Postmating Sexual Selection and its Role in Population Divergence in Beetles CLAUDIA FRICKE

Full text

(1)Digital Comprehensive Summaries of Uppsala Dissertations from the Faculty of Science and Technology 152. Postmating Sexual Selection and its Role in Population Divergence in Beetles CLAUDIA FRICKE. ACTA UNIVERSITATIS UPSALIENSIS UPPSALA 2006. ISSN 1651-6214 ISBN 91-554-6482-3 urn:nbn:se:uu:diva-6583.

(2)

(3)

(4)

(5)

(6)

(7)

(8)

(9)

(10) ! " # $ % $ $ &% %' (%

(11)

(12) )

(13)

(14)

(15) %' * ' !' &

(16) + +

(17)

(18) ,

(19) &

(20)

(21)

(22) - '

(23)

(24) '

(25)

(26)

(27)

(28) ".' /. ' ' 0+-1 2"3../3!/435' +

(29) %

(30) % $

(31) $

(32) )

(33) % % $

(34) '

(35) %

(36)

(37) $ ' (% %

(38) $

(39)

(40) %.

(41)

(42)

(43)

(44) ' 1

(45)

(46) ) 6

(47)

(48) %

(49)

(50) %

(51)

(52)

(53) ' 0

(54)

(55) %

(56) 7 %

(57) ' (%$

(58)

(59)

(60)

(61)

(62)

(63)

(64)

(65)

(66) )

(67)

(68) $

(69) % $ 7 $

(70) ' 8 3

(71)

(72) $ % )

(73) $ $

(74)

(75) %

(76) * % ' (%

(77)

(78) % % $

(79)

(80)

(81)

(82) $ % $

(83)

(84)

(85)

(86)

(87)

(88)

(89) ' (% 6 $ ) $

(90) -%

(91)

(92) ) )

(93)

(94) ' 0 ) 6 % %' 9

(95) % *

(96) $ % $

(97)

(98)

(99)

(100) $

(101)

(102) %

(103) $ $

(104)

(105) ) % $

(106)

(107) ' +

(108) 0

(109)

(110)

(111)

(112) % 6

(113) $$ $

(114)

(115)

(116)

(117)

(118)

(119)

(120)

(121)

(122) ' 0

(123)

(124) %

(125) %

(126)

(127)

(128) )$

(129)

(130) $

(131)

(132)

(133) )

(134)

(135) ' %

(136)

(137)

(138)

(139) % $ $

(140)

(141)

(142)

(143)

(144) $$

(145) $ ) )

(146)

(147) $

(148) ' :%

(149)

(150)

(151)

(152) 0 $

(153) %.

(154)

(155) ) %

(156)

(157)

(158)

(159)

(160) % ' +

(161)

(162)

(163) $

(164)

(165)

(166)

(167)

(168)

(169) ' 0

(170)

(171)

(172)

(173) $

(174)

(175)

(176) ) *

(177)

(178) ' (% % 6

(179) $$ $

(180)

(181)

(182)

(183)

(184)

(185)

(186) 3 .

(187) %

(188)

(189)

(190) $ $

(191) ' %

(192)

(193)

(194)

(195)

(196)

(197) . !"

(198) # # " $ %& " " #'()*+, " ; * ! 0++1 "!."3!"/ 0+-1 2"3../3!/435

(199) #

(200)

(201) ### 3!.45 <% #==

(202) '*'= >

(203) ?

(204) #

(205)

(206) ### 3!.45@.

(207) To my family, whom I love deeply.

(208) The beetle – drawings on the cover were designed by Dr. Gerhild Kaselow..

(209) List of Papers. This thesis is based on the following papers, which are referred to in the text by their Roman numerals.. I.. Fricke, C., Arnqvist, G. & Amaro, N. (2006) Female modulation of reproductive rate and its role in postmating prezygotic isolation in Callosobruchus maculatus. Functional Ecology (In press).. II.. Fricke, C. & Arnqvist, G. (2004) Conspecific sperm precedence in flour beetles. Animal Behaviour, 67: 729-732.. III.. Fricke, C. & Arnqvist, G. (2004) Divergence in replicated phylogenies: the evolution of partial post-mating prezygotic isolation in bean weevils. Journal of Evolutionary Biology, 17: 1345-1354.. VI.. Fricke, C. & Arnqvist, G. Rapid adaptation to a novel host in bean weevils: the role of sexual selection (Manuscript).. V. Fricke, C., Slevinsky, C. & Arnqvist, G. Natural selection hampers divergent evolution of reproductive traits in a seed beetle (Manuscript).. Papers I, II and III were reproduced with kind permission of the publishers: Elsevier Ltd and Blackwell Publishing Ltd. Contributions I performed all work on which this thesis is based, with the following exceptions. G. Arnqvist contributed towards the planning and analyses of all experiments as well as the final stages of writing papers I – V. Noelia Amaro participated in conducting the experiment on which paper II is based. Candice Slevinsky performed the P2 assay reported in paper V..

(210)

(211) Contents. Introduction...................................................................................................11 Divergence ...............................................................................................11 Natural selection ..................................................................................12 Sexual selection ...................................................................................12 Divergence patterns .............................................................................14 Postmating prezygotic isolation...........................................................16 Aims of the thesis..........................................................................................17 Model species and Methods..........................................................................18 Results and Discussion .................................................................................20 Conclusions...................................................................................................30 Summary in Swedish (Sammanfattning) ......................................................31 Summary in German (Zusammenfassung) ...................................................34 Acknowledgments.........................................................................................37 References.....................................................................................................40.

(212)

(213) ‘Ich bin selber Wissenschaftler und gewiss nicht wissenschaftsfeindlich eingestellt, aber ich habe mir auch meine mythische und ein wenig animistische Weltanschauung nie nehmen lassen. Ich habe mir durch Newton und Darwin niemals das eigentliche Mysterium des Lebens rauben lassen.’ Aus ‘Das Orangenmädchen’ Jostein Gaarder.

(214)

(215) Introduction. The amazing biodiversity we observe today is the result of millions of years of evolution leading to divergence and ultimately to speciation. Speciation, the process that causes the formation of new species, is accomplished when reproductive isolation prevents the exchange of genetic material between diverged forms (Dobzhansky 1937; Mayr 1970). Traditionally, we recognize two types of barriers to gene flow: premating and postzygotic reproductive isolation. Premating isolation occurs when mating between forms is less likely than within forms, such as occurs when different cues for mate recognition are involved in the acceptance or rejection of mates by the two forms. Postzygotic isolation is the result of genetic incompatibilities in hybrids, resulting in hybrid infertility or inviability (Coyne and Orr 2004). More recently, attention has been given to a third form of reproductive isolation: postmating prezygotic, or cryptic isolation (see Howard 1999; Eady 2001, for reviews). This latter form is less well investigated and reduces gene flow after a mating has occurred but prior to zygote formation. The evolution of reproductive isolation is central to understanding divergence and speciation. Hence, if we want to understand speciation we need to know, what evolutionary processes cause divergence and aid in the evolution of reproductive isolation.. Divergence Every speciation event starts with divergence between two or more populations. If a population splits into two or several subunits, each sub-population typically proceeds along a different evolutionary trajectory through differing selection pressures or random effects. As two populations gradually accumulate genetic dissimilarities, they become less likely to be reproductively compatible. Thus, it is expected that populations with decreasing degrees of relatedness should exhibit increased reproductive incompatibility. However, different types of selection are thought to result in different divergence patterns (see below). Primarily, there are two main types of selection, that cause populations to diverge; natural and sexual selection.. 11.

