ACTAUNIVERSITATISUPSALIENSISUPPSALA2006

Digital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 157

Morphological and BehaviouralDifferentiation in a Pipefish

SARAH ROBINSON-WOLRATH

ISSN 1651-6214ISBN 91-554-6501-3urn:nbn:se:uu:diva-6666

List of papers

This thesis is based on the following papers, which are referred to in the text by their Roman numerals:

I Robinson-Wolrath, S.I., Hoffman, E.A., Berglund, A. & Jones, A. Phenotypic differentiation without genetic divergence among populations of pipefish. Submitted manuscript.

II Robinson-Wolrath, S.I., Berglund, A. & Rosenqvist, G. Individual and population differences in behavioural traits of a sex-role reversed pipefish. Submitted manuscript.

III Robinson-Wolrath, S.I. Adaptive strategies of a female pipefish in response to perceived mortality risks. Submitted manuscript.

IV Robinson-Wolrath, S.I. & Billing, A. Sequential mate choice in a male pipefish. Manuscript.

V Robinson-Wolrath, S.I. Video playback versus live stimuli for assessing mate choice in a pipefish. Environmental Biology of Fishes In press.

Paper V is reprinted with kind permission of Springer Science and Business Media.

Contributions I performed all work on which this thesis is based, with the following exceptions. E. Hoffman and A. Jones carried out the genetic analyses reported in Paper I. A. Berglund contributed towards the planning and running of the experiment reported in Paper II, and was involved in the final stages of writing papers. G. Rosenqvist contributed towards Paper II, specifically, the running of the experiment and the final stages of writing. A. Billing contributed towards Paper IV, specifically, the running of the experiment and the final stages of writing.

Contents

Introduction.....................................................................................................7Population differentiation...........................................................................7

Behavioural differentiation....................................................................7Mate sampling behaviour ...........................................................................8

Study Species ..................................................................................................9

Thesis aims ...................................................................................................11

Results and discussion ..................................................................................12Phenotypic differentiation without genetic divergence............................12Behavioural differentiation ......................................................................14

Personalities.........................................................................................14Alternative reproductive investment strategies....................................16

Sequential mate choice.............................................................................19Video playback.........................................................................................20

Conclusions...................................................................................................21

Sammanfattning på svenska..........................................................................22

Acknowledgements.......................................................................................27

References.....................................................................................................29

7

Introduction

Population differentiation Genetic and phenotypic differences among populations of organisms are at the core of modern evolutionary and ecological theory. A wide array of evolutionary mechanisms can produce population divergence. Perhaps the most significant mechanisms are genetic drift (at low population densities) and natural selection, but other mechanisms, such as phenotypic plasticity and sexual selection, also potentially play a role in population differentiation at the level of the genotype or the phenotype. If distinct populations occupy environments that differ with respect to important ecological parameters, then local adaptation or phenotypic plasticity may be important mechanisms producing population differentiation (Piersma & Drent 2003). We expect the fitness of an individual to be greater when its phenotype matches the environment in which it occurs (Kingslover et al. 2002), and such phenotypic fine-tuning can be achieved through selection or plasticity. Given the wide array of mechanisms that can produce population differentiation, a major goal of evolutionary biology is to understand the evolution of divergence among distinct natural populations (Rogers et al. 2002).

Behavioural differentiation Population divergence is often closely associated with behavioural differentiation by the process of adaptation. A meaningful approach here is to investigate suites of behavioural traits for individuals (i.e., “personalities”) of a species which inhabit different environments.

Personalities/Adaptive strategies The study of animal personalities (and adaptive strategies) has critical implications for evolution, because when different behaviours are correlated they evolve not in isolation, but as a package. These correlations are not always adaptive and can generate trade-offs across situations that may be important in evolution (Sih et al. 2004a; Dingemanse & Reale 2005). For instance, if boldness and aggression are positively correlated with each other, then aggressive individuals might do well in situations where high levels of aggression are favoured, for example during competition for resources, but do poorly when high levels of boldness are deleterious, for

8

example under predation risk (Sih et al. 2004c). Behavioural correlations can also be adaptive (Neff & Sherman 2004; Sih et al. 2004b; Dingemanse & Reale 2005), as they may simply represent alternative behavioural strategies for coping with a variable environment (e.g. Hedrick 2000). Indeed, recent work showing that populations differ in the structure of their behavioural correlations implies that the correlations themselves are not constrained, but can evolve in an adaptive fashion (e.g. sticklebacks, Bell & Stamps 2004; Bell 2005).

