0.23 (P ¼ 0.045) after the removal of the first discriminant axis. The first discriminant axis was signi- ficantly affected by the centroid size of the wing and the landmarks (LM) 1, 7, 9 and 10, whereas the second discriminant axis was affected by several traits (Table 1, Fig. 2). The first discriminant axis separated the Vancou- ver population from the Oulanka and Colorado popula- PC4 tions, whereas the second axis separated all populations (Fig. 4). Cross-validated classification of the strains into the correct population of origin was 57.1% for the Oulanka strains, 83.3% for Colorado strains and 95% for Vancouver strains. Oulanka strains were misclassified mainly to Colorado (28.6%) and vice versa (16.7%).

PC7 Male genitalia size and shape The analysis of male genitalia shape gave a complicated pattern of correlations between EFDs, resulting in 14 PCs. Altogether 30 harmonics were used for the extraction of EFDs, the normalization method being based on the first harmonic. Variance–covariance matrices of the resulting 117 EFDs (X and Y dimensions, sine and cosine compo- nents for each harmonic) were analysed with PCA embed- ded in the SHAPE 1.2 software. The PCA with 14 axes 100 μm explained 93.3% of the total variation. The genital hook length and the genital size (area) were treated separately. The between populations variance level for the male Fig. 3 Genitalia outlines. Shapes corresponding to the mean and genitalia traits was not as high as for the song and wing two times the standard deviation in both directions are shown for the three axes that contribute to significant variation among traits. Nested ANOVAs resulted to only three traits, out of populations. A long distance between the lines indicates the highest the original 16 traits, varying between populations after variance. Shapes are normalized, but the approximate scale is sequential Bonferroni correction (pcrit ¼ 0.025). PC7 shown. showed the highest among population variance (18%),

ª 2007 THE AUTHORS 20 (2007) 1591–1601 JOURNAL COMPILATION ª 2007 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY Male song, wing and genital characters 1597

Courtship Songs Wings 4 4 POP Oulanka Colorado Vancouver 2 2

0 0 Axis 2 Axis 2

–2 –2

–4 –4

–6–4–20246 –6–4–20246 Axis 1 Axis 1

Genitalia mtDNA 4 2

2 1

Fig. 4 Scatterplots of the discriminant axes and multi-dimensional scaling plot for mit- 0 Axis 2 0 ochondrial DNA. Each point is the mean Axis 2 value for an isofemale strain. The Colorado population (six strains) is denoted by circles, –2 the Oulanka population (seven strains) by –1 squares and the Vancouver population (20 strains) by triangles. Note that fife Oulanka –4 strains have coincident positions in the –6–4–20246 –3 –2 –1 0 1 mtDNA plot around position ()1, )1). Axis 1 Axis1

followed by PC3 (13.3%) and PC4 (13.1%). PC3 des- Genetic vs. phenotypic divergence of the strains from cribes variation in overall thickness of distiphallus, PC4 different populations mainly in the angle and extension of the genital hook Molecular data concerning the strains used in this study and the basal curvature of the distiphallus and PC7 in a were obtained from Mirol et al. (2007). Mitochondrial specific curvature just above the hook (Fig. 3.). The hook (mt)DNA sequences (COI and II, 1358 bp) from seven length (F2,30 ¼ 3.61, P ¼ 0.039) did not vary signifi- Oulanka, six Colorado and 20 Vancouver strains were cantly among populations after sequential Bonferroni plotted in a MDS figure, based on F84 distances. The correction. North American strains form two overlapping clusters The first and the second discriminant axis for the male with the mtDNA haplotypes mixed between the two genitalia traits accounted for 83.3% (eigenvalue 2.13) populations. The Oulanka strains are clearly differenti- and 16.7% (eigenvalue 0.43) of among-population ated from the North American strains, this population variance respectively. Wilks’ k for the two discriminant showing high within population variation in mtDNA axes had a combined effect of 0.22 (P < 0.001) and an haplotypes. AMOVA results for the mtDNA were Fst ¼ effect of 0.70 (P ¼ 0.006) after the removal of the 0.187 (P < 0.0001), 18.7% variation among populations. influence of the first discriminant axis. The first discri- minant axis was mainly affected by PC4 and the second Covariation between wing and song traits one by PC7 and PC3 (Table 1). The first axis separated all three populations from each other and the second one Covariation between the sexually selected pulse charac- separated Colorado from the rest of the populations ters of male song (PL, CN and FRE; Ritchie et al., 2001) (Fig. 4). The cross-validated classification into correct and the wing traits was studied with ANCOVA to find out population was 85.7% for the Oulanka strains, 50% for whether variation in wing morphology could have an the Colorado strains and 80% for the Vancouver strains. effect on the quality of male song. Significant covariation A total of 33.3% of the strains from Colorado was was detected in ANCOVA for the wing landmark charac- misclassified to the Oulanka population. ter LM X4 as the dependent variable, the populations as

ª 2007 THE AUTHORS 20 (2007) 1591–1601 JOURNAL COMPILATION ª 2007 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY 1598 J. ROUTTU ET AL.

the independent variable and the song characters PL but partly different set of strains; data in Mirol et al., (F1,27 ¼ 10.92, P ¼ 0.003), CN (F1,27 ¼ 9.05, P ¼ 0.006) 2007). Even though the data on the phenotypic and and FRE (F1,27 ¼ 11.32, P ¼ 0.002) as covariates. The genetic divergence of populations are not strictly com- landmark character LM X4 pinpoints the endpoint of the parable (analysis of variance comprising different levels second long vein on the outline of a wing (Houle et al., and partly different strains for microsatellites), the 2003), i.e. the shorter the vein, the longer the PL and the magnitude of the population divergence in phenotypic higher the CN and the FRE. characters is clearly higher than their genetic divergence. The most important selection pressures liable to have Discussion an effect on population divergence in male song charac- ters are directional sexual selection exercised by females We analysed the extent of population divergence in on male traits (e.g. Heisler et al., 1987), diversifying phenotypic characters and compared it with the genetic selection towards the increase of divergence among data to trace the effects of selection in the phenotypic sympatric species (Etges et al., 2006) and balancing characters. We discovered strong divergence in the song selection for maintaining species specificity (Lambert & and wing characters and milder divergence in the Henderson, 1986). The song traits may be affected genital characters. Population divergence in these traits simultaneously by different kinds of selective pressures did not coincide with the extent of their genetic and they may also change through alterations in other divergence, suggesting that the characters are not genetically correlated song traits. In D. montana, the song evolving neutrally. is an essential element of male courtship rituals and a The three D. montana populations, two from North requirement for mating (Liimatainen et al., 1992). The America and one from Finland, showed significant females of this species (Finnish population) have been geographic variation both in behavioural and morpholo- found to exercise choice on male song traits, both in the gical characters. As all the flies used in this analysis were wild (Aspi & Hoikkala, 1995) and in playback experi- reared under standardized laboratory conditions, this ments with synthetic songs (Ritchie et al., 2001). In these variation must have a genetic basis even though the studies, the females have shown a preference for effects of maternal and environmental factors cannot be songs with a short PL and high CN and FRE, opening fully excluded. The divergence between populations in a possibility for Fisherian runaway or good-genes the studied characters did not coincide with the degree of selection (Hoikkala et al., 1998). genetic divergence of the same populations (Mirol et al., When studying variation in the male courtship songs 2007), suggesting that the phenotypic characters are not of the laboratory strains of D. montana from a wide evolving neutrally. If the phenotypic traits had evolved geographic area, Mirol et al. (2007) found the largest in line with the neutral genetic divergence between divergence to be accounted for by the pulse characters of populations, the Finnish Oulanka population should the song, mainly the song frequency (FRE). In our study, have formed a separate cluster from the North American we detected high variation between D. montana popula- Colorado and Vancouver populations. At the phenotypic tions in the pulse characters CN and PL, variation in FRE level, the divergence between the North American becoming nonsignificant after Bonferroni correction. The populations in song and wing data was, however, as fact that the progenies of wild-caught females showed no high as or even higher than their divergence from the clear divergence in FRE in our study was mainly because Oulanka population. of high within-strain variation. Our study strains had not Low concordance between the genetic and phenotypic been maintained for multiple generations in laboratory divergence of populations has earlier been found in conditions, which can decrease within strain and several studies comparing the genotypic and phenotypic increase between strain variation in FRE (S. Huttunen, divergence between conspecific populations (see McKay J. Aspi, A. Hoikkala, J. Routtu, C. Schlo¨tterer, unpub- & Latta, 2002) or subspecies (e.g. Chan & Arcese, 2003), lished). The song frequency is quite sensitive to changing the divergence at the phenotypic level usually exceeding environmental factors (Hoikkala & Suvanto, 1999) and it the divergence at the genetic level. We were not able to also has a low heritability (Aspi & Hoikkala, 1993; use the FST and QST statistics (e.g. Merila¨ & Crnokrak, Suvanto et al., 1998), and thus one would expect any 2001; McKay & Latta, 2002) for our study populations, change under selection in this character to be slow. mainly because of the low number of populations. In the In addition to PL and CN, also IPI (the interpulse present study, variation between the Colorado, Vancou- interval) varied significantly among populations. This ver and Oulanka populations in phenotypic traits was trait has been found to play an important role in species 66.3% for the most diverged song trait and 59.9% and recognition in D. montana (Saarikettu et al., 2005) as well 18% for the most diverged wing and genitalia traits as in several other Drosophila species (e.g. Cowling & respectively. In comparison, the mtDNA divergence level Burnet, 1981). The songs of all species of the montana between the three populations for the same set of strains phylad of the D. virilis group have a species-specific as used in the present study was about 18.7% and the IPI (Hoikkala & Lumme, 1987). Variation between microsatellite divergence level 8.78% (same populations, D. montana populations in male courtship song could

ª 2007 THE AUTHORS 20 (2007) 1591–1601 JOURNAL COMPILATION ª 2007 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY Male song, wing and genital characters 1599

have been enhanced by character displacement, if the pressure affecting these traits could be caused by cryptic flies of different populations interact with different female choice or by variation in the ability of males to species in the wild. In Finland, interspecific courtships remove a previous male’s sperm with their genitalia. are quite common between sympatric species (Liimatai- Also sexual conflict over the duration of copulation nen & Hoikkala, 1998). In Colorado, D. montana occurs (K. Klappert & D. Mazzi, unpublished data) may sympatrically with Drosophila borealis and Drosophila influence the curvature of the genitalia and the genital flavomontana (A. Hoikkala & D. Mazzi, personal observa- hook angle to increase the males’ ability to prolong tion) and in Oulanka with Drosophila ezoana, Drosophila copulation (Arnqvist & Rowe, 2005). littoralis and Drosophila lummei (Aspi et al., 1993), whereas Adaptation of D. montana populations to different in Vancouver it is probably the only representative of the environmental conditions together with restricted gene D. virilis group (K. Klappert & L. Orsini, unpublished flow between populations has led to the divergence of observation). It is not possible to trace the historical populations for several behavioural and morphological patterns of sympatry between different species, but it is characters, as predicted. The phenotypic divergence of intriguing that a species-specific trait shows such high allopatric populations in the male courtship songs, wings variation between conspecific populations. On the other and genitalia has clearly been enhanced by selection. It hand, it is important for the species-recognition signals to remains to be seen whether this divergence results in any differ from those of other sympatric species to effectively reproductive isolation between populations. prevent hybridization. The wing size and shape characters separated the two Acknowledgments North-American D. montana populations from each other, the differentiation between the Colorado and Oulanka We are grateful to the European Commission for funding populations being of lower level. Drosophila species are the Research Training Network ‘Co-evolved Traits’ known to evolve latitudinal wing morphology clines (HPRN-CT-2002-00266). JR was supported also by Marie when introduced into novel environments (Gilchrist Curie scholarship, AH by the Finnish Centre of Excel- et al., 2000; Santos et al., 2004), reflecting increasing lence in Evolutionary Research and KvdL by US National body size in colder climate (Huey et al., 2000). Similarly, Science Foundation grant DEB-0129219. Special thanks altitude has been show to have an effect on wing for Kirsten Klappert, Mike Ritchie, Martin Scha¨fer, morphology by increasing wing load because of colon- Christian Schlo¨ tterer and two anonymous referees for ization or the scarcity of food resources (Norry et al., constructive criticism and David Houle for the help with 2001). In the present study, the morphological diver- the wing analysis. Thanks are also because of the people gence in wing traits was most pronounced in the who helped in collecting and rearing the flies, especially Vancouver population, which was separated from the Kirsten Klappert, Luisa Orsini, Susanna Huttunen and Oulanka and Colorado populations by changes influen- Mari Saarikettu, and to Oulanka Biological Station and cing mainly the size and the internal landmarks of the the Rocky Mountain Biological Laboratory and LAPBIAT wings. Each population differed from the other two project. populations by changes mainly on the tip and the front and back edges of the wings. The differentiation, likely to References have aerodynamic consequences, could be because of natural selection causing the wing form to adjust to the Arnqvist, G. & Rowe, L. 2005. Sexual Conflict. Princeton aerodynamic optimum of a local climate. Also, a wing University Press, Princeton and Oxford. landmark pinpointing the endpoint of the second long Aspi, J. & Hoikkala, A. 1993. Laboratory and natural heritabil- vein on the outline of the wing showed covariation with ities of male courtship song characters in Drosophila montana the combined effect of three pulse characters of the song. and D. littoralis. Heredity 70: 400–406. Aspi, J. & Hoikkala, A. 1995. Male mating success and survival in However, further experiments are required to find out the field with respect to size and courtship song characters in what is the precise role of the wing morphology on song Drosophila littoralis and D. montana (Diptera: Drosophilidae). production. J. Insect Behav. 8: 67–87. The genitalia shape and size showed divergence Aspi, J. & Lankinen, P. 1991. Frequency of multiple insemina- between the Vancouver and Oulanka and Vancouver tion in a natural population of Drosophila montana. Hereditas and Colorado populations in the DA, but the cross- 117: 169–177. validated classification test revealed an overlap in these Aspi, J., Lumme, J., Hoikkala, A. & Heikkinen, E. 1993. traits between the Colorado and Oulanka populations. Reproductive ecology of the boreal riparian guild of Drosophila. Ecography 16: 65–72. Nested ANOVAs (see Table S1) revealed strong within strain variance for all the genitalia shape traits suggesting Chan, Y.L. & Arcese, P. 2003. Morphological and Microsatellite differentiation in Melospiza melodia (Aves) at a Microgeographic them to be sensitive to environmental factors or to effects Scale. J. Evol. Biol. 16: 936–947. of sample preparation. D. montana females mate repeat- Cowling, D.E. & Burnet, B. 1981. Courtship songs and genetic edly in the wild, with the last male siring most of the control of their acoustic characteristics in sibling species of offspring (Aspi & Lankinen, 1991), and so selective