(216) Natural selection Darwin (1859) argued that natural selection is the driving selective agent for variation between populations. Populations exposed to a changing or novel environment adapt to the prevailing conditions and reproductive isolation can evolve as a by-product (Dobzhansky 1937; Mayr 1970). Different ecological settings cause contrasting selection pressures in populations inhabiting two or more environments, causing eventual reproductive isolation (Schluter 2000, 2001; Via 2001; Rundle and Nosil 2005). There is evidence from laboratory studies that premating isolation can evolve quite rapidly in populations experiencing divergent natural selection (reviewed in Rice and Hostert 1993 and Coyne and Orr 2004). A link between reproductive isolation and adaptive change is often established, and this forms the basis of assortative mating (Schluter 2000, 2001; Via 2001; Rundle and Nosil 2005). Differences in the biological or physical properties of different habitats can translate into changes in male signals/ornamentation accompanied by changes in female preferences. Potential examples of such differences in selective environment are differences in predator communities (e.g., Endler 1995; Nosil 2004), light environment (e.g., Marchetti 1993) and sound transmission (e.g., Slabbekoorn and Smith 2002). Here, changes in sexual selection with environment are essentially predictable/ repeatable and can result in parallel speciation (Nosil et al. 2002; Boughman et al. 2005). Thus, it is widely accepted that natural selection is an important generator of divergence.. Sexual selection The importance of sexual selection in causing variation between populations was widely acknowledged two decades ago (West-Eberhard 1983; Panhuis et al. 2001; Coyne and Orr 2004). Female mate choice has the potential to drive divergence of secondary sexual characters contributing to rapid reproductive isolation (Lande 1981, 1982). Sexual selection is considered to be a powerful selective agent as the actual traits involved in premating isolation (sexual traits) get modified directly, with reproductive isolation being a likely outcome. Evidence comes from comparative studies showing that more speciose genera contain on average more sexually dimorphic species (Barraclough et al. 1995; Mitra et al. 1996; Møller and Cuervo 1998; Owens et al. 1999; Arnqvist et al. 2000, but see Gage et al. 2002 and Morrow et al. 2003a). Furthermore, sexual selection should be arbitrary to some extent (Lande 1981; Arak and Enquist 1993; Schluter and Price 1993; Rice 1998) as there is a multitude of potential evolutionary trajectories along which male signals and female responses can coevolve. For internally fertilised animals, there are a great number of steps involved in each fertilization event (Eberhard 1996; Markow 1997; Bloch Qazi et al. 2003) with a large number of male ejaculatory substances interacting with female receptors (Gillot 12.

(217) 1996; Wolfner 1997, 2002; Chapman 2001). While the potential for sexual selection to generate divergence was recognised quite recently, do we now have compelling evidence from nature of the amazing power of this selective agent in causing phenomena as the radiation in African cichlids (Seehausen and van Alphen 1999). Sexually antagonistic coevolution With the advent of molecular tools it became clear that most animal species are polygamous. This insight changed the view evolutionary biologists had on reproduction from one of a harmonious event between the sexes to one that comprises an act incurring different costs and benefits to each sex. Therefore, polyandry leads to sexual conflict as the evolutionary interests of the sexes differ (Parker 1979). Bateman (1948) stated that male reproductive success depends critically on the number of females he mates with, while females often can secure lifelong fecundity with a single or few copulations. Thus, the sexes differ in their optimal mating rate, giving rise to the potential for male manipulation and female counter-adaptation over the number of copulations (Rowe et al. 1994; Arnqvist and Rowe 1995). Moreover, sexual conflict can arise over various reproductive decisions in the sexes (Arnqvist and Rowe 2005). Theory predicts that the form of direct sexual selection that results from interlocus sexual conflict should be particularly efficient in generating population divergence in reproductive characters (Rice 1998; Gavrilets 2000; Gavrilets et al. 2001; Arnqvist and Rowe 2005). Here, the harm that males cause to females arises as a pleiotropic by-product of male-male competition over fertilisations. The response of females to male traits then evolves (in essence a form of resistance) to mitigate male harm (Holland and Rice 1998; Rice 1998; Coyne and Orr 2004). This is thought to commonly result in perpetual sexual antagonistic coevolution between the sexes (Chapman et al. 1995; Holland and Rice 1999; Morrow et al. 2003b; Arnqvist and Rowe 2005). Because internal fertilization is associated with many potential forms of sexual conflict and because various traits contribute to male persistence and female resistance (Arnqvist and Rowe 2005), allopatric populations can in theory diverge along different arbitrary coevolutionary trajectories. Evidence for this scenario comes from comparative studies (Arnqvist 1998; Arnqvist et al. 2000; Swanson and Vacquier 2002 a,b). Martin and Hosken (2003) also provided some direct evidence in line with these ideas when they showed that laboratory populations evolving under more intense sexual conflict discriminated to a higher extent against mates from other selection lines.. Postmating sexual selection This form of sexual selection describes events that determine reproductive success after mating has taken place. Parker (1970) acknowledged that competition between males does not end with a successful copulation, but continues 13.

(218) with sperm competition when females mate multiply. Thus male-male competition does not end until the successful fertilisation of a females ovum. Females, however, are not passive bystanders, as long believed, but can actively bias fertilisation of their eggs towards certain preferred male types (Eberhard 1996; Wilson et al. 1997). This causes male and female genotypes to interact in their effect on patterns of fertilisation (Clark et al. 1999; Brown and Eady 2001; Hosken et al. 2002; Nilsson et al. 2002; 2003). Postmating sexual selection has gained more attention over the last years, suggesting that it is a powerful selective agent in shaping reproductive traits such as physiology, behaviour and genital morphology (Eberhard 1996; Arnqvist 1998). In line with predictions from theory is the finding that reproductive proteins evolve on average faster than non-reproductive proteins (Civetta and Singh 1995, 1998; Swanson et al. 2001; Galindo et al. 2003; for a review see: Swanson and Vacquier 2002 a,b). Hence, postmating sexual selection has the potential to cause reproductive incompatibility between diverging forms and is thus a prime candidate as an engine driving divergence in the early stages of speciation.. Divergence patterns The types of selection described above can all lead to divergence between populations. They should however leave different traces as different patterns of divergence are expected to emerge, reflected in the relationship between phylogeny and the evolution of reproductive incompatibilities (Arnqvist and Rowe 2005). One important aspect of divergence by natural selection is that it should in some sense be repeatable, such that independent episodes of adaptation to a given ecological environment yield similar results. Thus allopatric populations experiencing the same environment often exhibit parallel evolution (Rundle et al. 2000; Johannesson 2001; Nosil et al. 2002; Boughman et al. 2005). Sexual selection in contrast should be arbitrary (see section on sexual selection) as there is a multitude of potential evolutionary trajectories on which male signals and female responses can coevolve independent of the environment. The difference between natural and sexual selection described here can only be observed if we can compare replicated populations experiencing similar conditions. However, this might be rare in nature and further it does not allow us to distinguish between different sexual selection scenarios. An alternative approach is to compare rates of divergence with patterns revealed by comparing reproductive responses of crosses between several population of increasing phylogenetic distance. Reproductive incompatibility will eventually result from the accumulation of genetic changes rendering male signals and female receptors a poor fit. Evolution of reproductive divergence is slow if a character causing isolation changes as a correlated response to direct natural selection on a pleiotropically linked trait (Kirkpatrick and Ravigné 2002; see curve A in Figure 1). 14.