Mate sampling behaviourMate-sampling behaviour has been used to describe the process by which individuals gather information on potential mates and make decisions based on that information (reviewed by Jennions & Petrie 1997). One reason for the interest in this behaviour is the growing awareness of its influence on the outcomes of coevolutionary models of male tactics and female preferences (Pomiankowski 1987; de Jong & Sabelis 1991; Kokko & Mappes 2005). If a female with strong preferences runs a slight risk of remaining unmated, the opportunity for a “runaway” evolution of sexual ornaments is greatly lowered (known as the wallflower effect: de Jong & Sabelis 1991). Consequently, there is a drive for individuals to adopt optimal mate-sampling behaviours, i.e. to trade-off the risk of remaining unmated against the risk of mating with the ‘wrong’ individuals.

Multiple mating has been suggested as a behavioural strategy which may ensure successful fertilization (Sheldon 1994; Jennions & Petrie 2000; Preston et al. 2005). To determine whether a lack of mates and consequently a fear of unsuccessful reproduction is responsible for animals adopting a multiple mating strategy is challenging (Kokko & Mappes 2005). However, based upon a theoretical model proposed by Kokko and Mappes (2005), it is possible to begin by experimentally investigating a particular adaptation of the choosing sex in multiply mating species. Specifically, to investigate the ability of individuals to adjust levels of choosiness when encountering sequential mates. It is this adjustment which has been suggested to optimise mating success and ensure successful fertilisation. In the case of conventional (not reversed) sex roles, it is often costly for females to delay reproduction, and it has been suggested that an optimally behaving female should first mate unselectively, and then increase in selectivity to improve mate quality in later matings (“trade-up”, Jennions & Petrie 2000). This trade-up adaptation (form of mate-sampling behaviour) may help explain the persistence of multiple mating patterns in nature.

9

Study Species

The deep-snouted pipefish (Syngnathus typhle) inhabits meadows of eelgrass (Zostera marina) along the European coast. In this species of pipefish the sex roles are reversed, such that males actively choose among and reject some females, whereas females vigorously display towards males, often in temporary groups in a leklike fashion (for review see, Berglund & Rosenqvist 2003). After a courtship dance, the female transfers her eggs to the male’s brood pouch, where they are fertilized. The male broods the offspring in a pouch formed by two skin folds along the tail which merge during pregnancy and a placenta-like structure develops. The long male pregnancy lowers the potential reproductive rate of males below that of females (Berglund et al. 1989; Berglund & Rosenqvist 1990).

Reproductive performance in this pipefish depends on body size in both males and females, with the larger females producing more and larger eggs, and larger males accommodating more offspring (Berglund et al. 1986a-b). Therefore, with both sexes aiming to attract the best quality and most numerous mates, there is a drive to be as large as one can possibly be. It is, therefore, likely that the environment and resulting optimal mate-sampling behaviour in this species will have a strong influence, to either slow down or speed up, the ‘runaway’ preference for larger body size.

Population structure has not, until now, been studied in this species on any geographic scale, but observations of the natural history of S. typhlehave led to the belief that they are poor dispersers that should exhibit substantial genetic differentiation. This assumption is based on this species’ poor swimming ability and cryptic phenotype in its eelgrass environment, and the lack of planktonic larvae. Individuals align themselves vertically within the eelgrass, where their body shape and coloration closely approximate the surrounding vegetation (Vincent et al. 1995). Newborns are miniature versions of the adults that are believed to remain in the local seagrass bed. Thus, movement of these slow swimmers between eelgrass meadows, which are separated by bare sandy substrate, would make the fish extremely vulnerable to predation.