ª 2007 THE AUTHORS 20 (2007) 1591–1601 JOURNAL COMPILATION ª 2007 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY 1600 J. ROUTTU ET AL.

the Drosophila melanogaster subgroup. Anim. Behav. 29: 924– Kuhl, F.P. & Giardina, C.R. 1982. Elliptic Fourier features of a 935. closed contour. Comp. Graph. Imag. Proc. 18: 236–258. Dobzhansky, T. 1951. Genetics and the Origin of Species. Columbia Kulikov, A.M., Melnikov, A.I., Gornostaev, N.G., Lazebny, O.E. University Press, New York. & Mitrofanov, V.G. 2004. Morphological analysis of male Dryden, I.L. & Mardia, K.V. 1998. Statistical Shape Analysis. John mating organ in the Drosophila virilis species group: a multi- Wiley and Sons, Chichester, UK. variate approach. J. Zool. Syst. Evol. Res. 42: 135–144. Dufour, L. 1844. Anatomie generale des dipteres. Ann. Sci. Nat. 1: Lambert, D.M. & Henderson, N.R. 1986. The stability of the 244–264. specific-mate recognition system of Drosophila melanogaster. Eberhard, W.G. 1985. Sexual Selection and Animal Genitalia. Behav. Genet. 16: 369–373. Harvard University Press, Cambridge. Lande, R. 1982. Rapid origin of sexual isolation and character Etges, W.J., Over, K.F., Cardoso de Oliveira, C. & Ritchie, M.G. divergence in a cline. Evolution 36: 213–223. 2006. Inheritance of courtship song variation among geogra- Liimatainen, J. & Hoikkala, A. 1998. Interactions of the males phically isolated populations of Drosophila mojavensis. Anim. and females of three sympatric Drosophila virilis group species, Behav. 71: 1205–1214. D. montana, D. littoralis and D. lummei, in intra- and Fisher, R.A. 1930. The Genetical Theory of Natural Selection. interspecific courtships in the wild and in the laboratory. J. Clarendon Press, Oxford. Insect Behav. 11: 399–417. Gilchrist, G.W., Huey, R.B. & Serra, L. 2000. Rapid evolution of Liimatainen, J., Hoikkala, A., Aspi, J. & Welbergen, Ph. 1992. wing size clines in Drosophila subobscura. Genetica 112–113: Courtship in Drosophila montana: the effects of male auditory 273–286. signals on the behaviour of the flies. Anim. Behav. 43: 35– Grandcolas, P. & D’Haese, C. 2003. Testing adaptation with 48. phylogeny: how to account for phylogenetic pattern and van der Linde, K. 2005–2006. Geometrics Package. http://www. selective value together? Zoologica Scripta 32: 483–490. kimvdlinde.com/professional/geometrics.html Grimaldi, D.A. 1990. A phylogenetic, revised classification of van der Linde, K. & Houle, D. 2004–2006. Wings. http://www. genera in the Drosophilidae (Diptera). Bull. Am. Mus. Nat. Hist. kimvdlinde.com/professional/wings.html 197: 1–139. van der Linde, K. & Houle, D. 2006. Applied Usage of the Heisler, L., Andersson, M.B., Arnold, S.J., Boak, C.R., Borgia, G., Minimum-Volume Ellipsoid. [www document]. URL http:// Hausfater, G., Kirkpatrik, M., Lande, R., Maynard Smith, J., www.kimvdlinde.com/professional/mvepaper.html. O’Donald, P., Thornhill, A.R. & Weissing, F.J. 1987. The Lu, F. & Houle, D. 1995–1997. FindWing. http://bio.fsu.edu/ evolution of mating preferences and sexually selected traits. dhoule/Software/ In: Sexual Selection: Testing the Alternatives (J. W. Bradbury & M. Mayr, E. 1963. Animal Species and Evolution. Harvard University B. Andersson, eds), pp. 97–118. John Wiley and Sons, New Press, Cambridge. York. McKay, J.K. & Latta, R.G. 2002. Adaptive population diver- Hewitt, G.M. 2004. Genetic consequences of climatic oscillations gence: markers, QTL and traits. Trends Ecol. Evol. 17: 285– in the Quaternary. Philos. Trans. R. Soc. Lond. B. 359: 183–195. 291. Hoikkala, A. & Lumme, J. 1987. The genetic basis of evolution of Merila¨, J. & Crnokrak, P. 2001. Comparison of genetic the male courtship sounds in the Drosophila virilis group. differentiation at marker loci and quantitative traits. J. Evol. Evolution 41: 827–845. Biol. 14: 892–903. Hoikkala, A. & Suranto, L. 1999. Male courtship song frequency Mezey, J.G. & Houle, D. 2005. The dimensionality of genetic as an indicator of male mating success in Drosophila montana. variation for wing shape in Drosophila melanogaster. Evolution J. Insect Behav. 12: 599–609. 59: 1027–1038. Hoikkala, A., Aspi, J. & Suvanto, L. 1998. Male courtship song Mirol, P., Scha¨fer, M., Orsini, L., Routtu, J., Schlo¨ tterer, C., frequency as an indicator of male genetic quality in an insect Butlin, R.K. & Hoikkala, A. 2007. Phylogeographic patterns in species, Drosophila montana. Proc. R. Soc. Lond. B. 265: 503– Drosophila montana. Mol. Ecol. 16: 1085–1097. 508. Moorhead, P.S. 1954. Chromosome variation in giant forms of Holm, S. 1979. A simple sequentially rejective multiple test Drosophila montana. Univ. Texas Publ. 5422: 106–129. procedure. Scand. J. Stat. 6: 65–70. Norry, F.M., Bubliy, O.A. & Loeschcke, V. 2001. Developmental Hosken, D.J. & Stockley, P. 2004. Sexual selection and genital time, body size and wing loading in Drosophila buzzatii from evolution. Trends Ecol. Evol. 19: 87–93. lowland and highland populations in Argentina. Hereditas 135: Houde, A.E. 1993. Evolution by sexual selection: what can 35–40. population comparisons tell us?. Am. Nat. 141: 796–803. Pa¨a¨llysaho, S., Vieira, C.P., Hoikkala, A. & Vieira, J. 2005. Houle, D., Mezey, J.G., Galpern, P. & Carter, A. 2003. Automa- Evidence for introgression in differentiated North-American ted measurement of Drosophila wings. BMC Evol. Biol. 3: 25. and Finnish Drosophila montana populations. Genetica 123: Huey, R.B., Gilchrist, G.W., Carlson, M.L., Berrigan, D. & Serra, 285–293. L. 2000. Rapid evolution of a geographic cline in size in an Parker, G.A. 1970. Sperm competition and its evolutionary introduced fly. Science 287: 308–309. consequences in insects. Biol. Rev. Camb. Philos. Soc. 45: 525– Iwata, H. & Ukai, Y. 2002. SHAPE: a computer program package 567. for quantitative evaluation of biological shapes based on Rice, W.R. 1989. Analyzing tables of statistical tests. Evolution 43: elliptic Fourier descriptors. J. Hered. 93: 384–385. 223–225. Kennington, W.J., Killeen, J.R., Goldstein, D.B. & Partridge, L. Ritchie, M., Saarikettu, M., Livingstone, S. & Hoikkala, A. 2001. 2003. Rapid laboratory evolution of adult wing area in Characterization of female preference functions for Drosophila Drosophila melanogaster in response to humidity. Evolution 57: montana courtship song and a test of the temperature coupling 932–936. hypothesis. Evolution 55: 721–727.

ª 2007 THE AUTHORS 20 (2007) 1591–1601 JOURNAL COMPILATION ª 2007 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY Male song, wing and genital characters 1601

Rousseeuw, P.J. & Leroy, A.M. 1987. Robust Regression and Outlier Wright, S. 1931. Evolution in Mendelian populations. Genetics Detection. Wiley-Interscience, New York. 16: 97–159. Rousseeuw, P.J. & van Zomeren, B.C. 1990. Unmasking multi- variate outliers and leverage points. J. Am. Stat. Assoc. 85: 633–639. Supplementary material Saarikettu, M., Liimatainen, J.O. & Hoikkala, A. 2005. The role of male courtship song in species recognition in Drosophila The following supplementary material is available for this 35 montana. Behav. Genet. : 257–263. article: Santos, M., Iriarte, P.F., Ce´spedes, W., Balanya`, J., Fontdevila, A. Table S1 Nested ANOVAs. & Serra, L. 2004. Swift laboratory thermal evolution of wing shape (but not size) in Drosophila subobscura and its relation- This material is available as part of the online article from: ship with chromosomal inversion polymorphism. J. Evol. Biol. http://www.blackwell-synergy.com/doi/abs/10.1111/j. 17: 841–855. 1420-9101.2007.01323.x Sokal, R.R. & Rohlf, F.J. 1997. Biometry. The Principles and Practice Please note: Blackwell Publishing are not responsible for of Statistics in Biological Research, 3rd edn. Freeman, New York. the content or functionality of any supplementary Suvanto, L., Liimatainen, J. & Hoikkala, A. 1998. Variability and materials supplied by the authors. Any queries (other evolvability of male song characters in Drosophila montana than missing material) should be directed to the populations. Hereditas 130: 13–18. corresponding author for the article. Throckmorton, L.H. 1982. The Genetics and Biology of Drosophila, Vol. 3b. Academic Press, New York. Received 8 December 2006; revised 22 January 2007; accepted 24 Weber, K., Johnson, N., Champlin, D. & Patty, A. 2005. Many January 2007 P-element insertions affect wing shape in Drosophila melano- gaster. Genetics 169: 1461–1475.