(219) Reproductive response. A. C. B. Male relatedness/Time since divergence Figure 1. The predicted relationship between female reproductive response and male relatedness in population crosses for different types of selection: natural selection (A); sexual selection by female choice (B) and sexually antagonistic coevolution (C). (Adapted after Arnqvist and Rowe 2005). In contrast, sexual selection is expected to evolve rapidly as a result of direct selection on reproductive characters (Lande 1982). Furthermore, different sexual selection scenarios are hypothesised to produce distinct patterns of reproductive response between diverging forms differing in their degrees of relatedness (Parker and Partridge 1998; Clark et al. 1999; Andrés and Arnqvist 2001; Arnqvist and Rowe 2005). Models of sexual selection via female choice for indirect benefits predict a tight match for the coevolution of male signals and female preferences. This results in a rapid but steady evolution of reproductive incompatibility between diverging forms (Arnqvist and Rowe 2005, see curve B in Figure 1). Under sexually antagonistic coevolution, reproductive compatibility will also decrease rapid and steadily, however, theory predicts a ‘window’ where females might respond more strongly than average to signals from males in slightly diverged populations (Parker and Partridge 1998). Females evolve resistance to reproductive signals from their own males, while they do not share an evolutionary history with allopatric males. Hence, closely related allopatric males might be able to elicit stronger reproductive responses in females than males from the same population (see curve C in Figure 1).. 15.

(220) Postmating prezygotic isolation Even though incompatibilities between gametes of different taxa have been known for some time (Dobzhansky 1937), scientists have only recently realised that isolation can be caused by traits acting after mating but before zygote formation. With the identification of the traits involved in postmating sexual selection it became clear that these same traits might play a vital role in reducing gene flow. To date conspecific sperm or pollen precedence is a widely acknowledged mechanism of cryptic isolation and has been wellstudied in a number of taxa (Nakano 1985; Hewitt et al. 1989; Robinson et al. 1994; Wade et al. 1994; Carney et al. 1996; Price 1997; Howard et al. 1998). Conspecific sperm precedence can only be observed under competition in multiply mating females. Conspecific sperm precedence describes the phenomenon that females preferably use sperm from their own conspecific males to fertilise their eggs over heterospecific sperm, thus effectively preventing the formation of hybrid offspring. Even though this phenomenon has received much attention we still have a poor understanding of how it evolves. Reproductive proteins are prime candidates in causing differential fertilisation and therefore are potentially important during speciation (Markow 1997; Rice 1998) as they are known to evolve rapidly (Civetta and Singh 1995, 1998; Swanson et al. 2001; Galindo et al. 2003; for a review see: Swanson and Vacquier 2002 a,b) and affect postcopulatory behaviour in females (Chen 1984; Wolfner 1997, 2002; Chapman 2001; Gillot 2003). These postmating reproductive ‘signals’ can be species specific (Chen 1984) and are known to sometimes cause assortative fertilisation (Palumbi 1999). There is a great variety of male seminal ‘signals’ (e.g. >80 in Drosophila; Chapman 2001), which influence different stages of the fertilisation process (Wolfner 1997, 2002; Bloch Qazi et al. 2003). Thus male-female coevolution could potentially occur along a near infinite number of trajectories, and cryptic barriers to fertilisation can involve a combination of several separate mechanisms. This means that it can be unique to hybridisation between any one species-pair (Albuquerque et al. 1996; Markow 1997; Alipaz et al. 2001; Eady 2001; Price et al. 2001).. 16.

(221) Aims of the thesis. This thesis addresses several topics in an attempt to gain a better understanding of the processes that cause early population divergence. (1) What is the prevailing selective agent causing divergence, or more, specifically, is postmating sexual selection important? (2) What types of selection influence the rate of adaptation? (3) What is the joint effect of natural and sexual selection concurrently acting on a population? These questions were addressed empirically by performing laboratory experiments on different species of beetles. The methods used and the results obtained will be presented in the following sections.. 17.

(222) Model species and Methods. In all experiments presented in this thesis, the bean weevil Callosobruchus maculatus was used, with the exception of paper II, where the experiment was conducted with flour beetles Tribolium spp. Both species are worldwide pests, with Tribolium living on flour/grains and Callosobruchus feeding on leguminose seeds in bean storages. Despite being pests, both species are beneficial for the study of evolutionary processes, as their life-cycle is fast and they are easily reared in the laboratory. In the following account I will concentrate on C. maculatus, as it is the main organism used in these studies. Mated female bean weevils cement their eggs on the host bean, and newly-hatched larvae then burrow into the seed. The larvae complete their development and pupate inside a single host seed. Adult beetles live for an average of 10 days when kept without food and water. The entire life-cycle from egg to egg is completed in about 21-24 days at 30°C. These biological features facilitate rearing of these bean weevils in the laboratory, making them suitable model organisms for long-term experiments. Colonies of C. maculatus were held in incubators under constant conditions at 30°C ±0.5° and 45% RH ±10% with a 12-12 hour light-dark cycle. I employed two main experimental methods to study the scientific questions addressed in this thesis. In all five studies presented here, I used allopatric populations, which originated from distinct geographical locations prior to transfer to laboratory conditions. These populations are expected to have taken separate evolutionary paths and the studies employed investigate the degree of divergence and the traits under selection. In the first two contributions of this thesis (paper I and II) a phylogenetic approach was taken, where a focal female strain was chosen and each was mated to males with different degrees of relatedness. Here the effect of male postmating prezygotic signals on reproductive responses in focal females was measured. I then tested whether a pattern between increasing phylogenetic distance and reproductive compatibility was discernable. Employing this logic, we investigated if a female’s early reproductive rate was modulated by copulations with certain male types (paper I), while in paper II we focused on females’ sperm utilization patterns. The second main methodological approach in this thesis employed the experimental manipulation of selection in laboratory cultures of Callosobruchus maculatus. In a long-term study we exposed replicated selection lines to different selection regimes by changing the intensity of natural and 18.