Overall, the existing knowledge on pipefish make it an ideal model to investigate population differentiation, at genetic, morphological and behavioural levels, and examine mechanisms underlying the persistence of multiple mating in this species.

10

All papers in this thesis are based on data collected from fish collected near Kristenberg Marine Research Station in Sweden (58o15'N, 11o28'E) during May and June 2002, 2003, 2004 and 2005. The fishes were captured in May by trawling in three eelgrass meadows in the Gullmar Fjord (Figure 1), before the breeding season. I kept the male and female fish separate in 225 L barrels and fed them live Artemia and frozen mysids twice a day before any experiments were undertaken. Water was continuously renewed so temperature and salinity followed the natural variation at 4m depth. All fish were released back to the wild after experiments, when possible.

Fiske-bäcks-kil

X

X

X

1

2

3

1 km

Figure 1. Map showing location of the populations sampled in this study.

11

Thesis aims

The main aim of this thesis was to investigate the interplay between individuals and the environment which they inhabit. The processes of phenotypic fine-tuning to maximise the fitness of an individual in its environment is fundamental to these studies. I have three principal aims with this thesis, namely:

I. To investigate population differentiation:

To assess the population structure of pipefish on a microgeographic scale (Paper I).To determine if morphological differentiation among populations is related to behavioural differences (Paper II).To determine if morphological differentiation among populations is a consequence of alternative reproductive investment strategies of females (Paper III).

II. To investigate the persistence of multiple mating :

To determine whether the persistence of multiple mating in the pipefish can be described by the trade-up hypothesis as a process to ensure successful fertilization (Paper IV).

III. To investigate new experimental methods:

To determine the suitability of the video playback technique to study mate choice in the pipefish. (Paper V).

12

Results and discussion

Phenotypic differentiation without genetic divergence The aim of paper I was to assess the population structure of pipefish on a microgeographic scale. The assumed poor dispersal ability of the pipefish (Vincent et al. 1995) led to the prediction that they should exhibit substantial genetic and morphological differentiation among sites. To investigate their population structure we, firstly, used a suite of microsatellite markers to establish whether or not there is significant genetic structure among pipefish populations in different eelgrass meadows. Second, we looked for evidence of local adaptation by characterizing patterns of pipefish phenotypic variation within and between the different eelgrass meadows. Finally, the physical environment of the three sites was sampled to provide a detailed description of their habitats. Taken together, this provided us with information on the population structure of the pipefish studied in this thesis.

Contrary to predicted, genetic analyses revealed that the populations were genetically panmictic (FST ,p>0.05) Interestingly, however, the morphometric analyses revealed significant differences among the populations for females (p<0.05), but not for males (p>0.05). Females became characteristically larger in size from site 1 to site 2 to site 3. In relation to qualitative measures, for females, the contrast in colouration between their dorsal and ventral surfaces was significantly different between sites 1 and 3 (ANOVA: F=4.46, df=2, p=0.01; Tukey HSD: p<0.05).

Sampling of the eelgrass meadows which they inhabit revealed a gradual difference between sites, from site 1 to site 2 to site 3, with an increase in depth, increase in the percent of living eelgrass, increase in visibility, decrease in the number of other species present, and a decrease in the percent of algae (Table 1). These characteristics allow for interpretation of morphological differences between populations to be attributed to environmental differences (local adaptation).

13

Table 1. Description of sampling sites, with numbers representing averages over the 10 quadrats (per m2) scored at each site.

Site 1 2 3 Substrate Sand Sand/shell Mud Depth (m) 1.11 1.41 2.47 Temperature (˚C) 16.5 14.5 16.8 Visibility (m) 2.3 3 3.5 % living eelgrass 14.05 15.5 43 Number of anemones 0.4 0 35.2 Total number of ‘other species’ 8 6 3 Number of parasite-bearing snails 10.8 7.6 0 % algae cover on eelgrass 10.6 9.5 4

Our results are surprising, because we were expecting to find at least some genetic differentiation among distinct eelgrass meadows. Instead, our results indicate that the pipefish are essentially genetically panmictic on small geographic scales (up to at least 4 km). Given the genetic panmixia, one question is, how females from different populations can remain phenotypically distinct. Here I provide, amongst many, three explanations for our results.