ª 2007 THE AUTHORS 20 (2007) 1591–1601 JOURNAL COMPILATION ª 2007 EUROPEAN SOCIETY FOR EVOLUTIONARY BIOLOGY

V

MICROSATELLITE-BASED SPECIES IDENTIFICATION METHOD FOR DROSOPHILA VIRILIS GROUP SPECIES

by

Jarkko Routtu, Anneli Hoikkala and Maaria Kankare

Submitted Manuscript

Microsatellite-based species identification method for Drosophila virilis group species

Jarkko Routtu, Anneli Hoikkala and Maaria Kankare

Department of Biological and Environmental Science, University of Jyväskylä, Jyväskylä, Finland

ABSTRACT

Species of the D. virilis group are widely used in evolutionary research, but the individuals of different species are difficult to distinguish from each other morphologically. We constructed a fast and easy microsatellite-based identification method for the species of the group occurring sympatrically in northern Europe. The neighbor joining tree based on 14 microsatellite loci also gave a good resolution of a species divergence pattern in the whole group. 2

INTRODUCTION

The Drosophila virilis group species have been extensively used in studies concerning genome structure, speciation and evolution of species-specific behavioural characters (see e.g. Patterson and Stone 1952; Throckmorton 1982; Hoikkala et al. 2005; Routtu et al. 2007). The group consists of 11-14 species or subspecies, depending on whether D. canadiana is included, whether D. borealis is separated into eastern and western forms and whether D. americana americana and D. americana texana are classified as subspecies. Spicer and Bell (2002) have divided the group into four phylads based on the phylogeny of mitochondrial 12S and 16S rRNA genes: the virilis phylad (D. virilis, D. americana americana, D. a. texana, D. novamexicana and D. lummei), the montana phylad (D. montana, D. lacicola, D. borealis eastern and D. borealis western and D. flavomontana), the littoralis phylad (D. littoralis, D. canadiana and D. ezoana) and the kanekoi phylad represented by D. kanekoi. The species of D. virilis group are especially suitable for speciation studies as they represent different levels of divergence with an estimated time of 9.0 ± 0.7 mya between virilis and montana phylad species and 2.6 ± 0.4 mya between D. virilis and the other species in the virilis phylad (Nurminsky et al.1996). Many of these species can be crossed with each other to study the genetic basis of interspecific differences (Throckmorton 1982). Also, geographically isolated populations of some species of the group offer a good opportunity to study the effects of ecological and behavioural factors on population divergence, and to trace the genetic mechanisms underlying the process. For example the divergence time between the D. montana populations from different continents is estimated to be from 0.45 to 0.9 mya (Päällysaho et al. 2005; Mirol et al. 2007). Species of the D. virilis group share some morphological characters such as a relatively large size for Drosophilidae and a dark shade around the longest cross-vein of the wing, which make identification at the group level easy. The species are, however, difficult to distinguish from each other. Researchers have used different methods for identifying the species of wild-collected flies, such as classifying the flies on the basis of external male genitalia (Lakovaara and Hackman 1973; Kulikov et al. 2004) female spermathecae (Pitnick et al. 1999), mating flies with laboratory-reared flies of known species (Liimatainen and Hoikkala 1998) and recording and analyzing the songs of wild-caught males (Huttunen et al. 2002). Species-identification has also been made by using protein gel electrophoresis (e.g. Lakovaara et al. 1976; Spicer 1991) and RAPD fingerprinting (Mikhailovsky et al. 2007). Unfortunately, these methods are not very accurate and/or they are very time consuming and require special expertise. Microsatellite markers have been used earlier to explore the phylogeny of the D. virilis group species (Orsini et al. 2004) and to map genes that affect male song traits (Huttunen et al. 2004). Orsini et al. (2004) focused their study on the species of the virilis phylad. We have used here a similar approach with a 3 special emphasis on North-European species. The main aim of this study was to develop a fast and easy species identification method for the D. virilis group species occurring sympatrically in northern Europe (D. montana, D. littoralis, D. ezoana and D. lummei; see Aspi et al. 1993; Bächli et al. 2005). For this purpose the microsatellite locus Vir72ms appeared to be especially informative. The 14 polymorphic microsatellite loci used in the study amplified in all D. virilis group species, which gave us an opportunity also to check how well they cluster the fly strains of different species throughout the group and to identify possible misclassifications.

MATERIAL AND METHODS

50 microsatellite loci (20 loci from Orsini and Schlötterer 2004 and 30 microsatellites designed on the basis of the available D. virilis scaffolds (Schäfer et al. in prep.)) were screened with PCR for all D. virilis group species. The list of strains/individual flies representing the study species with information on their geographical origin is given in Appendix 1. A total of 14 loci could be amplified in all species and therefore they were selected for further analyses (Table 1). Genomic DNA was extracted from the fly samples using the high salt extraction method (Miller et al. 1988) adapted to Drosophila flies. PCR was performed using fluorescent 6-Fam or Yakima yellow (TAGC, Copenhagen, Denmark) labelled forward primer in a 10 μl reaction volume with 50-100 ng genomic DNA, 1 μM of each primer, 200 μM dNTPs, 1.5 mM MgCl2 and 1 unit of Taq DNA Polymerase (BioTools). The PCR profile was kept for 5 min at 95°C, followed by 30 cycles of 1 min at 95°C, 30 s at 52-59°C and 30 s at 72°C, and finally one cycle of 10 min at 72°C in a GeneAmp PCR system 9700 thermal cycler (Applied Biosystems). Alleles were separated with formamide and analysed with an ABI 3100 automatic sequencer (Applied Biosystems). The resulting spectrograms were genotyped with GeneMapper 3.7 software (Applied Biosystems). The Excel Microsatellite Toolkit (Park 2001) was used to calculate the mean number of alleles (MNA) and allele ranges over all loci for each species. An individual-based microsatellite genetic distance matrix with unequal sample sizes for different species was constructed with GenAlEx 6.031 (Peakall and Smouse 2006), where a genetic distance for a single locus with i, j, k and l alleles is defined as a set of squared distances; d2(ii,ii) =0, d2(ij,ij) = 0, d2(ii,ij) = 1, d2(ij,ik)=1, d2(ij,kl)=2, d2(ii,jk)=3 and d2(ii,jj)=4. Genetic distances were visualised with MEGA 3.1 (Kumar et al. 2004) (Fig. 1).The program STRUCTURE (Pritchard et al. 2000) with a Bayesian model-based clustering method was used to assign individuals to homogenous clusters (species). This method estimates the fraction of individual multilocus genotypes belonging to each cluster. The number of clusters is inferred by calculating the probability P(X|K) of the data, given a certain prior value of K (number of clusters) over a number of Markov 4 chain Monte Carlo (MCMC) iterations. The scores of individuals in the clusters correspond to the probability of ancestry in any one of them. An admixture model with correlated allele frequencies was used for the clustering analysis. Several runs of various lengths (10,000) were performed for each number of clusters (K), testing K values from 7 to 14. Finally, to choose the best value of K, two independent runs (200,000 iterations, 200,000 steps) were done using K values 12 and 13.

RESULTS AND DISCUSSION

Species of the D. virilis group are difficult to identify at the species level and several of them could be classified as cryptic species. This causes a lot of difficulties in research, especially as the populations of many of these species occur sympatrically in the wild. Our aim was to make the identification of the flies collected in northern Europe easier, as well as to study species clustering at the group level and to detect possible misclassifications in laboratory strains. The new species identification method was based on a set of 14 polymorphic microsatellite loci, which can be amplified in all D. virilis group species (except one locus which did not amplify in D. borealis). Samples for the three species found in the wild in northern Europe (D. ezoana, D. littoralis and D. montana) and D. flavomontana included wild-caught flies in addition to the flies of laboratory strains of the species. Other species were represented by the flies of one to six laboratory strains (see Table 2). The mean number of alleles across all microsatellite loci ranged from 1.1 to 9.3 (Table 2). A distribution of allele frequencies among loci showed species- specific variation, which was most pronounced in the Vir72ms locus with several species specific diagnostic alleles with no overlap in size (Table 3). This locus separated D. kanekoi, D. littoralis/D. canadiana, D. ezoana, D. novamexicana and D. virilis into their own groups. Interestingly, D. a. americana shared one allele with D. a. texana and one with D. lummei. In addition, D. montana, D. flavomontana, D. borealis and D. lacicola shared one fixed or nearly fixed allele. The microsatellite locus Vir72ms appeared to be especially informative for identification of the North-European species of the group (D. montana, D. littoralis, D. ezoana and D. lummei); all of these species had diagnostic allele or alleles at this locus (Table 3). This locus will be especially valuable for detecting rare species like D. lummei which, to our knowledge, has not been collected and identified in Fennoscandia during the last 20 years. The neighbor joining tree based on 14 microsatellite loci (Fig. 1.) gave a good resolution of an overall species divergence pattern. It clustered D. virilis phylad species in one brand close to D. ezoana. Contrary to the groupings in Orsini et al. (2004) and Spicer and Bell (2002), D. a. americana grouped together with D. a. texana. This grouping is supported by Schäfer et al. (2006), suggesting that the two subspecies are just chromosomal forms of a single species. D. 5 montana populations were clustered in one branch subdivided into three different clades according to the geographical origin of the strains/individuals (Finland, Canada and USA). Other species of the montana phylad clustered close to D. montana as expected. One D. lacicola, one D. flavomontana, one D. lummei and one D. borealis strain clustered among nonconspecifics. D. kanekoi had its own well differentiated branch which is similar to phylogenies based on mtDNA 12S and 16S rRNA (Spicer and Bell 2002) and nuclear adh gene (Nurminsky et al. 1996). Multilocus genotypes were clustered into groups which were similar to those constructed with the neighbor joining tree in this study (Fig. 1). The assignment success was high (>81%) in almost all samples (Table 4). The model that explained the data best (P ~1.000) partitioned the D. virilis group species into twelve clusters, while all of the other models were inadequate (P<0.001). Clustering of multilocus genotypes placed all D. virilis group species except D. montana strains, one D. borealis strain, one D. lummei strain, two D. flavomontana strains and two D. lacicola strains into a single cluster. D. montana flies were assigned into three separate clusters according to their geographic origin (Finland, Canada and USA), except for the three Canadian individuals that had admixture origins of Canadian and USA clusters. As in the neighbor joining tree, one western D. borealis strain (0961.03) clustered with D. montana from Colorado, USA (with 98.5% success), D. canadiana clustered with D. littoralis (with 98.6% success), and D. a. americana clustered together with D. a. texana to a single cluster with 81.4% and 91.3% success, respectively. The species status for some members of the D. virilis group has been recently under debate (Nurminsky et al. 1996; Schlötterer 2000; Spicer and Bell 2002; Orsini and Schlötterer 2004; Schäfer et al. 2006). To get more information on the species status of D. canadiana and to clarify the misclassified D. borealis strain, we performed some interstrain crosses (data not shown). Both the neighbor joining tree and the clustering method clustered D. canadiana with D. littoralis. Interspecific hybrids between these two species appeared to be fully fertile showing no significant pre- and postzygotic barriers. Also the male courtship songs of the two species are indistinguishable (Hoikkala, unpublished results). Thus, the species status of D. canadiana is questionable. The D. borealis strain, which clustered with D. montana (Colorado) strains in the neighbor joining tree, mated and produced offspring with a D. montana (Colorado) strain without any pre- or postzygotic barrier symptoms, which suggested it to be D. montana. The fact that the flies of this strain did not produce offspring with the flies of western or eastern D. borealis strains gives further support to this conclusion. Finally, we tested our species identification method for the flies collected in Jyväskylä, Finland (620 15’ N, 250 45’ E) in spring 2007. The Vir72ms locus separated the flies of different species accurately and fast: among thirty genotyped individuals, 22 flies were D. littoralis and 8 flies D. montana. 6

Acknowledgements

We are especially grateful to Martin Schäfer for designing the microsatellite markers. Additionally, we would like to thank Maarit Kokkonen for excellent laboratory assistance. Thanks are also due to Mikael Mökkönen for reviewing the language. Laboratory strains were obtained from the University of Oulu, Finland (Jaana Liimatainen) and Tucson Drosophila Stock Center. The work has been supported by The Finnish Centre of Excellence in Evolutionary Research. In addition, JR was funded by Finnish Cultural Foundation and MK by Academy of Finland, project number 7114064. 7

REFERENCES

Aspi, J., Lumme, J., Hoikkala, A. and Heikkinen, E. 1993. Reproductive ecology of the boreal riparian guild of Drosophila. - Ecography 16: 65-72. Bächli, G., Vilela, C.R., Andersson Esher, S. and Saura, A. 2005. The Drosophilidae (Diptera) of Fennoscandia and Denmark. Fauna Entomologica Scandinavica, Vol. 39. - Brill, Leiden. Hoikkala, A., Klappert, K. and Mazzi, D. 2005. Factors affecting male song evolution in Drosophila montana. - Curr. Top. Devel. Biol. 67: 225-250. Huttunen S., Vieira J. and Hoikkala A. 2002. Nucleotide and repeat length variation at the nonA gene of the Drosophila virilis group species and its effects on male courtship song. -Genetica 115: 159-167. Huttunen, S., Aspi, J., Hoikkala, A. and Schlötterer, C. 2004. QTL analysis of variation in male courtship song characters in Drosophila virilis. - Heredity 92: 263-269. Kulikov, A.M., Melnikov, A.I., Gornostaev, N.G., Lazebny, O.E., and Mitrofanov, V.G. 2004. Morphological analysis of male mating organ in the Drosophila virilis species group: a multivariate approach. - J. Zool. Syst. Evol. Res. 42:135-144. Kumar, S., Tamura, K. and Nei, M. 2004. MEGA3: Integrated software for Molecular Evolutionary Genetics Analysis and sequence alignment. – Brief. Bioinf. 5:150-163. Lakovaara, S. and Hackman, W. 1973. Drosophila ovivorum sp. n., a new species of the Drosophila virilis group (Diptera, Drosophilidae). - Notul. Entomol. 53:167-169. Lakovaara, S.A., Saura, A., Lankinen, P., Pohjola, L. and Lokki, J. 1976. The use of isoenzymes in tracing evolution and in classifying Drosophilidae. - Zool. Scripta 5:173-179. Liimatainen, J.O. and Hoikkala, A. 1998 Interactions of the males and females of three sympatric Drosophila virilis group species, D. montana, D. littoralis and D. lummei, (Diptera : Drosophilidae) in intra- and interspecific courtships in the wild and in the laboratory. - J. Insect Behav. 11: 399-417. Mikhailovsky, S.S., Kulikov, A.M., Potapov, S.G., Lazebny, O.E. & Mitrofanov, V.G. 2007. A RAPD fingerprinting of sibling species of the Drosophila virilis group. - Rus. J. Genet. 43: 105-109. Miller, S.A., Dykes, D.D. and Polesky, H.F. 1988. A simple salting out procedure for extracting DNA from human nucleated cells. - Nuc. Acids Res. 16: 1215. Mirol, P., Schäfer, M.A., Orsini, L., Routtu, J., Schlötterer, C., Hoikkala, A. and Butlin, R.K. 2007. Phylogeographic patterns in Drosophila montana. - Mol. Ecol. 16: 1085-1097. Nurmisky, D.I., Moriyama, E.N., Lozovskaya, E.R. and Hartl D.L 1996. Molecular phylogeny and genome evolution in the Drosophila virilis species group: duplications of the alcohol dehydrogenase gene. - Mol. Biol. Evol. 13: 132-149. 8