(223) sexual selection. We performed a laboratory natural selection experiment (sensu Fuller et al. 2005), by imposing a host bean and a mating system treatment on 16 selection lines. Half of the selection lines were retained on their ancestral laboratory host (black-eyed beans, Vigna unguiculata), while we experimentally induced a host shift by introducing the remaining lines to a novel host (chick peas, Cicer arietinum). In addition, we manipulated the intensity of postmating sexual selection in the selection lines by enforcing monogamy (M lines) on half of the selection lines whereas the other eight lines were kept polygamous (P lines). These two selection treatments were applied in a full factorial, orthogonal design with four replicates for each of the four treatment combinations. This experimental approach made it possible to investigate the joint action of natural and sexual selection on the rate of adaptation and population divergence (paper IV and V).. 19.

(224) Results and Discussion. Postmating prezygotic isolation In internally fertilised organisms, a multitude of processes are involved in the successful fusion of two gametes (Eberhard 1996; Markow 1997; Bloch Qazi et al. 2003). Thus, every step can potentially contribute to reproductive incompatibilities reducing gene flow between diverging forms. We investigated this hypothesis in papers I and II by testing if the populations used had diverged in male reproductive signals that are transferred in the ejaculate, and whether differences in females’ response to these signals had evolved. We employed the phylogenetic approach described above in both studies (see paper I and II for details). In paper I we tested if C. maculatus females modulate their reproductive rate according to male relatedness. Although females showed differences in early reproductive output depending on male type, they did not respond less to males with decreasing phylogenetic relatedness. In fact, the opposite was true. Females consistently showed the highest early reproductive response to signals from heterospecific males (F2 = 297.10, df = 3, P = <0.001) (see Figure 2). 40. Fecundity (Day 1). 35. 30. 25. 20 B ra z il fe m a le s Y e m e n fe m a le s M a li fe m a le s 15 1. 2. 3. 4. M a le s. Figure 2. Fecundity early in females’ life, measured here as the number of eggs laid by a female during 24 hours after her first mating. Males are coded according to their phylogenetic distance relative to females, with 1 representing males from the own strain through 4 representing heterospecific C.analis males.. 20.

(225) That male type had an effect on female reproductive rate implies a potential for divergence in these reproductive signal – receptor systems that could ultimately cause reproductive incompatibility. However, the results from paper I failed to find any mismatch between the heterospecific male signal and female receptors reflected in early reproductive response. In paper II we document that, in the red flour beetle Tribolium castaneum, males of different genotypes differently affected female sperm utilisation patterns (male sperm defense P1: F2 = 15.25, df = 6, P = 0.018; male sperm offense: P2: F2 = 27.68, df = 6, P = 0.0001). Females discriminated against sperm from unrelated males and this effect intensified with increasing phylogenetic distance (sperm defence P1: rs = -0.50, N = 20, P < 0.05; sperm offence (P2: rs = -0.14, N = 20, P > 0.5) (see Figure 3). This pattern shows that in this species postmating prezygotic reproductive incompatibility seems to accumulate steadily. Female discrimination against unrelated sperm evolves rapidly and in the early stages of population differentiation and is an effective mechanism in reducing gene flow between the diverging forms. 1.0. Sperm precedence. 0.8. 0.6. 0.4. 0.2. 0.0. G. C. A. N. T. F. M. Phylogenetic distance. Figure 3. Average (rSE) sperm precedence in focal females (T. castaneum [G]) measured as either P1 (filled circles) or P2 (open circles), in relation to the phylogenetic distance of their mate. G-T represent five different conspecific populations of T. castaneum, while F represents T. freemani and M represents T. madens. Note that position along the abscissa indicates increased relative rather than absolute phylogenetic distance between T. castaneum [G] and the male type. (P1 is the proportion of offspring sired by the focal male mated first to a female in a double mating experiment. P2 is the proportion of offspring sired by the focal male mated second to a female in a double mating experiment). These two studies exemplify that divergence in postmating prezygotic traits can occur rapidly between allopatric forms and often before the divergence of either premating or postzygotic traits. None of the sexes discrimi21.

(226) nated against a mate from another population in either paper I or II and there was no apparent genetic incompatibilities in hybrid offspring. This finding supports previous theoretical and empirical studies (Markow 1997; Clark et al. 1999; Howard 1999; Eady 2001). Due to postmating sexual selection, rapid divergence of reproductive characters is possible (see also Nilsson et al. 2002, 2003). While we showed, that conspecific sperm precedence is already effective at reducing gene flow in the early stages of divergence, we could find no evidence for female reducing there reproductive output after a copulation with a heterospecific to prevent the production of hybrid offspring. However, there still remains the possibility that other postcopulatory prezygotic traits prove a hindrance to gene flow (see Albuquerque et al. 1996; Markow 1997; Alipaz et al. 2001; Eady 2001; Price et al. 2001). Rates of adaptation The effect of sexual selection on the rate of adaptation is controversial. Sexual selection is predicted to either reinforce natural selection or hamper the rate of adaptation. Some models of sexual selection suggest that the rate of fixation of beneficial alleles is increased (Whitlock 2000; Lorch et al. 2003), or that deleterious alleles are purged more effectively (Agrawal 2001; Siller 2001), under sexual selection and that adaptation is accelerated as a consequence. Other sexual selection models, assuming no net benefits due to female choice (e.g. Fishers run-away process), show that sexual selection can instead impose a load on populations that may constrains populations from reaching their fitness optimum (Lande 1980; Kirkpatrick 1982). There is comparative support for the idea that sexual selection can be detrimental, increasing the probability of extinction (McLain et al. 1995, 1999; Sorci et al. 1998; Doherty et al. 2003; Morrow and Pitcher 2003, but see Morrow and Fricke 2004). Additionally, Kokko and Brooks (2003) pointed out that sexual conflict is likely to cause extinction to a higher extent, because costs and benefits are not carried by the same individual. Theory predicts that sexual selection that results from sexual conflict (Gavrilets et al. 2001) can result in a reproductive load inflicted on populations as females are harmed or are forced to expend on the evolution of resistance (Arnqvist and Rowe 2005) and there is some empirical support for this (Holland and Rice 1999). Furthermore, sexual selection among males can in theory lead to sub-optimal phenotypes in females due to genetic correlations across the sexes. Such intralocus sexual conflict can certainly depress population level fitness (Chippindale et al. 2001; Rice and Chippindale 2001; Arnqvist and Rowe 2005) and provides one important way in which sexual selection can impede the rate of adaptation. There is little consensus with regards to the net effect of sexual selection, and whether natural and sexual selection generally reinforce one another or, conversely, act in opposite directions in adaptive evolution is an open question. 22.