Firstly, the pipefish may exhibit phenotypic plasticity, such that, for females, the developmental instability, maintenance, production and genetic costs of plasticity (DeWitt et al. 1998) may be outweighed by the potential gains from greater survivorship, higher reproductive success or increased mating success. Plasticity may be more advantageous for females than males, because males are considered the choosing sex in this species whereas females compete among themselves for access to mates.

A second explanation could be that the morphological differentiation between sites for females could be a consequence of natural selection imposed by the different physical environments (for an example on guppies, see Arendt & Reznick 2005). When environments within the range of a species differ it is unlikely that a single phenotype will confer high fitness in all situations (Via et al. 1995; Viera et al. 2000; Mackay 2001; Borevitz et al. 2002; Sgro & Hoffmann 2004). Thus, although the populations start at birth with identical phenotypic distributions (due to the homogenizing effects of gene flow), differential survival due to selection in dissimilar habitats may result in different phenotypic distributions in the adults. In this study we, again, propose that the females may be more susceptible to selection, because of the risks they must take when initiating and carrying out courtship.

Finally, there could be differential migration from a common over-winter site. However, the long, barren and deep migration routes between sites makes this explanation less likely.

14

Overall, the study on the population structure of pipefish, which revealed discordance between patterns of genetic differentiation and patterns of morphological divergence, provided the basis for further investigation into the mechanisms producing interpopulation phenotypic differences in pipefish.

Behavioural differentiationMorphological divergence between sexes and among geographically distinct populations is often closely associated with behavioural differentiation by the process of adaptation. In papers II and III we characterised suites of behavioural traits of individuals from three populations of pipefish (Syngnathus typhle) off the west coast of Sweden in an attempt to understand why there is morphological differentiation, without genetic divergence, between these populations. Specifically, we characterized: firstly, suites of correlated behaviours (“personalities”) for both sexes across populations, and secondly, behavioural responses to perceived mortality risks for females across populations.

PersonalitiesThe aim of paper II was to test a prediction that both the populations and the sexes may elicit a variety of habitat selection strategies to use resources, and thus may differ in the behavioural traits used to exploit those resources in different sites. If such behavioural differences exist, it may provide, at least in part, an explanation for the morphological differentiation among populations. Specifically, four related questions were addressed. Firstly, do the pipefish have correlated suites of behavioural traits (personalities)? Secondly, are there any personality differences between the sexes? Thirdly, are there any personality differences between the sites? And finally, are there any individual behaviours which consistently differ between sites?

Under controlled experimental conditions, behavioural measures of individuals, of both sexes, from the three pipefish populations were collected. Four different behaviours were recorded: activity level, aggression, choosiness/courtship, and boldness under predation risk, from which personality types could be described.

This study revealed a personality which explained 20% of the behavioural variation among pipefish individuals in the study populations. This personality was, however, not characteristic of a particular sex or site. Analyses also revealed that the only behaviour which was characteristically different between sites was courtship swimming rate (the amount of time an individual was observed swimming in its compartment during the courtship/choosiness context) (ANCOVA: site, F2,119 = 7.08, p=0.001; Figure 2).

15

1 2 3

site

2

3

4

5

6

7

8

9

10

11co

urts

hip

swim

min

g fr

eque

ncy

Figure 2. Differences between sites with respect to courtship swimming frequency, with site 3 fish having a significantly lower frequency of measured courtship swimming compared to the other two sites.

Firstly, why did this study only characterise one personality, explaining a comparatively small percentage in all behavioural variation? This can be explained by the possibility that pipefish benefit from exhibiting a large degree of behavioural flexibility as a means to survive in an unpredictable and varying environment.