Orsini, L. and Schlötterer, C. 2004. Isolation and characterization of microsatellites in Drosophila montana and their cross-species amplification in D. virilis. - Mol. Ecol. Notes. 4: 412–414. Orsini, L., Huttunen, S. and Schlötterer, C. 2004. A multilocus microsatellite phylogeny of the Drosophila virilis group. - Heredity 93: 161-165. Park, S. D. E. 2001. Trypanotolerance in West African cattle and the population genetic effects of selection [Ph.D. thesis] - University of Dublin. Patterson, J. T. and Stone, W. S. 1952. Evolution in the genus Drosophila. - New York: Macmillan. Peakall, R. and Smouse, P.E. 2006. GENALEX 6: genetic analysis in Excel. Population genetic software for teaching and research. - Mol. Ecol. Notes. 6: 288-295. Pritchard, J.K., Stephens, M. and Donnelly, P. 2000. Inference of population structure using multilocus genotype data. - Genetics 155: 945–959. Pitnick, S., Markow, T. and Spicer, G.S. 1999. Evolution of multiple kinds of female sperm-storage organs in Drosophila. - Evolution 53: 1804-1822. Päällysaho, S., Vieira, C., Hoikkala, A. and Vieira, J. 2005. Evidence for introgression in differentiated North-American and Finnish Drosophila montana populations. - Genetica 123: 285-293. Routtu, J., Mazzi, D. Van der Linde, K., Mirol, P., Butlin, R.K. and Hoikkala, A. 2007. The extent of variation in male song, wing and genital characters among allopatric Drosophila montana populations. - J. Evol. Biol. 20:1591- 1601. Schlötterer, C. 2000. Microsatellite analysis indicates genetic differentiation of the neo-sex chromosomes in Drosophila americana americana. - Heredity 85: 610-616. Schäfer, M.A., Orsini, L., McAllister, B.F. and Schlötterer, C. 2006. Patterns of microsatellite variation through a transition zone of a chromosomal cline in Drosophila americana. - Heredity 97: 291-295. Spicer, G.S. 1991. Molecular evolution of the Drosophila virilis species group as inferred by 2-dimensional electrophoresis. - J. Mol. Evol. 33: 379-394 Spicer, G.S. and Bell, C.D. 2002. Molecular phylogeny of the Drosophila virilis species group (Diptera: Drosophilidae) inferred from mitochondrial 12S and 16S ribosomal RNA genes. - Ann. Entom. Soc. Amer. 95: 156-161. Throckmorton, L.H. 1982. The Genetics and Biology of Drosophila Vol. 3b. - Academic Press, New York. 9

FIGURE 1 Unrooted neighbor joining tree based on 14 microsatellite loci. Misclassified individuals: *1; D. flavomontana strain 15010-0981.00, *2; D. lummei strain 1100, *3; D. borealis strain 15010-0961.03 and *4; D. lacicola strain 15010-0991.13. 10

TABLE 1 Microsatellite loci used in this study with chromosome position, expected size, locus name, annealing temperature, repeat type and primer sequences. Microsatellites developed by Schäfer et al. (in prep.), except for Mon26, Mon6, Mon17a and Mon20, developed by Orsini and Schlötterer (2004).

Chromosome Expected Locus TaqC Microsatellite Primer size repeat sequences (5’ – 3’) X152 Vir99ms 55 CA F: acaatgcttgcacaatgacg R: ccatgcaaattgtgaactgc X140 Vir72ms 55 TA F: ggtccctgctcagaacaact R: cgctgcttaagccaacaata 2204 Vir32ms 55 CA F: gggtgttgatgtcgagtgtg R: aagaagtccaaagcgctcaa 2155 Vir4ms 55 GT F: ttgcaatattcccccatttc R: gcggcagaaatgacattgac 2245 Vir38ms 55 CA F: gaattcgcaatgcacgtaaa R: cgacgtatctgtgagccact 3157 Vir90ms 55 GT F: caattaaaaggaccgcctga R: tcattatgcggaaatgctga 3 139 Vir93ms 55 GA F: acgtggtccaagcaatttgt R: tgagctcccgaccagtttag 3149 Vir11ms 55 TA F: gcaaaacatgaataatgcgaac R: ctttgacaatggcaccacac 4162 Vir17ms 55 GT F: tgcaacgttcggtagtcaag R: cggctttgggtcttagattg 5149 Vir74ms 55 TA F: ggcaggttttatcaggcaac R: tggcatctcgatacgcataa X 114–144 Mon26 52 TG F: gagtggcagacacaacctca R: gccaacagtgcacgtaatttt 2 156–168 Mon6 59 GTT F: gtccgaaccacgcaataact R: gctgttgatgatgatgaggc Unknown 134–162 Mon17a 56 GT F: atatctgtgcagaggcaggg R: tgaaattcaagtgcagcgac Unknown 144–148 Mon20 58,5 TG F: gcagcagccacaatatcaaa R: ggctgctgttgttaaaggct 11

TABLE 2 Species of D. virilis group with information on their sample size, the number of microsatellite loci typed and the mean number of alleles over all studied loci.

Species Number Number of Number Mean number of strains wild-caught of of individuals loci alleles D. a. americana 4 14 2.8 D. a. texana 4 14 3.1 D. borealis 3 13 2.3 D. canadiana 1 14 1.1 D. ezoana* 3 24 14 5.8 D. flavomontana 2 3 14 2.3 D. kanekoi 1 14 1.0 D. lacicola 2 14 2.0 D. littoralis* 5 32 14 6.8 D. lummei* 5 14 2.1 D. montana* 32 25 14 9.3 D. novamexicana 3 14 1.5 D. virilis 6 14 3.1 * Species occurring sympatrically in northern Europe (D. lummei has not been found in the wild for several years).

TABLE 3 Allele frequencies (%) in the Vir72ms locus for all species in the D. virilis group. The frequencies of the alleles separating the species of the North- European species community are shown in bold. Abbreviations are AME (D. a. americana), BOR (D. borealis), CAN (D. canadiana), EZO (D. ezoana), FLA (D. flavomontana), KAN (D. kanekoi), LAC (D. lacicola), LIT (D. littoralis), LUM (D. lummei), MON (D. montana), NOV (D. novamexicana), TEX (D. a. texana) and VIR (D. virilis).

VIR AME TEX LUM NOV LIT CAN MON FLA BOR LAC EZO KAN 84 100.0 86 100.0 100.0 88 97.4 100.0 100.0 100.0 90 100.0 92 100.0 94 2.6 100 66.7 100.0 102 100.0 104 33.3 100.0 12 TABLE 4 Minimum percentage of D. virilis group species belonging to the 12 clusters inferred by the multilocus clustering method. Numbers of individuals assigned to a certain cluster are given in the parentheses.

Species N 1 2 3 4 5 6 7 8 9 10 11 12 D. a. americana 4 81.4 (4) D. a. texana 4 91.3 (4) D. borealis 3 98.5 (1) 98.7 (2)* D. canadiana 1 98.6 (1) D. ezoana 27 84.7 (27) D. flavomontana 4 98.6 (1) 97.8 (3) D. kanekoi 1 98.2 (1) D. lacicola 2 97.8 (1) 74.0 (1) D. littoralis 36 89.8 (36) D. lummei 6 97.6 (5) 78.7 (1) D. novamexicana 3 98.6 (3) D. virilis 5 89.1 (5) D. montana 57 87.4 (9) 83.6 (36) 77.4 (11) * One D. borealis individual was assigned to this cluster with 62.9% success. 13

APPENDIX 1 The strains/individuals of the Drosophila virilis group species with their ID number, strain code, geographical origin and collection year, if known.

SPECIES ID STRAIN/ INDIVIDUAL ORIGIN YEAR D. a. americana AME1 G 96.47 Gary, Indiana, USA AME2 G 96.48 Gary, Indiana, USA AME3 NN 97.2 Niobrara, Nebraska, USA AME4 NN 97.8 Niobrara, Nebraska, USA D. a. texana TEX1 ML 97.3 Monroe, Louisiana, USA TEX2 ML 97.5 Monroe, Louisiana, USA TEX3 LP 97.7 Lone Star, Texas, USA TEX4 CD 97.5 Lone Star, Texas, USA D. borealis BOR1 C3F7 Crested Butte, Colorado, USA 2003 BOR2 0961.00 Itasca Park, Minnesota, USA 1950 BOR3 0961.03 Chester, Idaho, USA 1949 BOR4 0961.05 Quebec, Canada 1949 D. canadiana CAN1 1091.00 British Columbia, Canada D. ezoana EZO1 EZO 2 Kemi, Finland 1995 EZO2 E 04.3 Oulanka, Finland 2004 EZO3 EZO / OU Oulu, Finland 2003 EZO4 O 25* Oulanka, Finland 2002 EZO5 O 59* Oulanka, Finland 2002 EZO6 O 70* Oulanka, Finland 2002 EZO7 O 71* Oulanka, Finland 2002 EZO8 O3F13* Oulanka, Finland 2003 EZO9 O3F17* Oulanka, Finland 2003 EZO10 03F22* Oulanka, Finland 2003 EZO11 O3F29* Oulanka, Finland 2003 EZO12 O3F46* Oulanka, Finland 2003 EZO13 O3F49* Oulanka, Finland 2003 EZO14 O3F55* Oulanka, Finland 2003 EZO15 O3F62* Oulanka, Finland 2003 EZO16 O3F69* Oulanka, Finland 2003 EZO17 O3F70* Oulanka, Finland 2003 EZO18 O3F75* Oulanka, Finland 2003 EZO19 O3F76* Oulanka, Finland 2003 EZO20 O3F84* Oulanka, Finland 2003 EZO21 O3F88* Oulanka, Finland 2003 EZO22 O3F102* Oulanka, Finland 2003 EZO23 O3F108* Oulanka, Finland 2003 EZO24 O3F115* Oulanka, Finland 2003 EZO25 O3F118* Oulanka, Finland 2003 EZO26 O3F119* Oulanka, Finland 2003 EZO27 O3F124* Oulanka, Finland 2003 D. flavomontana FLA1 0981.00 Chester, Idaho, USA 1949 FLA2 0981.02 Craig, Colorado, USA 1949 FLA3 COL 1* Colorado, USA 2003 FLA4 COL 3* Colorado, USA 2003 FLA5 COL 5* Colorado, USA 2003 D. kanekoi KAN1 1061.00 Sapporo, Japan D. lacicola LAC1 0991.13 Manitoba, Canada 1949 LAC2 0991.12 Fenske Lake, Minnesota, USA 1947 D. littoralis LIT1 LITTO 1 Jyväskylä, Finland 2006 LIT2 LI 10 Don-river, Rostov, Russia 2000 LIT3 1052 Batumi, Georgia, Russia 1973 14