(227) The question how natural and sexual selection jointly influence divergence and in particular rates of adaptation was addressed in paper IV. Sixteen selection lines were established and the natural and sexual selection regime was altered by creating a novel versus ancestral habitat, and by varying the intensity of sexual selection. We tracked response to selection in these laboratory selection lines by conducting repeated fitness assays in generation 6, 20 and 35 after establishment of the selection lines (see paper IV for details). After 35 generations we furthermore tested how well these lines had adapted to the novel conditions. Our study revealed that the novel habitat effectively reduced population fitness and this was largely through it being primarily an unacceptable oviposition site for C. maculatus. Juvenile survival was actually higher on the novel host than on the ancestral host (F1,12 = 31.272; P = 0.001) and changed over time depending on host (Time × Host type: F2,24 = 4.576; P = 0.021) (see table 1). Adaptation to the novel host evolved fast as reflected by an overall improvement in performance (see table 1). Over time, there was an overall increase in the total number of offspring emerging, a rateindependent measure of population fitness, in the lines on the novel host. This was a result of an increase in the acceptance of chickpeas by females as a suitable oviposition host over the course of the experiment. Table 1. The mean values(rs.e.) for several components of fitness for selection lines on the ancestral or the novel host for all three fitness assays.. Host. Generation. Ancestral Novel. Juvenile survival. Host acceptance. Total offspring. Development rate. 6. 0.64r0.01. 0.88r0.01. 432.28r2.99. 0.83r0.004. 20. 0.64r0.01. 0.87r0.01. 388.35r4.86. 0.85r0.01. 35. 0.60r0.01. 0.89r0.01. 408.41r4.47. 0.82r0.004. 6. 0.76r0.01. 0.46r0.01. 228.35r4.92. 0.83r0.01. 20. 0.74r0.01. 0.55r0.01. 241.34r4.46. 0.83r0.01. 35. 0.67r0.01. 0.66r0.01. 296.68r5.31. 0.84r0.004. The rate of change in host acceptance was higher in the polygamous lines on the novel host, while it increased more rapidly under monogamy in lines kept on the old host (Time × Host type × Mating system: F2,24 = 4.215; P = 0.027) (see Figure 4). The same pattern of faster responses to selection in polygamous lines on the novel host and in monogamous lines on the ances23.

(228) tral host emerged also for populations’ fitness and development rate. Thus, our results reveal that contrasting effects of sexual selection dominated under the two natural selection regimes. 1.2 Novel - M ono Novel - Poly O ld - M ono O ld - Poly. Host acceptance. 1.0. 0.8. 0.6. 0.4. 0.2. 0.0 0. 5. 10. 15. 20. 25. 30. 35. 40. G eneration. Figure 4. Evolution of host acceptance in the four selection treatment combinations, with chick peas being the novel and black-eyed beans the old (i.e., ancestral) host, respectively. Acceptance rate was measured as the proportion of eggs laid on beans over the total number of eggs laid (arcsine transformed).. Sexual selection might act to facilitate the fixation of beneficial alleles under novel conditions (Whitlock 2000; Lorch et al. 2003), reinforcing natural selection more effectively. In contrast, it seems that for a population close to its adaptive peak, fitness costs due to a sexual selection load and/or sexually antagonistic adaptations are removed as genetic monogamy is enforced. Our study suggests that sexual selection may indeed critically affect the rate of adaptation, but that this depends on whether selection is acting in a novel or ancestral environment. Empirical data directly addressing this question are rare and somewhat equivocal. While some studies have found positive effects of sexual selection on components of adaptation (Promislow et al. 1998; Crudgington et al. 2005), others have found no clear net effect (Holland 2002; Martin and Hosken 2003) and yet others have documented negative effects (Holland and Rice 1999; Martin and Hosken 2004; Martin et al. 2004). Holland and Rice (1999) pointed to the fact that different environmental conditions should change the relative costs and benefits to populations from sexual selection. Our data support this idea.. 24.

(229) Divergence The type of selection causing divergence between populations in a given species leaves certain traces, which can be detected in experimental studies. Populations experiencing the same environment should be subject to the same natural selection resulting in parallel evolution. Sexual selection, in contrast, leads populations along arbitrary evolutionary trajectories resulting in distinct end points. In paper III we employed this logic and tested for the relative contribution of postmating sexual selection on population divergence using replicated populations. Laboratory systems provide us with a unique possibility of studying replicated episodes of allopatric speciation. We obtained two separate strains of Callosobruchus maculatus, which were split over the same set of laboratories (see Figure 5). We assumed each laboratory setting differed from the others, yet was stable over time, thus exposing the two strains in each laboratory to similar ecological conditions. On the other hand, the three populations belonging to each strain experienced different natural selection regimes. If natural selection shaped the reproductive traits in these populations, the pattern of divergence should be similar for the two strains. However, if sexual selection caused divergence each population should follow its own evolutionary path and we expect dissimilar patterns of divergence between the two strains. We found divergence in postmating prezygotic traits across these populations. Females varied significantly in their responses to mating signals/ ejaculatory substances from the different male types (see Figure 6). However, the pattern of male u female interaction was different in the two strains (eg. early fecundity: strain u male u female interaction: F4,329 = 4.02, P 0.01). As the pattern of divergence is not replicable and distinct for the two strains, we conclude that sexual selection has been the the primary selective agent causing divergence in postmating traits.. 25.

(230) Brazil USA. UK 1. South India UK 2. USA. UK 1. 1997. 1980-82. 1976. 1975. UK 2. 1992. 1979. Figure 5. Two strains of C. maculatus (Brazil and South India) with independent origin have been kept in the same set of laboratories and thus share a similar evolutionary history, in essence representing ‘replicated’ phylogenies. Brazil was collected in 1975 in Campinas, Brazil (Credland and Wright 1989), while South India was collected in 1979 in Tirunelveli (Mitchell 1991). Brazil came in 1976 to the National Resource Institute in England and was further distributed in 1980 to the University of Leicester (UK 2) and in 1982 to the University of London (UK1). The University of London (UK 1) population of South India was derived in 1997 from the population at the University of Leicester (UK 2), which came from the original strain in 1992 from the USA to England. Note that branch lengths in the figure are arbitrary.. While we cannot exclude the importance of additional natural selection driving divergence, our experiments suggest that sexual selection has been a powerful force operating in C. maculatus. The selection experiment (see Model species and Methods) offers the opportunity to more directly study the relative importance of natural and sexual selection on divergence in reproductive characters (paper V). After 35 generations of selection, we measured several male and female reproductive traits. We tested for an overall response to selection by crossing our lines back to the base lines as a common control. Individuals from selection lines differed significantly from those from the base population in many reproductive responses (e.g., P1: F1,331 = 10.58, P = 0.001; P2: F1,358 = 0.26, P = 0.614). Furthermore, the selection lines showed divergent evolution in reproductive characters as revealed by one-way analyses of variance (e.g., P1: F15,296 = 2.85, P < 0.001; P2: F15,323 = 2.42, P = 0.002).. 26.