Secondly, why were there no sex differences in behaviour? It is unique that a behavioural study, which has simultaneously investigated numerous behavioural traits in both sexes, has found no difference between males and females in a sex-role reversed species. We discuss our finding in terms of, the mutual mate choice prevalent in this species (Sandvik et al. 2000; Berglund et al. 2005; Sandvik Widemo 2006). A characteristic of mutual mate choice is that both sexes are actively choosing their mate (as they do in S. typhle, where larger mates are preferred: Berglund et al. 1986a, 2005) and as a consequence both sexes may be aggressive towards potential competitors for a ‘good’ mate (as they are in S. typhle: Berglund et al. 2005). Hence, when a male or a female is put in the same experimental situation their behaviours may be identical. This may explain, for example, the absence of a difference between the sexes in their respective interest to be involved in aggression interactions. Furthermore, the similarity between the sexes in boldness (measured in this study as willingness to take a foraging

16

risk) can be explained in terms of both sexes requiring resources as much as the other, just for different purposes. In nature, however, males and females do not experience identical situations, so therefore different behaviours are recorded in field situations. Nevertheless, in this experiment the lack of sex differences in behaviour in pipefish does challenge theories of sex differences as a consequence of differential parental investment (Trivers 1972; Johnstone et al. 1996; Wade & Shuster 2002; Borgerhoff Mulder 2004), with males in this species contributing markedly more resources than females to parental investment. This study, therefore, urges for careful interpretation of behavioural studies when characterising the sexes (also suggested by, Kolm & Berglund 2004).

A final point which warrants discussion is why the level of courtship swimming behaviour was significantly lower in site 3 when compared to sites 1 and 2. We suggest that differences between their habitats hold the key here. Site 3 is characterised by a greater visibility than the other sites, and the females are significantly larger (Paper I, Table 1). Consequently, less time may be required for mate search at this site. Simply, our result may be a consequence of differential allocation to sexual traits (less to a behavioural trait like courtship and more to a morphological trait like body size). This idea could be tested in a mate search tactics experiment comparing the three sites and different morphologies of individuals.

Alternative reproductive investment strategies The aim of paper III was to investigate whether female pipefish adopt alternative reproductive investment strategies depending on their expected lifespan. The pipefish in this study inhabit three different habitats which may be characterised by different risks of mortality. It is, therefore, possible that the expected lifespan of individuals may differ. Here I aimed to determine whether the morphological differences which exist between populations (Paper I) may be a result of differential allocation to, for example, current reproduction (low expected lifespan) or growth towards future reproductive attempts (high expected lifespan).

I experimentally investigated the reproductive behaviour of individual females, from each population, in response to various combinations of environmental stress/mortality risks. Reproductive behaviour was measured in terms of their mating interest towards a potential mate. The females environment was manipulated in terms of presence/absence of a predator (cod), presence/absence of eelgrass, and whether or not they were exposed to three days of food deprivation prior to behavioural recordings. I aimed to mimic those conditions which may invoke a mortality risk response from the fish in their natural environment.

Specific predictions for this study were that females from site 1, which appear to experience a high level of potential mortality in the wild, may be

17

comparatively more willing to take the risk to reproduce as their expected lifespan may be low, i.e. adopt a terminal investment strategy. By contrast, females from site 3, which appear to experience a low level/negligible level of mortality in the wild, may be comparatively less willing to take the risk to reproduce as their expected lifespan may be higher, therefore adopting a cautious strategy (avoiding all potential risk). I, therefore, predicted a gradi-ent response from sites 1, to site 2 (intermediate in mortality risk), to site 3, where females change from adopting a terminal investment through to a cautious strategy, respectively.

Results from this study revealed that, firstly, there are two reproductive strategies (current reproduction or growth towards future reproductive attempts) which describe over 80% of the variance in behaviour of individual females, and secondly, there are behavioural differences between the three sites.

In more detail, analyses revealed a significant difference between the sites in the frequency of dancing (F2,178 = 5.49, p=0.005; Figure 3). Specifically, site 2 females displayed at a lower rate than both site 1 and 2 females.

1 2 3

Site

0

1

2

3

4

5

6

7

8

9

10

11

12

13

Dan

cing

Figure 3. Differences between sites with respect to frequency of displaying behaviour (number of dances), with sites 2 females displaying at a lower frequency compared to the other two sites. Error bars represent 95% confidence intervals.