LIT4 OU 7 Oulanka, Finland 2004 LIT5 KU 05.5 Kuopio, Finland 2005 LIT6 LIT 2* Jyväskylä, Finland 2006 LIT7 LIT 3* Jyväskylä, Finland 2006 LIT8 LIT 4* Jyväskylä, Finland 2006 LIT9 LIT 5* Jyväskylä, Finland 2006 LIT10 LIT 1* Jyväskylä, Finland 2006 LIT11 04.1198* Jyväskylä, Finland 2006 LIT12 04.1241* Jyväskylä, Finland 2006 LIT13 04.1254* Jyväskylä, Finland 2006 LIT14 04.1264* Jyväskylä, Finland 2006 LIT15 04.1281* Jyväskylä, Finland 2006 LIT16 04.1304* Jyväskylä, Finland 2006 LIT17 04.1308* Jyväskylä, Finland 2006 LIT18 04.1323* Jyväskylä, Finland 2006 LIT19 04.1326* Jyväskylä, Finland 2006 LIT20 04.1335* Jyväskylä, Finland 2006 LIT21 04.1346* Jyväskylä, Finland 2006 LIT22 04.1351* Jyväskylä, Finland 2006 LIT23 04.1401* Jyväskylä, Finland 2006 LIT24 04.1455* Jyväskylä, Finland 2006 LIT25 04.1491* Jyväskylä, Finland 2006 LIT26 04.1506* Jyväskylä, Finland 2006 LIT27 04.1514* Jyväskylä, Finland 2006 LIT28 04.1555* Jyväskylä, Finland 2006 LIT29 04.1580* Jyväskylä, Finland 2006 LIT30 04.1591* Jyväskylä, Finland 2006 LIT31 K 34* Kemi, Finland 2002 LIT32 K 47* Kemi, Finland 2002 LIT33 L 2* Lappajärvi, Finland 2002 LIT34 L 3 * Lappajärvi, Finland 2002 LIT35 L 4* Lappajärvi, Finland 2002 LIT36 L 5* Lappajärvi, Finland 2002 LIT37 L 6* Lappajärvi, Finland 2002 D. lummei LUM1 200 Moscow, Russia 1988 LUM2 LU 1 Kemi, Finland 1990 LUM3 1100 Vaajasalo, Kuopio, Finland 1969 LUM4 1101S, W;60 Överkalix, Sweden 1970 LUM5 1143 Sakata, Yamagata, Japan 1974 LUM6 LuJapFu Hokkaido, Japan 1974 D. montana MON1 C3F6 Crested Butte, Colorado, USA 2003 MON2 CAN3F4 Vancouver, Canada 2003 MON3 03F66 Oulanka, Finland 2003 MON4 MO 4 Oulu, Finland 2000 MON5 JKL Jyväskylä, Finland 2006 MON6 1021.13 Kawasaki, Japan MON7 C3F2 Crested Butte, Colorado, USA 2003 MON8 C3F13 Crested Butte, Colorado, USA 2003 MON9 CAN3F1 Vancouver, Canada 2003 MON10 CAN3F2A Vancouver, Canada 2003 MON11 CAN3F2B Vancouver, Canada 2003 MON12 CAN3F3 Vancouver, Canada 2003 MON13 CAN3F5 Vancouver, Canada 2003 MON14 CAN3F8 Vancouver, Canada 2003 MON15 CAN3F12 Vancouver, Canada 2003 15

MON16 CAN3F13 Vancouver, Canada 2003 MON17 CAN3F14 Vancouver, Canada 2003 MON18 CAN3F15 Vancouver, Canada 2003 MON19 O3F20 Oulanka, Finland 2003 MON20 O3F39 Oulanka, Finland 2003 MON21 O3F43 Oulanka, Finland 2003 MON22 O3F52 Oulanka, Finland 2003 MON23 O3F63 Oulanka, Finland 2003 MON24 O3F77 Oulanka, Finland 2003 MON25 O3F79 Oulanka, Finland 2003 MON26 O3F80 Oulanka, Finland 2003 MON27 O3F94 Oulanka, Finland 2003 MON28 O3F97 Oulanka, Finland 2003 MON29 C3F1* Crested Butte, Colorado, USA 2003 MON30 C3F5* Crested Butte, Colorado, USA 2003 MON31 C3F8* Crested Butte, Colorado, USA 2003 MON32 C3F9* Crested Butte, Colorado, USA 2003 MON33 L 1* Lappajärvi, Finland 2002 MON34 L 5* Lappajärvi, Finland 2002 MON35 L 7* Lappajärvi, Finland 2002 MON36 L 7.2* Lappajärvi, Finland 2002 MON37 L 7.3* Lappajärvi, Finland 2002 MON38 K 10.1* Kemi, Finland 2002 MON39 K 12.2* Kemi, Finland 2002 MON40 K 14.1* Kemi, Finland 2002 MON41 K 15.1* Kemi, Finland 2002 MON42 K 22.1* Kemi, Finland 2002 MON43 JKL 1* Jyväskylä, Finland 2006 MON44 JKL 2* Jyväskylä, Finland 2006 MON45 JKL 3* Jyväskylä, Finland 2006 MON46 JKL 4* Jyväskylä, Finland 2006 MON47 JKL 5* Jyväskylä, Finland 2006 MON48 K 21* Kemi, Finland 2002 MON49 K 36* Kemi, Finland 2002 MON50 K 48* Kemi, Finland 2002 MON51 O 61* Oulanka, Finland 2002 MON52 O3F26 Oulanka, Finland 2003 MON53 O3F42 Oulanka, Finland 2003 MON54 O3F106 Oulanka, Finland 2003 MON55 O3F112 Oulanka, Finland 2003 MON56 COL 2* Colorado, USA 2003 MON57 COL 4* Colorado, USA 2003 D. novamexicana NOV1 1031.00 Grand Junction, Colorado, USA NOV2 1031.04 Moab, Utah, USA 1949 NOV3 1031.08 San Antonio, New Mexico, USA 1947 D.virilis VIR1 V-WW-03 Wuwei, China 2002 VIR2 V-QUFU Qufu, China 2002 VIR3 TOYAMA-1 Toyama, Japan 2004 VIR4 TOYAMA-11 Toyama, Japan 2004 VIR5 DNH Dunghuang, China 2002 VIR6 HUNAN Hunan, China 2002 *Wild-caught individual

BIOLOGICAL RESEARCH REPORTS FROM THE UNIVERSITY OF JYVÄSKYLÄ

1RAATIKAINEN, M. & VASARAINEN, A., Damage 10 LAKE PÄIJÄNNE SYMPOSIUM. 199 p. 1987. caused by timothy flies (Amaurosoma spp.) 11 SAARI, V. & OHENOJA, E., A check-list of the in Finland, pp. 3-8. larger fungi of Central Finland. 74 p. 1988. SÄRKKÄ, J., The numbers of Tubifex tubifex and 12 KOJOLA, I., Maternal investment in semi- its cocoons in relation to the mesh size, domesticated reindeer (Rangifer t. tarandus pp. 9-13. L.). 26 p. Yhteenveto 2 p. 1989. ELORANTA, P. & ELORANTA, A., Keurusselän 13 MERILÄINEN, J. J., Impact of an acid, polyhumic kalastosta ja sen rakenteesta. - On the fish river on estuarine zoobenthos and vegetation fauna of Lake Keurusselkä, Finnish Lake in the Baltic Sea, Finland. 48 p. Yhteenveto 2 p. District, pp. 14-29. 1989. ELORANTA, P. & ELORANTA, A., Kuusveden veden 14 LUMME, I., On the clone selection, ectomy- laadusta, kasviplanktonista ja kalastosta. - On corrhizal inoculation of short-rotation will- the properties of water, phytoplankton and ows (Salix spp.) and on the effects of some fish fauna of Lake Kuusvesi, Central Finland, nutrients sources on soil properties and pp. 30-47. 47 p. 1975. plant nutrition. 55 p. Yhteenveto 3 p. 1989. 2ELORANTA, V., Effects of different process 15 KUITUNEN, M., Food, space and time constraints wastes and main sewer effluents from pulp on reproduction in the common treecreeper mills on the growth and production of (Certhia familiaris L.) 22 p. Yhteenveto 2 p. Ankistrodesmus falcatus var. acicularis 1989. (Chlorophyta), pp. 3-33. 16 YLÖNEN, H., Temporal variation of behavioural ELORANTA, P. & KUNNAS, S., A comparison of and demographical processes in cyclic littoral periphyton in some lakes of Central Clethrionomys populations. 35 p. Yhteenveto Finland, pp. 34-50. 2 p. 1989. ELORANTA, P., Phytoplankton and primary 17 MIKKONEN, A., Occurrence and properties of production in situ in the lakes Jyväsjärvi and proteolytic enzymes in germinating legume North Päijänne in summer 1974, pp. 51-66. seeds. 61 p. Yhteenveto 1 p. 1990. 66 p. 1976. 18 KAINULAINEN, H., Effects of chronic exercise and 3RAATIKAINEN, M., HALKKA, O., VASARAINEN,A. & ageing on regional energy metabolism in heart HALKKA, L., Abundance of Philaenus muscle. 76 p. Yhteenveto 1 p. 1990. spumarius in relation to types of plant 19 LAKSO, MERJA, Sex-specific mouse testosterone community in the Tvärminne archipelago, 16 “-hydroxylase (cytochrome P450) genes: southern Finland. 38 p. 1977 characterization and genetic and hormonal 4HAKKARI, L., On the productivity and ecology regulations. 70 p. Yhteenveto 1 p. 1990. of zooplankton and its role as food for fish in 20 SETÄLÄ, HEIKKI, Effects of soil fauna on some lakes in Central Finland. 87 p. 1978. decomposition and nutrient dynamics in 5KÄPYLÄ, M., Bionomics of five woodnesting coniferous forest soil. 56 p. Yhteenveto 2 p. solitary species of bees (Hym., Megachilidae), 1990. with emphasis on flower relationships. 89 p. 21 NÄRVÄNEN, ALE, Synthetic peptides as probes 1978. for protein interactions and as antigenic 6KANKAALA, P. & SAARI, V., The vascular flora of epitopes. 90 p. Yhteenveto 2 p. 1990. the Vaarunvuoret hills and its conservation, 22 ECOTOXICOLOGY SEMINAR, 115 p. 1991. pp. 3-62. 23 ROSSI, ESKO, An index method for TÖRMÄLÄ, T. & KOVANEN, J., Growth and ageing environmental risk assessment in wood of magpie (Pica pica L.) nestlings, pp. 63-77. processing industry. 117 p. Yhteenveto 2 p. 77 p. 1979. 1991. 7VIITALA, J., Hair growth patterns in the vole 24 SUHONEN, JUKKA, Predation risk and Clethrionomys rufocanus (Sund.), pp. 3-17. competition in mixed species tit flocks. 29 p. NIEMI, R. & HUHTA, V., Oribatid communities in Yhteenveto 2 p. 1991. artificial soil made of sewage sludge and 25 SUOMEN MUUTTUVA LUONTO. Mikko Raatikaiselle crushed bark, pp. 18-30. 30 p. 1981. omistettu juhlakirja. 185 p. 1992. 8TÖRMÄLÄ, T., Structure and dynamics of 26 KOSKIVAARA, MARI, Monogeneans and other reserved field ecosystem in central Finland. parasites on the gills of roach (Rutilus rutilus) 58 p. 1981. in Central Finland. Differences between four 9ELORANTA, V. & KUIVASNIEMI, K., Acute toxicity lakes and the nature of dactylogyrid of two herbicides, glyphosate and 2,4-D, to communities. 30 p. Yhteenveto 2 p. 1992. Selenastrum capricornuturn Printz 27 TASKINEN, JOUNI, On the ecology of two (Chlorophyta), pp. 3-18. Rhipidocotyle species (Digenea: ELORANTA, P. & KUNNAS, S., Periphyton Bucephalidae) from two Finnish lakes. 31 p. accumulation and diatom communities on Yhteenveto 2 p. 1992. artificial substrates in recipients of pulp mill 28 HUOVILA, ARI, Assembly of hepatitis B surface effluents, pp. 19-33. antigen. 73 p. Yhteenveto 1 p. 1992. ELORANTA, P. & MARJA-AHO, J., Transect studies 29 SALONEN, VEIKKO, Plant colonization of on the aquatic inacrophyte vegetation of Lake harvested peat surfaces. 29 p. Yhteenveto 2 p. Saimaa in 1980, pp. 35-65. 65 p. 1982. 1992. BIOLOGICAL RESEARCH REPORTS FROM THE UNIVERSITY OF JYVÄSKYLÄ