(231) (a). 60 55. Number of eggs. 50 45 40 35 30. B r U S A m ales B r U K 2 m ale s B r U K 1 m ale s. 25 20 Br USA. Br UK 2. Br UK 1. F em ale p o pu la tio n. (b). 50. Number of eggs. 45 40 35 30 25 20. SI U SA m ales SI U K 2 m ales SI U K 1 m ales. 15 S I US A. S I UK 2. SI U K 1. Fem ale population. Figure 6. Female early fecundity, measured as the number of eggs laid during the first day following mating (r SE), in the various population crosses for the Brazil (a) and the South Indian (b) strains.. We tested for divergence by analysing 14 reproductive response variables with a geometrical approach (see Figure 6 and paper V for details). Reproductive traits diverged less across replicated selection lines evolving under intense natural selection than when selection was weak (F1,14 = 6.867; P = 0.020) (mean Euclidean distances; black-eyed bean-lines: 3.654, s.e. = 0.155; chick pea-lines: 3.084, s.e. = 0.153; see Figure 7). In contrast, there was no significant effect of the mating system treatment (F1,14 = 0.071; P = 0.794) nor did our two treatments interact significantly (F3,12 = 2.373; P = 0.121).. 27.

(232) Female net resistance to remating. 2. B/P. 1. B/M. 0. B/P C/M B/M C/P C/P C/P C C/M C/MC/P B B/P. -1. B/P. C/M. B/M. -2 -2. -1. 0. B/M. 1. 2. Male relative defense ability (P1) Figure 7. Illustration of the strategy used to analyze reproductive divergence. The figure shows ordination of all replicated selection lines in a two-dimensional space described by two reproductive variables (standardized). The linear distance between each line and its treatment level mean (exemplified here with food treatment – denoted by encircled letters; B: black eyed beans; C: chick peas) represents the Euclidean distance for each line to its level mean. For a given food treatment level, the eight Euclidean distances collectively describes the degree of evolutionary divergence within that treatment level. Our analysis shows that evolutionary divergence across lines within treatment level, in 14 dimensions simultaneously, is significantly larger when kept on black eyed beans (solid lines) compared to chick peas (hatched lines). Ellipses represent Gaussian bivariate confidence (P = 0.6) ellipses. (Food - C: chick peas; B: black eyed beans; Mating system - M: monogamy; P: polygamy).. Hence, under strong natural selection reproductive traits diverged less and more slowly compared to weak natural selection. This effect was independent of the intensity of sexual selection in these selection lines. It seems that the selection lines under strong natural selection were constrained to respond to the sexual selection treatment. This might be because of antagonistic selection on some of the traits measured or genetic correlations restricting independent responses to selection (Dickerson 1955; Hansen et al 2003). Thus genetic constraints can effectively limit evolutionary responses if populations are subject to strong natural selection and this only allows populations to evolve along a limited number of evolutionary pathways (e.g. multivariate lines of least genetic resistance, sensu Schluter [1996]). This finding implies, 28.

(233) that natural selection can actually hamper divergent evolution. When reproductive isolation evolves as a pleiotropic by-product of divergence in allopatric populations under different ecological conditions (Dobzhansky 1937; Mayr 1970; Schluter 2000, 2001; Rundle and Nosil 2005) our results imply that strong natural selection might decelerate the pace of speciation.. 29.

(234) Conclusions. This thesis presents convincing evidence that postmating sexual selection is a powerful engine in driving population divergence. Traits underlying variables such as female reproductive output, female mating rate or male success in sperm competition evolved rapidly and could in some cases effectively reduce gene flow between allopatric populations of a single species. Furthermore, as revealed by significant male-female interaction terms for a number of postcopulatory reproductive traits, divergence can potentially occur at different steps leading to fertilization and partially reduce gene flow. Thus, there is ample opportunity for reproductive incompatibilities to evolve in an arbitrary fashion with different allopatric populations taking distinct evolutionary trajectories. Postmating sexual selection is a prime candidate causing cryptic isolation between incipient species and fueling the early stages of speciation. However, the outcome of evolution depends critically on the joint effects of natural and sexual selection acting on a certain population. Sexual selection can act like a double-edged sword and its net benefit depends on the prevailing natural selection regime. While sexual selection can reinforce natural selection in its effect, it can also inflict a reproductive load effectively impeding adaptation in a given population. I also present compelling evidence that strong natural selection slows down evolution in reproductive characters causing reduced divergence. These results illustrate the importance of studying sexual selection in conjunction with natural selection to gain a deeper understanding of the processes causing early divergence between incipient forms.. 30.

(235) Summary in Swedish (Sammanfattning). Sexuell selektion efter parningen och dess betydelse för populationsdivergens hos skalbaggar Det finns en fantastisk mångfald av former och organismer i naturen, vilket avspeglar många miljoner års biologiska evolution. Arter är biodiversitetens grundstenar och vår förståelse för arternas uppkomst har länge varit central inom evolutionsbiologin. Vår grundläggande kunskap om arternas uppkomst är idag god, men vissa detaljer återstår att klarlägga. Artbildning, processen som i förlängningen ger upphov till nya arter, börjar ofta med uppdelningen av en population till två eller flera separata undergrupper vilka är skilda från varandra (s.k. allopatriska populationer). Dessa undergrupper formas sedan det naturliga urvalet (s.k. naturlig selektion) till att följa olika evolutionära ”vägar”. Genom sådan divergens uppstår med tiden reproduktiva inkompatabiliteter, vilket gör att individer ej kan reproducera sej lika framgångrikt mellan som inom grupperna. Sådana inkompatabiliteter hindrar, eller minskar, därför flödet av genetiskt material mellan allopatriska populationer. Artbildningen kan anses komplett när olika former är helt oförmögna att reproducera sej med varandra, något som kallas reproduktiv isolering, och ekologiska skillnader tillåter då ofta arternas samexistens. I sina banbrytande bidrag betonade Darwin att selektion för individers förmåga att överleva i sin miljö, d.v.s. finna föda, undvika rovdjur, etc. (naturlig selektion i smal bemärkelse), har orsakat mycket av den artrikedom vi kan observera. Populationer som lever inom olika miljöer tenderar att anpassa sig till de rådande förutsätningarna för överlevnad. I sådana fall kan adaptiv evolution leda till reproduktiv isolering på två olika sätt. För det första så kan anpassningar till miljön, som en ren bieffekt, leda till att karaktärer som påverkar reproduktionen också förändras. Detta kan ske genom evolution av reproduktiva karaktärer som är genetiskt korrelerade med de som förändras genom anpassning. För det andra så kan karaktärer vilka är inblandade i anpassning till miljön även bilda basen för ett könsurval (dvs., sexuell selektion). Honor kan t.ex. välja partners vilka liknar dem själva, vilket ger upphov till reproduktiv isolerning i och med parning då ”sorteras” mellan individer som på något sätt är olika. De senaste decennierna har betydelsen av sexuell selektion, vilken orsakas av variation mellan individer i deras förmåga att reproducera sig framgångsrikt, vunnit mer och mer uppmärksamhet som en potentiellt viktig art31.