30 JOKINEN, ILMARI, Immunoglobulin production 49 MARTTILA, SALLA, Differential expression of by cultured lymphocytes of patients with aspartic and cycteine proteinases, glutamine rheumatoid arthritis: association with disease synthetase, and a stress protein, HVA1, in severity. 78 p. Yhteenveto 2 p. 1992. germinating barley. 54 p. Yhteenveto 1 p. 1996 31 PUNNONEN, EEVA-LIISA, Ultrastructural studies 50 HUHTA, ESA, Effects of forest fragmentation on on cellular autophagy. Structure of limiting reproductive success of birds in boreal forests. membranes and route of enzyme delivery. 26 p. Yhteenveto 2 p. 1996. 77 p. Yhteenveto 2 p. 1993. 51 OJALA, JOHANNA, Muscle cell differentiation in 32 HAIMI, JARI, Effects of earthworms on soil vitro and effects of antisense oligode- processes in coniferous forest soil. 35 p. oxyribonucleotides on gene expression of Yhteenveto 2 p. 1993. contractile proteins. 157 p. Yhteenveto 2 33 ZHAO, GUOCHANG, Ultraviolet radiation induced p.1996. oxidative stress in cultured human skin 52 PALOMÄKI, RISTO, Biomass and diversity of fibroblasts and antioxidant protection. 86 p. macrozoobenthos in the lake littoral in Yhteenveto 1 p. 1993. relation to environmental characteristics. 27 p. 34 RÄTTI, OSMO, Polyterritorial polygyny in the Yhteenveto 2 p. 1996. pied flycatcher. 31 p. Yhteenveto 2 p. 1993. 53 PUSENIUS, JYRKI, Intraspecific interactions, space 35 MARJOMÄKI, VARPU, Endosomes and lysosomes use and reproductive success in the field vole. in cardiomyocytes. A study on morphology 28 p. Yhteenveto 2 p. 1996. and function. 64 p. Yhteenveto 1 p. 1993. 54 SALMINEN, JANNE, Effects of harmful chemicals 36 KIHLSTRÖM, MARKKU, Myocardial antioxidant on soil animal communities and enzyme systems in physical exercise and decomposition. 28 p. Yhteenveto 2 p. 1996. tissue damage. 99 p. Yhteenveto 2 p. 1994. 55 KOTIAHO, JANNE, Sexual selection and costs of 37 MUOTKA, TIMO, Patterns in northern stream sexual signalling in a wolf spider. 25 p. (96 p.). guilds and communities. 24 p. Yhteenveto Yhteenveto 2 p. 1997. 2 p. 1994. 56 KOSKELA, JUHA, Feed intake and growth 38 EFFECT OF FERTILIZATION ON FOREST ECOSYSTEM 218 variability in Salmonids. 27p. (108 p.). p. 1994. Yhteenveto 2 p. 1997. 39 KERVINEN, JUKKA, Occurrence, catalytic 57 NAARALA, JONNE, Studies in the mechanisms of properties, intracellular localization and lead neurotoxicity and oxidative stress in structure of barley aspartic proteinase. human neuroblastoma cells. 68 p. (126 p.). 65 p. Yhteenveto 1 p. 1994. Yhteenveto 1 p. 1997. 40 MAPPES, JOHANNA, Maternal care and 58 AHO, TEIJA, Determinants of breeding reproductive tactics in shield bugs. 30 p. performance of the Eurasian treecreeper. 27 p. Yhteenveto 3 p. 1994. (130 p.). Yhteenveto 2 p. 1997. 41 SIIKAMÄKI, PIRKKO, Determinants of clutch-size 59 HAAPARANTA, AHTI, Cell and tissue changes in and reproductive success in the pied perch (Perca fluviatilis) and roach (Rutilus flycatcher. 35 p. Yhteenveto 2 p. 1995. rutilus) in relation to water quality. 43 p. 42 MAPPES, TAPIO, Breeding tactics and (112 p.). Yhteenveto 3 p. 1997. reproductive success in the bank vole. 28 p. 60 SOIMASUO, MARKUS, The effects of pulp and Yhteenveto 3 p. 1995. paper mill effluents on fish: a biomarker 43 LAITINEN, MARKKU, Biomonitoring of approach. 59 p. (158 p.). Yhteenveto 2 p. 1997. theresponses of fish to environmental stress. 61 MIKOLA, JUHA, Trophic-level dynamics in 39 p. Yhteenveto 2 p. 1995. microbial-based soil food webs. 31 p. (110 p.). APPALAINEN EKKA 44 L , P , The dinuclear CuAcentre of Yhteenveto 1 p. 1997. cytochrome oxidase. 52 p. Yhteenveto 1 p. 62 RAHKONEN, RIITTA, Interactions between a gull 1995. tapeworm Diphyllobothrium dendriticum 45 RINTAMÄKI, PEKKA, Male mating success and (Cestoda) and trout (Salmo trutta L). 43 p. female choice in the lekking black grouse. 23 p. (69 p.). Yhteenveto 3 p. 1998. Yhteenveto 2 p. 1995. 63 KOSKELA, ESA, Reproductive trade-offs in the 46 SUURONEN, TIINA, The relationship of oxidative bank vole. 29 p. (94 p.). Yhteenveto 2 p. 1998. and glycolytic capacity of longissimus dorsi 64 HORNE, TAINA, Evolution of female choice in the muscle to meat quality when different pig bank vole. 22 p. (78 p.). Yhteenveto 2 p. 1998. breeds and crossbreeds are compared. 112 p. 65 PIRHONEN, JUHANI, Some effects of cultivation on Yhteenveto 2 p. 1995. the smolting of two forms of brown trout 47 KOSKENNIEMI, ESA, The ecological succession (Salmo trutta). 37 p. (97 p.). Yhteenveto 2 p. and characteristics in small Finnish 1998. polyhumic reservoirs. 36 p. Yhteenveto 1 p. 66 LAAKSO, JOUNI, Sensitivity of ecosystem 1995. functioning to changes in the structure of soil 48 HOVI, MATTI, The lek mating system in the food webs. 28 p. (151 p.). Yhteenveto 1 p. 1998. black grouse: the role of sexual selection. 30 p. 67 NIKULA, TUOMO, Development of radiolabeled Yhteenveto 1 p. 1995. monoclonal antibody constructs: capable of transporting high radiation dose into cancer cells. 45 p. (109 p.). Yhteenveto 1 p. 1998. BIOLOGICAL RESEARCH REPORTS FROM THE UNIVERSITY OF JYVÄSKYLÄ

68 AIRENNE, KARI, Production of recombinant 73 SIPPONEN, MATTI, The Finnish inland fisheries avidins in Escherichia coli and insect cells. system. The outcomes of private ownership of 96 p. (136 p.). Yhteenveto 2 p. 1998. fishing rights and of changes in administrative 69 LYYTIKÄINEN, TAPANI, Thermal biology of practices. 81 p. (188 p.). Yhteenveto 2 p. 1999. underyearling Lake Inari Arctic Charr 74 LAMMI, ANTTI, Reproductive success, local Salvelinus alpinus. 34 p. (92 p.). adaptation and genetic diversity in small plant Yhteenveto 1 p. 1998. populations. 36 p. (107 p.). Yhteenveto 4 p. 1999. 70 VIHINEN-RANTA, MAIJA, Canine parvovirus. 75 NIVA, TEUVO, Ecology of stocked brown trout in Endocytic entry and nuclear import. 74 p. boreal lakes. 26 p. (102 p.). Yhteenveto 1 p. 1999. (96 p.). Yhteenveto 1 p. 1998. 76 PULKKINEN, KATJA, Transmission of 71 MARTIKAINEN, ESKO, Environmental factors Triaenophorus crassus from copepod first to influencing effects of chemicals on soil animals. coregonid second intermediate hosts and Studies at population and community levels. 44 effects on intermediate hosts. 45 p. (123 p.). p. (137 p.). Yhteenveto 1 p. 1998. Yhteenveto 3 p. 1999. 72 AHLROTH, PETRI, Dispersal and life-history 77 PARRI, SILJA, Female choice for male drumming differences between waterstrider (Aquarius characteristics in the wolf spider Hygrolycosa najas) populations. 36 p. (98 p.). rubrofasciata. 34 p. (108 p.). Yhteenveto 1 p. 1999. Yhteenveto 2 p. 1999.

JYVÄSKYLÄ STUDIES IN BIOLOGICAL AND ENVIRONMENTAL SCIENCE

78 VIROLAINEN, KAIJA, Selection of nature reserve 86 HUOVINEN, PIRJO, Ultraviolet radiation in networks. - Luonnonsuojelualueiden valinta. aquatic environments. Underwater UV 28 p. (87 p.). Yhteenveto 1 p. 1999. penetration and responses in algae and 79 SELIN, PIRKKO, Turvevarojen teollinen käyttö ja zooplankton. - Ultraviolettisäteilyn vedenalai- suopohjan hyödyntäminen Suomessa. - nen tunkeutuminen ja sen vaikutukset leviin Industrial use of peatlands and the re-use of ja eläinplanktoniin. 52 p. (145 p.). Yhteenveto cut-away areas in Finland. 262 p. Foreword 3 2 p. 2000. p. Executive summary 9 p. 1999. 87 PÄÄKKÖNEN, JARI-PEKKA, Feeding biology of 80 LEPPÄNEN, HARRI, The fate of resin acids and burbot, Lota lota (L.): Adaptation to profundal resin acid-derived compounds in aquatic lifestyle? - Mateen, Lota lota (L), ravinnon- environment contaminated by chemical wood käytön erityispiirteet: sopeumia pohja- industry. - Hartsihappojen ja hartsihappope- elämään? 33 p. (79 p.). Yhteenveto 2 p. 2000. räisten yhdisteiden ympäristökohtalo kemial- 88 LAASONEN, PEKKA, The effects of stream habit lisen puunjalostusteollisuuden likaamissa restoration on benthic communities in boreal vesistöissä. 45 p. (149 p.). headwater streams. - Koskikunnostuksen Yhteenveto 2 p.1999. vaikutus jokien pohjaeläimistöön. 32 p. (101 81 LINDSTRÖM, LEENA, Evolution of conspicuous p.). Yhteenveto 2 p. 2000. warning signals. - Näkyvien varoitussignaa- 89 PASONEN, HANNA-LEENA, Pollen competition in lien evoluutio. 44 p. ( 96 p.). Yhteenveto 3 p. silver (Betula pendula Roth). An 2000. evolutionary perspective and implications for 82 MATTILA, ELISA, Factors limiting reproductive commercial seed production. - success in terrestrial orchids. - Kämmeköiden Siitepölykilpailu koivulla. 41 p. (115 p.). lisääntymismenestystä rajoittavat tekijät. 29 p. Yhteenveto 2 p. 2000. (95 p.). Yhteenveto 2 p. 2000. 90 SALMINEN, ESA, Anaerobic digestion of solid 83 KARELS, AARNO, Ecotoxicity of pulp and paper poultry slaughterhouse by-products and mill effluents in fish. Responses at biochemical, wastes. - Siipikarjateurastuksen sivutuottei- individual, population and community levels. den ja jätteiden anaerobinen käsittely. 60 p. - Sellu- ja paperiteollisuuden jätevesien (166 p.). Yhteenveto 2 p. 2000. ekotoksisuus kaloille. Tutkimus kalojen 91 SALO, HARRI, Effects of ultraviolet radiation on biokemiallisista, fysiologisista sekä the immune system of fish. - Ultravioletti- populaatio- ja yhteisövasteista. 68 p. (177 p.). säteilyn vaikutus kalan immunologiseen Yhteenveto 1 p. Samenvatting 1 p. 2000. puolustusjärjestelmään. 61 p. (109 p.). 84 AALTONEN, TUULA, Effects of pulp and paper Yhteenveto 2 p. 2000. mill effluents on fish immune defence. - Met- 92 MUSTAJÄRVI, KAISA, Genetic and ecological säteollisuuden jätevesien aiheuttamat consequences of small population size in immunologiset muutokset kaloissa. 62 p. (125 Lychnis viscaria. - Geneettisten ja ekologisten p.). 2000. tekijöiden vaikutus pienten mäkitervakko- 85 HELENIUS, MERJA, Aging-associated changes in populaatioiden elinkykyyn. 33 p. (124 p.). NF-kappa B signaling. - Ikääntymisen vaiku- Yhteenveto 3 p. 2000. tus NF-kappa B:n signalointiin. 75 p. (143 p.). Yhteenveto 2 p. 2000. JYVÄSKYLÄ STUDIES IN BIOLOGICAL AND ENVIRONMENTAL SCIENCE

93 TIKKA, PÄIVI, Threatened flora of semi-natural predators. - Hyönteisten väritys puolustukses- grasslands: preservation and restoration. - sa vihollisia vastaan. 44 p. (92 p.) Yhteenveto Niittykasvillisuuden säilyttäminen ja 3 p. 2001. ennallistaminen. 35 p. (105 p.). Yhteenveto 2 p. 103 NIKKILÄ, ANNA, Effects of organic material on 2001. the bioavailability, toxicokinetics and toxicity 94 SIITARI, HELI, Ultraviolet sensitivity in birds: of xenobiotics in freshwater organisms. - consequences on foraging and mate choice. - Orgaanisen aineksen vaikutus vierasaineiden Lintujen ultraviolettinäön ekologinen mer- biosaatavuuteen, toksikokinetiikkaan ja kitys ravinnon- ja puolisonvalinnassa. 31 p. toksisuuteen vesieliöillä. 49 p. (102 p.) (90 p.). Yhteenveto 2 p. 2001. Yhteenveto 3 p. 2001. 95 VERTAINEN, LAURA, Variation in life-history 104 LIIRI, MIRA, Complexity of soil faunal traits and behaviour among wolf spider communities in relation to ecosystem (Hygrolycosa rubrofasciata) populations. - functioning in coniferous forrest soil. A Populaatioiden väliset erot rummuttavan disturbance oriented study. - Maaperän hämähäkin Hygrolycosa rubrofasciata) kasvus- hajottajaeliöstön monimuotoisuuden merkitys sa ja käyttäytymisessä. 37 p. (117 p.) metsäekosysteemin toiminnassa ja häiriön- Yhteenveto 2 p. 2001. siedossa. 36 p. (121 p.) Yhteenveto 2 p. 2001. 96 HAAPALA, ANTTI, The importance of particulate 105 KOSKELA, TANJA, Potential for coevolution in a organic matter to invertebrate communities of host plant – holoparasitic plant interaction. - boreal woodland streams. Implications for Isäntäkasvin ja täysloiskasvin välinen vuoro- stream restoration. - Hiukkasmaisen orgaanisen vaikutus: edellytyksiä koevoluutiolle? 44 p. aineksen merkitys pohjoisten metsäjokien pohja- (122 p.) Yhteenveto 3 p. 2001. eläinyhteisöille - huomioita virtavesien 106 LAPPIVAARA, JARMO, Modifications of acute kunnostushankkeisiin. 35 p. (127 p.) Yhteenveto 2 physiological stress response in whitefish p. 2001. after prolonged exposures to water of 97 NISSINEN, LIISA, The collagen receptor integrins anthropogenically impaired quality. - - differential regulation of their expression and Ihmistoiminnan aiheuttaman veden laadun signaling functions. - Kollageeniin sitoutuvat heikentymisen vaikutukset planktonsiian integriinit - niiden toisistaan eroava säätely ja fysiologisessa stressivasteessa. 46 p. (108 p.) signalointi. 67 p. (125 p.) Yhteenveto 1 p. 2001. Yhteenveto 3 p. 2001. 98 AHLROTH, MERVI, The chicken avidin gene 107 ECCARD, JANA, Effects of competition and family. Organization, evolution and frequent seasonality on life history traits of bank voles. recombination. - Kanan avidiini-geeniperhe. - Kilpailun ja vuodenaikaisvaihtelun vaikutus Organisaatio, evoluutio ja tiheä metsämyyrän elinkiertopiirteisiin. rekombinaatio. 73 p. (120 p.) Yhteenveto 2 p. 29 p. (115 p.) Yhteenveto 2 p. 2002. 2001. 108 NIEMINEN, JOUNI, Modelling the functioning of 99 HYÖTYLÄINEN, TARJA, Assessment of experimental soil food webs. - Kokeellisten ecotoxicological effects of creosote- maaperäravintoverkkojen toiminnan contaminated lake sediment and its mallintaminen. 31 p. (111 p.) Yhteenveto remediation. - Kreosootilla saastuneen 2 p. 2002. järvisedimentin ekotoksikologisen riskin 109 NYKÄNEN, MARKO, Protein secretion in ja kunnostuksen arviointi. 59 p. (132 p.) Trichoderma reesei. Expression, secretion and Yhteenveto 2 p. 2001. maturation of cellobiohydrolase I, barley 100 SULKAVA, PEKKA, Interactions between faunal cysteine proteinase and calf chymosin in Rut- community and decomposition processes in C30. - Proteiinien erittyminen Trichoderma relation to microclimate and heterogeneity in reeseissä. Sellobiohydrolaasi I:n, ohran boreal forest soil. - Maaperän eliöyhteisön ja kysteiiniproteinaasin sekä vasikan hajotusprosessien väliset vuorovaiku-tukset kymosiinin ilmeneminen, erittyminen ja suhteessa mikroilmastoon ja laikut-taisuuteen. kypsyminen Rut-C30-mutanttikannassa. 107 36 p. (94 p.) Yhteenveto 2 p. 2001. p. (173 p.) Yhteenveto 2 p. 2002. 101 LAITINEN, OLLI, Engineering of 110 TIIROLA, MARJA, Phylogenetic analysis of physicochemical properties and quaternary bacterial diversity using ribosomal RNA structure assemblies of avidin and gene sequences. - Ribosomaalisen RNA- streptavidin, and characterization of avidin geenin sekvenssien käyttö bakteeridiver- related proteins. - Avidiinin ja streptavi-diinin siteetin fylogeneettisessä analyysissä. 75 p. kvaternäärirakenteen ja fysioke-miallisten (139 p.) Yhteenveto 2 p. 2002. ominaisuuksien muokkaus sekä avidiinin 111 HONKAVAARA, JOHANNA, Ultraviolet cues in fruit- kaltaisten proteiinien karakteri-sointi. 81 p. frugivore interactions. - Ultraviolettinäön (126 p.) Yhteenveto 2 p. 2001. ekologinen merkitys hedelmiä syövien eläin- 102 LYYTINEN, ANNE, Insect coloration as a defence ten ja hedelmäkasvien välisissä vuoro- mechanism against visually hunting vaikutussuhteissa. 27 p. (95 p.) Yhteenveto 2 p. 2002. 112 MARTTILA, ARI, Engineering of charge, biotin- binding and oligomerization of avidin: new JYVÄSKYLÄ STUDIES IN BIOLOGICAL AND ENVIRONMENTAL SCIENCE

tools for avidin-biotin technology. - Avidiinin vasta alueelliseen ja luonnonsuojelullinen varauksen, biotiininsitomisen sekä merkitys. 36 p. (121 p.) Yhteenveto 3 p. 2003. oligomerisaation muokkaus: uusia työkaluja 122 SUIKKANEN, SANNA, Cell biology of canine avidiini–biotiiniteknologiaan. 68 p. (130 p.) parvovirus entry. - Koiran parvovirusinfektion Yhteenveto 2 p. 2002. alkuvaiheiden solubiologia. 88 p. (135 p.) 113 JOKELA, JARI, Landfill operation and waste Yhteenveto 3 p. 2003. management procedures in the reduction of 123 AHTIAINEN, JARI JUHANI, Condition-dependence methane and leachate pollutant emissions of male sexual signalling in the drumming from municipal solid waste landfills. - Kaato- wolf spider Hygrolycosa rubrofasciata. - paikan operoinnin ja jätteen esikäsittelyn Koiraan seksuaalisen signaloinnin kunto- vaikutus yhdyskuntajätteen biohajoamiseen ja riippuvuus rummuttavalla susihämähäkillä typpipäästöjen hallintaan. 62 p. (173 p.) Hygrolycosa rubrofasciata. 31 p. (121 p.) Yhteen- Yhteenveto 3 p. 2002. veto 2 p. 2003. 114 RANTALA, MARKUS J., Immunocompetence and 124 KAPARAJU, PRASAD, Enhancing methane sexual selection in insects. - Immunokom- production in a farm-scale biogas production petenssi ja seksuaalivalinta hyönteisillä. 23 p. system. - Metaanintuoton tehostaminen (108 p.) Yhteenveto 1 p. 2002. tilakohtaisessa biokaasuntuotanto- 115 OKSANEN, TUULA, Cost of reproduction and järjestelmässä. 84 p. (224 p.) Yhteenveto 2 p. offspring quality in the evolution of 2003. reproductive effort. - Lisääntymisen kustan- 125 HÄKKINEN, JANI, Comparative sensitivity of nukset ja poikasten laatu lisääntymispanos- boreal fishes to UV-B and UV-induced tuksen evoluutiossa. 33 p. (95 p.) Yhteenveto 2 p. 2002. phototoxicity of retene. - Kalojen varhais- vaiheiden herkkyys UV-B säteilylle ja reteenin 116 HEINO, JANI, Spatial variation of benthic UV-valoindusoituvalle toksisuudelle. 58 p. macroinvertebrate biodiversity in boreal streams. Biogeographic context and (134 p.) Yhteenveto 2 p. 2003. conservation implications. - Pohjaeläinyh- 126 NORDLUND, HENRI, Avidin engineering; teisöjen monimuotoisuuden spatiaalinen modification of function, oligomerization, vaihtelu pohjoisissa virtavesissä - eliömaan- stability and structure topology. - Avidiinin tieteellinen yhteys sekä merkitys jokivesien toiminnan, oligomerisaation, kestävyyden ja suojelulle. 43 p. (169 p.) Yhteenveto 3 p. 2002. rakennetopologian muokkaaminen. 64 p. 117 SIIRA-PIETIKÄINEN, ANNE, Decomposer (104 p.) Yhteenveto 2 p. 2003. community in boreal coniferous forest soil 127 MARJOMÄKI, TIMO J., Recruitment variability in after forest harvesting: mechanisms behind vendace, Coregonus albula (L.), and its responses. - Pohjoisen havumetsämaan consequences for vendace harvesting. - hajottajayhteisö hakkuiden jälkeen: muutok- Muikun, Coregonus albula (L.), vuosiluokkien siin johtavat mekanismit. 46 p. (142 p.) Yh- runsauden vaihtelu ja sen vaikutukset kalas- teenveto 3 p. 2002. tukseen. 66 p. (155 p.) Yhteenveto 2 p. 2003. ORTET AINE 118 K , R , Parasitism, reproduction and 128 KILPIMAA, JANNE, Male ornamentation and sexual selection of roach, Rutilus rutilus L. - immune function in two species of passerines. Loisten ja taudinaiheuttajien merkitys kalan - Koiraan ornamentit ja immuunipuolustus lisääntymisessä ja seksuaalivalinnassa. 37 p. varpuslinnuilla. 34 p. (104 p.) Yhteenveto 1 p. (111 p.) Yhteenveto 2 p. 2003. 2004. 119 SUVILAMPI, JUHANI, Aerobic wastewater 129 PÖNNIÖ, TIIA, Analyzing the function of treatment under high and varying nuclear receptor Nor-1 in mice. - Hiiren temperatures – thermophilic process tumareseptori Nor-1:n toiminnan tutkiminen. performance and effluent quality. - Jätevesien 65 p. (119 p.) Yhteenveto 2 p. 2004. käsittely korkeissa ja vaihtelevissa lämpöti- 130 WANG, HONG, Function and structure, loissa. 59 p. (156 p.) Yhteenveto 2 p. 2003. subcellular localization and evolution of the 120 PÄIVINEN, JUSSI, Distribution, abundance and encoding gene of pentachlorophenol 4- species richness of butterflies and monooxygenase in sphingomonads. 56 p. myrmecophilous beetles. - Perhosten ja (90 p.) 2004. muurahaispesissä elävien kovakuoriaisten 131 YLÖNEN, OLLI, Effects of enhancing UV-B levinneisyys, runsaus ja lajistollinen moni- irradiance on the behaviour, survival and muotoisuus 44 p. (155 p.) Yhteenveto 2 p. metabolism of coregonid larvae. - Lisääntyvän 2003. UV-B säteilyn vaikutukset siikakalojen 121 PAAVOLA, RIKU, Community structure of poikasten käyttäytymiseen, kuolleisuuteen ja macroinvertebrates, bryophytes and fish in metaboliaan. 42 p. (95 p.) Yhteenveto 2 p. boreal streams. Patterns from local to regional 2004. scales, with conservation implications. - Selkärangattomien, vesisammalten ja kalojen yhteisörakenne pohjoisissa virtavesissä – säännönmukaisuudet paikallisesta mittakaa- JYVÄSKYLÄ STUDIES IN BIOLOGICAL AND ENVIRONMENTAL SCIENCE