(236) bildningsprocess. För hanar innebär parningar inte nödvändigtvis framgångsrik reproduktion. Vi vet idag att honor av de flesta arter parar sig med flera olika hanar inom en reproduktiv episod, och spermier från olika hanar konkurerar därför med varandra över befruktandet av en viss honas ägg. Denna form av reproduktiv konkurrens, vilken alltså äger rum efter efter parningen, förekommer i två former (spermiekonkurens och s.k. kryptiskt honligt val). Sexuell selektion kan vara en kraftfull generator för divergens som snabbt kan reducera genflödet mellan divergerande former, dels för att den verkar direkt på reproduktiva karaktärer och dels för att den inbegriper väldigt många olika typer av egenskaper hos organismer. Det finns idag goda bevis från jämförande arbeten som stödjer idén att sexuell selektion kan orsaka artbildning. Det övergripande syftet med denna avhandling är att utröna huruvida sexuell selektion efter parningen kan bidra till de tidiga faserna av reproduktiv divergens mellan allopatriska populationer, och alltså kan medverka till reducerat genflöde och artbildning. För att kunna belysa dessa frågor genomfördes en serie laboratorie experiment, där allopatriska populationer av två olika skalbaggsarter användes som modellsystem (mjölbaggar, Tribolium castaneum, och bönvivlar, Callosobruchus maculatus). Två huvudsakliga metodiska strategier följdes. För det första utnyttjades en jämförande metodologi, där vi sökte mönster över allopatriska populationer vad gäller förhållandet mellan släktskap och reproduktiv divergens. Dessa mönster användes sedan för att söka utröna underliggande evolutionära processer. För det andra genomfördes ett artificiellt selektionsexperiment, där naturlig och sexuell selektion manipulerades i 40 efterföljande generationer varefter såväl anpassning till den rådande miljön som förändningar i reproduktiva karaktärer uppmättes. De allopatriska populationer som studerades uppvisade alla någon grad av reproduktiv divergens. I vissa fall kan denna divergens begränsa genflödet mellan populationer och därmed vara ett första steg mot artbildning. Mönstret vilket allopatriska populationer uppvisade, speciellt snabb och divergent evolution av karaktärer som påverkar spermiekonkurens och kryptiskt honligt val, indikerar att sexuell selektion efter parningar är en viktig arbildningsprocess hos dessa insekter. En mer komplex bild framträdde ur det artificiella selektionsexperimentet, vilket belyste de oberoende effekterna av naturlig och sexuell selektion på divergens. Vi fann att populationer divergerade i mindre grad under stark naturlig selektion. Detta innebär att naturlig selektion kan motverka divergens via sexuell selektion, förmodligen genom att evolutionen av reproduktiva karaktärer hindras av stark naturlig selektion. Vi observerade också att populationer i vilka sexuell selektion tilläts anpassade sig snabbare till en ny miljö. Detta försök visade alltså att sexuell selektion i vissa fall kan förstärka effekterna av naturlig selektion, förmodligen genom att öka hastigheten med vilken nya och fördelaktiga mutationer sprids i populationer. Vi visade emel32.

(237) lertid också att sexuell selektion kan ha den motsatta effekten. Under svag naturlig selektion observerades en nackdel med sexuell selektion, så att reproduktiv konkurrens medförde en börda för populationer. Effekterna av sexuell selektion på populationer berodde alltså kritiskt på naturlig selektion. De resultat som presenteras i denna avhandling utgör sammantaget goda bevis för att sexuell selektion efter parningen kan medverka till evolutionen av divergens mellan populationer och, i förlängningen, bildandet av nya arter. Resultaten har också viktiga konsekvenser för vår förståelse av hur reproduktiv konkurrens mellan individer (sexuell selektion) påverkar den hastighet med vilken populationer kan anpassa sig till sin miljö. Den övergripande slutsatsen är att sexuell selektion efter parningen är en effektiv process vad gäller evolutionen av reducerat genflöde mellan divergerande populationer, men att denna process måste ses i samspel med naturlig selektion för att fördjupa vår förståelse för artbildningens allra tidigaste stadier.. 33.

(238) Summary in German (Zusammenfassung). Die Bedeutung von postkopulativer sexueller Selektion auf die Divergenz von Käferpopulationen. Mit Erstaunen bewundern wir die Vielfalt der Organismen in unserer Welt. Diese Artenfülle ist das Resultat von jahrmillionenlanger biologischer Evolution. Arten sind die Grundbausteine von Biodiversität aber unser Wissen über die Prozesse, die zur Entstehung neuer Arten beitragen, ist in einigen Bereichen immer noch begrenzt. Artbildung beginnt mit der Aufspaltung einer existierenden Population in zwei oder mehrere Untereinheiten. Diese Untereinheiten befinden sich häufig in unterschiedlichen Habitaten und sind deshalb verschiedenen Selektionsdrücken ausgesetzt, die wiederum diese Untereinheiten entlang unterschiedlicher evolutionärer Pfade führen. In diesen Populationen werden sich folglich unterschiedliche biologische Merkmale etablieren, die zum Beispiel zu reproduktiver Unverträglichkeit (d.h. zu reproduktiver Inkompatibilität) zwischen den Populationen führen können, dadurch pflanzen sich Individuen von unterschiedlicher Gruppen weniger erfolgreich fort, als Individuen von der gleichen Gruppe. Dies trägt zu einem verringertem Genausstauch zwischen divergierenden Gruppen bei. Der Prozess der Artbildung ist dann abgeschlossen, wenn die divergierenden Populationen sich nicht mehr reproduzieren können und somit der Austausch von genetischem Material unterbleibt. In seinen bahnbrechenden Veröffentlichungen hat Darwin unsere Aufmerksamkeit auf die Bedeutung des natürlichen Auswahlprozesses gerichtet, der Eigenschaften wie die Futtersuche oder das Fluchtverhalten gegenüber Prädatoren in Organismen formt Diese Selektionsform liegt dem von uns beobachteten Artenreichtum in der Natur zugrunde. Durch natürliche Selektion passen sich Populationen in unterschiedlichen Habitaten den dort gegebenen Umständen an. In diesem Szenario entsteht reproduktive Inkompatibilität durch adaptive Evolution auf zwei Wegen. Erstens können Barrieren gegen den Genaustausch durch korrelierte Evolution entstehen. Hierbei wird ein Merkmal X durch natürliche Selektion verändert und zieht dabei ein durch Pleiotropie verbundenes Merkmal Y mit. Dieses Merkmal Y wirkt sich negativ auf die reproduktive Kompatibilität der divergierenden Populationen aus. Zweitens kann ein Merkmal unter natürlicher Selektion gleichzeitig als Auswahlkriterium im assortiativen Paarungsverhalten genutzt werden, wobei Weibchen Männchen bevorzugen, die ihnen selber ähnlich sind 34.