132 KUMPULAINEN, TOMI, The evolution and 142 PYLKKÖ, PÄIVI, Atypical Aeromonas salmonicida maintenance of reproductive strategies in bag -infection as a threat to farming of arctic charr worm moths (Lepidoptera: Psychidae). (Salvelinus alpinus L.) and european grayling - Lisääntymisstrategioiden evoluutio ja säily- (Thymallus thymallus L.) and putative means to minen pussikehrääjillä (Lepidoptera: prevent the infection. - Epätyyppinen Aero- Psychidae). 42 p. (161 p.) Yhteenveto 3 p. monas salmonicida -bakteeritartunta uhkana 2004. harjukselle (Thymallus thymallus L.) ja nieriälle 133 OJALA, KIRSI, Development and applications of (Salvelinus alpinus L.) laitoskasvatuksessa ja baculoviral display techniques. - Bakulo- mahdollisia keinoja tartunnan ennalta- virus display -tekniikoiden kehittäminen ja ehkäisyyn. 46 p. (107 p.) Yhteenveto 2 p. 2004. sovellukset. 90 p. (141 p.) Yhteenveto 3 p. 143 PUURTINEN, MIKAEL, Evolution of hermaphro- 2004. ditic mating systems in animals. - Kaksi- neuvoisten lisääntymisstrategioiden evoluu- 134 RANTALAINEN, MINNA-LIISA, Sensitivity of soil tio eläimillä. 28 p. (110 p.) Yhteenveto 3 p. decomposer communities to habitat 2004. fragmentation – an experimental approach. - 144 TOLVANEN, OUTI, Effects of waste treatment Metsämaaperän hajottajayhteisön vasteet technique and quality of waste on bioaerosols elinympäristön pirstaloitumiseen. 38 p. in Finnish waste treatment plants. - Jätteen- (130 p.) Yhteenveto 2 p. 2004. käsittelytekniikan ja jätelaadun vaikutus 135 SAARINEN, MARI, Factors contributing to the bioaerosolipitoisuuksiin suomalaisilla jätteen- abundance of the ergasilid copepod, käsittelylaitoksilla. 78 p. (174 p.) Yhteenveto Paraergasilus rylovi, in its freshwater 4 p. 2004. 145 BOADI, KWASI OWUSU, Environment and health molluscan host, Anodonta piscinalis. - in the Accra metropolitan area, Ghana. - Paraergasilus rylovi -loisäyriäisen esiintymi- Accran (Ghana) suurkaupunkialueen ympä- seen ja runsauteen vaikuttavat tekijät ristö ja terveys. 33 p. (123 p.) Yhteenveto 2 p. Anodonta piscinalis -pikkujärvisimpukassa. 2004. 47 p. (133 p.) Yhteenveto 4 p. 2004. 146 LUKKARI, TUOMAS, Earthworm responses to 136 LILJA, JUHA, Assessment of fish migration in metal contamination: Tools for soil quality rivers by horizontal echo sounding: Problems assessment. - Lierojen vasteet concerning side-aspect target strength. metallialtistukseen: käyttömahdollisuudet - Jokeen vaeltavien kalojen laskeminen sivut- maaperän tilan arvioinnissa. 64 p. (150 p.) taissuuntaisella kaikuluotauksella: sivu- Yhteenveto 3 p. 2004. aspektikohdevoimakkuuteen liittyviä ongel- 147 MARTTINEN, SANNA, Potential of municipal sewage treatment plants to remove bis(2- mia. 40 p. (82 p.) Yhteenveto 2 p. 2004. ethylhexyl) phthalate. - Bis-(2-etyyli- 137 NYKVIST, PETRI, Integrins as cellular receptors heksyyli)ftalaatin poistaminen jätevesistä for fibril-forming and transmembrane yhdyskuntajätevedenpuhdistamoilla. 51 p. collagens. - Integriinit reseptoreina fibril- (100 p.) Yhteenveto 2 p. 2004. laarisille ja transmembraanisille kolla- 148 KARISOLA, PIIA, Immunological characteri- geeneille. 127 p. (161 p.) Yhteenveto 3 p. 2004. zation and engineering of the major latex 138 KOIVULA, NIINA, Temporal perspective of allergen, hevein (Hev b 6.02). - Luonnon- humification of organic matter. - Orgaanisen kumiallergian pääallergeenin, heveiinin aineen humuistuminen tarkasteltuna ajan (Hev b 6.02), immunologisten ominaisuuksien funktiona. 62 p. (164 p.) Yhteenveto 2 p. 2004. karakterisointi ja muokkaus. 91 p. (113 p.) 139 KARVONEN, ANSSI, Transmission of Diplostomum Yhteenveto 2 p. 2004. spathaceum between intermediate hosts. 149 BAGGE, ANNA MARIA, Factors affecting the - Diplostomum spathaceum -loisen siirtyminen development and structure of monogenean communities on cyprinid fish. - Kidus- kotilo- ja kalaisännän välillä. 40 p. (90 p.) loisyhteisöjen rakenteeseen ja kehitykseen Yhteenveto 2 p. 2004. vaikuttavat tekijät sisävesikaloilla. 25 p. 140 NYKÄNEN, MARI, Habitat selection by riverine (76 p.) Yhteenveto 1 p. 2005. grayling, Thymallus thymallus L. - Harjuksen 150 JÄNTTI, ARI, Effects of interspecific relation- (Thymallus thymallus L.) habitaatinvalinta ships in forested landscapes on breeding virtavesissä. 40 p. (102 p.) Yhteenveto 3 p. 2004. success in Eurasian treecreeper. - Lajien- 141 HYNYNEN, JUHANI, Anthropogenic changes in välisten suhteiden vaikutus puukiipijän Finnish lakes during the past 150 years pesintämenestykseen metsäympäristössä. inferred from benthic invertebrates and their 39 p. (104 p.) Yhteenveto 2 p. 2005. sedimentary remains. - Ihmistoiminnan 151 TYNKKYNEN, KATJA, Interspecific interactions aiheuttamat kuormitusmuutokset suomalaisis- and selection on secondary sexual characters sa järvissä viimeksi kuluneiden 150 vuoden in damselflies. - Lajienväliset vuorovaikutuk- set ja seksuaaliominaisuuksiin kohdistuva aikana tarkasteltuina pohjaeläinyhteisöjen valinta sudenkorennoilla. 26 p. (86 p.) Yh- avulla. 45 p. (221 p.) Yhteenveto 3 p. 2004. teenveto 2 p. 2005. JYVÄSKYLÄ STUDIES IN BIOLOGICAL AND ENVIRONMENTAL SCIENCE

152 HAKALAHTI, TEIJA, Studies of the life history of a 164 ILMARINEN, KATJA, Defoliation and plant–soil parasite: a basis for effective population interactions in grasslands. - Defoliaatio ja management. - Loisen elinkiertopiirteet: kasvien ja maaperän väliset vuorovaikutukset perusta tehokkaalle torjunnalle. 41 p. (90 p.) niittyekosysteemeissä. 32 p. (111 p.) Yhteenve- Yhteenveto 3 p. 2005. to 2 p. 2006. 153 HYTÖNEN, VESA, The avidin protein family: 165 LOEHR, JOHN, Thinhorn sheep evolution and properties of family members and engineering behaviour. - Ohutsarvilampaiden evoluutio ja of novel biotin-binding protein tools. - Avidiini- käyttäytyminen. 27 p. (89 p.) Yhteenveto 2 p. proteiiniperhe: perheen jäsenten ominaisuuk- 2006. sia ja uusia biotiinia sitovia proteiiniyökaluja. 166 PAUKKU, SATU, Cost of reproduction in a seed 94 p. (124 p.) Yhteenveto 2 p. 2005. beetle: a quantitative genetic perspective. - 154 GILBERT, LEONA , Development of biotechnological Lisääntymisen kustannukset jyväkuoriaisella: tools for studying infectious pathways of kvantitatiivisen genetiikan näkökulma. 27 p. canine and human parvoviruses. 104 p. (84 p.) Yhteenveto 1 p. 2006. (156 p.) 2005. 167 OJALA, KATJA, Variation in defence and its 155 SUOMALAINEN, LOTTA-RIINA, Flavobacterium columnare in Finnish fish farming: fitness consequences in aposematic animals: characterisation and putative disease interactions among diet, parasites and management strategies. - Flavobacterium predators. - Puolustuskyvyn vaihtelu ja sen columnare Suomen kalanviljelyssä: merkitys aposemaattisten eläinten kelpoisuu- karakterisointi ja mahdolliset torjunta- teen: ravinnon, loisten ja saalistajien vuoro- menetelmät. 52 p. (110 p.) Yhteenveto 1 p. vaikutus. 39 p. (121 p.) Yhteenveto 2 p. 2006. 2005. 168 MATILAINEN, HELI, Development of baculovirus 156 VEHNIÄINEN, EEVA-RIIKKA, Boreal fishes and display strategies towards targeting to tumor ultraviolet radiation: actions of UVR at vasculature. - Syövän suonitukseen molecular and individual levels. - Pohjoisen kohdentuvien bakulovirus display-vektorien kalat ja ultraviolettisäteily: UV-säteilyn kehittäminen. 115 p. (167 p.) Yhteenveto 2 p. vaikutukset molekyyli- ja yksilötasolla. 52 p. 2006. (131 p.) 2005. 169 KALLIO, EVA R., Experimental ecology on the 157 VAINIKKA, ANSSI, Mechanisms of honest sexual interaction between the Puumala hantavirus signalling and life history trade-offs in three and its host, the bank vole. - Kokeellista cyprinid fishes. - Rehellisen seksuaalisen ekologiaa Puumala-viruksen ja metsämyyrän signaloinnin ja elinkiertojen evoluution välisestä vuorovaikutussuhteesta. 30 p. (75 p.) mekanismit kolmella särkikalalla. 53 p. Yhteenveto 2 p. 2006. (123 p.) Yhteenveto 2 p. 2005. 170 PIHLAJA, MARJO, Maternal effects in the magpie. 158 LUOSTARINEN, SARI, Anaerobic on-site - Harakan äitivaikutukset. 39 p. (126p.) wastewater treatment at low temperatures. Yhteenveto 1 p. 2006. Jätevesien kiinteistö- ja kyläkohtainen 171 IHALAINEN, EIRA, Experiments on defensive anaerobinen käsittely alhaisissa lämpötilois- mimicry: linkages between predator behaviour sa. 83 p. (168 p.) Yhteenveto 3 p. 2005. and qualities of the prey. - Varoitussignaalien 159 SEPPÄLÄ, OTTO, Host manipulation by jäljittely puolustusstrategiana: kokeita peto– parasites: adaptation to enhance saalis-suhteista. 37 p. (111 p.) Yhteenveto 2 p. transmission? Loisten kyky manipuloida 2006. isäntiään: sopeuma transmission tehostami- 172 LÓPEZ-SEPULCRE, ANDRÉS, The evolutionary seen? 27 p. (67 p.) Yhteenveto 2 p. 2005. ecology of space use and its conservation 160 SUURINIEMI, MIIA, Genetics of children’s consequences. - Elintilan käytön ja reviiri- bone growth. - Lasten luuston kasvun gene- käyttäytymisen evoluutioekologia tiikka. 74 p. (135 p.) Yhteenveto 3 p. 2006. luonnonsuojelullisine seuraamuksineen. 32 p. 161 TOIVOLA, JOUNI, Characterization of viral (119 p.) Yhteenveto 2 p. 2007. nanoparticles and virus-like structures by 173 TULLA, MIRA, Collagen receptor integrins: using fluorescence correlation spectroscopy evolution, ligand binding selectivity and the (FCS) . - Virus-nanopartikkelien sekä virusten effect of activation. - Kollageenireseptori- kaltaisten rakenteiden tarkastelu fluoresenssi integriiniien evoluutio, ligandin sitomis- korrelaatio spektroskopialla. 74 p. (132 p.) valikoivuus ja aktivaation vaikutus. 67 p. (129 Yhteenveto 2 p. 2006. p.) Yhteenveto 2 p. 2007. 162 KLEMME, INES, Polyandry and its effect on male 174 SINISALO, TUULA, Diet and foraging of ringed and female fitness. - Polyandria ja sen vaiku- tukset koiraan ja naaraan kelpoisuuteen 28 p. seals in relation to helminth parasite (92 p.) Yhteenveto 2 p. 2006. assemblages. - Perämeren ja Saimaan norpan 163 LEHTOMÄKI, ANNIMARI, Biogas production from suolistoloisyhteisöt ja niiden hyödyntäminen energy crops and crop residues. - Energia- hylkeen yksilöllisen ravintoekologian selvittä- kasvien ja kasvijätteiden hyödyntäminen misessä. 38 p. (84 p.) Yhteenveto 2 p. 2007. biokaasun tuotannossa. 91 p. (186 p.) Yhteen- veto 3 p. 2006. JYVÄSKYLÄ STUDIES IN BIOLOGICAL AND ENVIRONMENTAL SCIENCE

175 TOIVANEN, TERO, Short-term effects of forest restoration on beetle diversity. - Metsien ennallistamisen merkitys kovakuoriaislajiston monimuotoisuudelle. 33 p. (112 p.) Yhteenveto 2 p. 2007. 176 LUDWIG, GILBERT, Mechanisms of population declines in boreal forest grouse. - Kanalintu- kantojen laskuun vaikuttavat tekijät. 48 p. (138 p.) Yhteenveto 2 p. 2007. 177 KETOLA, TARMO, Genetics of condition and sexual selection. - Kunnon ja seksuaalivalin- nan genetiikka. 29 p. (121 p.) Yhteenveto 2 p. 2007. 178 SEPPÄNEN, JANNE-TUOMAS, Interspecific social information in habitat choice. - Lajienvälinen sosiaalinen informaatio habitaatinvalin- nassa. 33 p. (89 p.) Yhteenveto 2 p. 2007. 179 BANDILLA, MATTHIAS, Transmission and host and mate location in the fish louse Argulus coregoni and its link with bacterial disease in fish. - Argulus coregoni -kalatäin siirtyminen kalaisäntään, isännän ja parittelukumppanin paikallistaminen sekä loisinnan yhteys kalan bakteeritautiin. 40 p. (100 p.) Yhteenveto 3 p. Zusammenfassung 4 p. 2007. 180 MERILÄINEN, PÄIVI, Exposure assessment of animals to sediments contaminated by pulp and paper mills. - Sellu- ja paperiteollisuuden saastuttamat sedimentit altistavana tekijänä vesieläimille. 79 p. (169 p.) Yhteenveto 2 p. 2007. 181 ROUTTU, JARKKO, Genetic and phenotypic divergence in Drosophila virilis and D. montana. - Geneettinen ja fenotyyppinen erilaistuminen Drosophila virilis ja D. montana lajien mahlakärpäsillä. 34 p. (106 p.) Yhteen- veto 1 p. 2007.