(239) und dadurch reproduktive Isolation entsteht, da sich die Paare in verschiedene Gruppen „sortieren“ und damit den Genaustausch zwischen den Gruppen verringern. Während der letzten Jahrzehnte hat die Einsicht zugenommen, dass sexuelle Selektion ein wichtiger Faktor für die Divergenz von Populationen sein kann. Sexuelle Selektion entsteht, wenn Individuen in ihrem Vermögen sich fortzupflanzen und somit Gene in die nächste Generation weiterzugeben, variieren. Sexuelle Selektion wird als eine kraftvolle Antriebskraft für die Artbildung betrachtet, da sie reproduktive Merkmale direkt verändert, wobei diese Merkmale schnell zur reproduktiven Unverträglichkeiten führen und dadurch einen begrenzten Genaustausch verursachten können. Beweise für diese Theorie kommen von vergleichenden Studien, die zeigen, dass artreichere Gattungen im Durchschnitt eine höhere Anzahl sexuell dimorpher Arten enthalten. Durch Studien, die neu entwickelte molekularbiologische Methoden verwenden, zeichnet sich immer stärker ab, dass die meisten Tierarten ein polygames Paarungsverhalten haben und sich während einer Paarungssaison mit mehr als einem Partner paaren. Diese Einsicht führte zu einer Erweiterung der sexuellen Selektionstheorie, in die nun auch Ereignisse mit einbezogen werden, die nach einer erfolgreichen Kopulation aber vor der Zygotenbildung ablaufen. Somit konkurrieren in sich mehrmals paarenden Weibchen die Spermien von mehreren Männchen um die Befruchtung von Eiern. Weibchen verhalten sich dabei aber nicht passiv, sondern können aktiv die Spermien bestimmter Männchen für die Befruchtung ihrer Eier bevorzugen. Postkopulative sexuelle Selektion ist ein wichtiger Prozess, der zu schnellen Veränderungen führt und Auswirkungen auf die Physiologie, das Paarungsverhalten und die generelle Morphologie/Genitalienmorphologie der beiden Geschlechter haben kann. Das Ziel dieser Doktorarbeit ist es, die Bedeutung von postkopulativer sexueller Selektion für die anfängliche Divergenz von Populationen zu untersuchen und einschätzen zu können, ob verschiedene postkopulative Merkmale das Potential besitzen, den Genfluss zwischen divergierenden Populationen einzuschränken. Dazu habe ich Laborexperimente an allopatrischen Populationen von Mehlkäfern, Tribolium castaneum, oder Bohnenkäfern, Callosobruchus maculatus, durchgeführt. Hierbei wurden hauptsächlich zwei Methoden angewandt. 1.) Ein Selektionsversuch, um die Effekte von natürlicher und sexueller Selektion auf die Anpassungsgeschwindigkeit und frühzeitige Divergenz von Populationen entflechten zu können. 2.) Ein phylogenetischer Ansatz, in dem Weibchen mit Männchen mit abnehmendem Verwandtschaftsgrad gepaart wurden. Die reproduktive Reaktion der Weibchen ermöglicht es, durch das entstehende Muster zwischen reproduktiver Antwort und Verwandtschaftsgrad zwischen den Paarungspartnern, Rückschlüsse auf die historischen Selektionsdrücke zu ziehen, die diese Divergenz bewirkt haben. 35.

(240) Meine Doktorarbeit zeigt, dass postkopulative sexuelle Prozesse wichtig sind und schnell zur anfänglichen Divergenz von Populationen beitragen. Alle untersuchten Populationen zeigten Divergenz zu einem gewissen Grad. Weiterhin konnten in einigen Fällen die gemessenen postkopulativen Merkmale erfolgreich zur Reduktion von Genfluss zwischen diesen Populationen beitragen. Ich schliesse daraus, dass postkopulative sexuelle Selektion ein wichtiger Faktor in den Anfangsstadien der Artbildung ist. Wenn allerdings die gemeinsamen Effekte von natürlicher und sexueller Selektion auf divergierende Populationen betrachtet wird, entsteht ein komplizierteres Bild. Natürliche und sexuelle Selektion können sehr gegensätzliche Effekte aufeinander und damit auf den Differenzierungsprozess von Populationen haben. Starker natürlicher Selektionsdruck behindert die Divergenz von Merkmalen durch sexuelle Selektion vermutlich da starker natürlicher Selektionsdruck die Evolution von reproduktiven Merkmalen einschränkt. Im Gegensatz dazu haben wir beobachtet, dass sexuelle Selektion den Effekt von natürlicher Selektion verstärken kann, vermutlich durch eine erhöhte Fixierungsrate von neuen und vorteilhaften Mutationen, und dadurch eine erhöhte Anpassungsgeschwindigkeit von Populationen an neue Habitate bewirken kann. Ausserdem hängt der Nettonutzen von sexueller Selektion kritisch von den vorherrschenden natürlichen Bedingungen ab. Unter schwachem natürlichen Selektionsdruck kann sexuelle Selektion sich nachteilig auf die Adaptationsfähigkeit einer Population auswirken, da sexuelle Selektion eine reproduktiven Last mit sich führen kann und dadurch eine Population darin behindert, das volle Anpassungspotential zu erlangen. Das Fazit meiner Arbeit und der hier präsentierten Resultate ist, dass postkopulative sexuelle Selektion ein kraftvoller Prozess ist, der zur Evolution von Divergenz führt und zur Verminderung von Genaustausch zwischen Populationen beitragen kann, die sich auf dem Weg zu neuen Arten befinden. Allerdings sollten die Effekte von postkopulativer sexueller Selektion im Zusammenspiel mit natürlicher Selektion betrachtet werden, um unser Verständnis für den Artbildungsprozess zu vertiefen.. 36.

(241) Acknowledgments. I thank Göran Arnqvist, Mats Björklund, Marnie Demandt, Dietlinde Fricke, Gesine Fricke, Kirsten Fricke, Mari Katvala, Silke Langenheder, Johanna Rönn and Chris Wiley for commenting on all or parts of the thesis. The work in this thesis was made possible by financial support of several funding sources: a fellowship from the DAAD (Doktorandenstipendium im Rahmen des gemeinsamen Hochschulsonderprogramms III von Bund und Ländern) financed my first two years as a graduate student. I further thank Stiftelsen för Zoologisk Forskning, Lars Hiertas Minne, KVA and Sederholms för utrikes resor for financial support. Grants from Rektors Wallenbergmedel and Liljewalchs resestipendium made travels to international conferences possible. Tack, Thank you, Danke schön Saving the best for last. This thesis would not have come into existence without the help, encouragement and support from a whole bunch of different people, who made this whole experience so much fun. Göran, I am incredibly grateful to you for taking on the challenge and being such a fantastic supervisor. You were always extremely helpful, supportive and took your time to explain, teach and discuss all sorts of crazy ideas. Thanks, it would not have been possible without you. I would also like to thank Mats Björklund for functioning as my supervisor in the beginning and helping me all along the way. Anders B., Lars G. Ingrid A. and Anna Q. thanks for fun discussions about science and life in general (Anders, are you still organising that PhD student trip??). Ingela, thanks for all the administrative help and the good chats during these years. I probably would not have endeavoured on a four-year PhD in Sweden if I wouldn’t have had so much fun in Umeå. My two flat-mates Phil and Patrick allowed me to crash in a small room next to their kitchen for six month and then another six month and another until I left Umeå three years later with a crying and a laughing eye. All you guys, you know who you are -, it was a fantastic time with oh so many parties. I loved it. I had the pleasure to meet and work together with a great number of wonderful people, which made this whole experience unforgettable. There is one person, who was a constant through all these years and I have the deepest admiration for his scientific ideas, sharp mind, theories concerning everything in everyday life and great sense of humour. Urban, thank you for being 37.

No results found