The analyses with behaviour as a response variable, revealed significant differences between the sites (F2,178 = 6.20, p=0.002). Specifically, when

18

well-fed, females from site 2 actively avoided both the predator and their potential mate, whereas those from sites 1 and 3 avoided neither (one-way ANOVA, difference between sites when fed: F2,104 = 15.04, p<0.001). However, when females were starved these differences disappeared (one-way ANOVA, difference between sites when starved: F2,94 = 0.08, p = 0.920).

Results indicate that females from both sites 1 and 3 adopt what appears to be either a terminal investment strategy or a naïve response to perceived mortality risk, whereas females from site 2 adopt a more cautious, risk-sensitive strategy. I interpret these responses in terms of differing demo-graphic composition among sites and consequently the different breeding environments. As predicted, females from site 1 responded by modifying their behaviour to promote current, rather than possible future, reproduction (terminal investment), i.e. when those females were exposed to risk they actively attempted to attract a mate (similar ‘risk-reckless’ behaviour has been suggested for guppies, Evans et al. 2002). This suggestion is supported by the finding that females from site 1 take high risks when attempting to attract a potential mate. It is possible that a past experience of high mortality risks in site 1 has resulted in individuals being forced to adopt a terminal investment strategy if they have any chance of reproducing. Individuals from site 2 responded cautiously in response to mortality risks. The lower level of mortality risk experienced in the past (in the wild) may have, as predicted, resulted in a preference to wait for less risky future mating opportunities. Interestingly, however, females from site 3 did not respond as predicted. Rather than also adopting a cautious strategy in response to the mortality risks, they adopted a seemingly reckless strategy similar to that observed for females from site 1. One possible explanation for this unexpected result is based on the low predation, the high density of eelgrass (camouflage) and the high food availability which has been described for site 3 (Paper I). This environment could be compared to that on which island theories/studies are based (MacArthur & Wilson 1967; Grant 1998; Schoener et al. 2005). When an individual has not previously been in contact with the mortality risks im-posed in this experiment, that individual may be expected to react in a naïve fashion, similar to species on islands. In other words, what may appear to be risk-reckless behaviour may simply be a naïve response to risk.

This study, therefore, supports the prediction that individuals from different sites adopt alternative reproductive investment strategies depending on both their past and present experience. The alternate investment strategies (namely, current reproduction or growth towards future reproductive attempts) revealed here also provide an explanation for the existing morphological differences between sites.

19

Sequential mate choice Do individuals increase their level of choosiness in consecutive mating encounters? This is predicted by the trade-up hypothesis, and has been proposed to exist in multiply mating species as a means to ensure successful fertilization while securing high-quality mates in subsequent matings (Sheldon 1994; Jennions & Petrie 2000; Preston et al. 2005). The aim of paper IV was to test the trade-up hypothesis. Specifically, we tested the predictions that, firstly, males are indiscriminate during their first mating encounter, and secondly, that their acceptance of a second potential mate depends upon their first mating encounter.

The experiment allowed a focal male to freely interact with a first female for 2.5 hours and then, changing females, with a second female for another 2.5 hours. We carried out four treatments with females of different body sizes exposed in the following orders: small female first/small female second (n=18), large/large (n=19), small/large (n=18), and large/small (n=19).

Firstly, contrary to predicted, males were discriminate during their first mating encounter (significantly higher mating interest in the larger female: MANOVA: F(8,68) = 3.81, p=0.001; Figure 4).

Figure 4. Males spent more time courting if they were exposed to a large (n=38) versus a small (n=36) female during their first mating encounter (p<0.01). The graph shows the mean duration of time the males spent courting the respective female (large or small), and the bars represent 95% confidence intervals.

20

Also, contrary to predicted, after controlling for the reproductive interest of the male and his potential mates, we failed to reveal any interaction between the first female which the male encountered and his corresponding response to the second female encountered (MANCOVA: F(5,62) = 0.59, p=0.709).

Results from this study suggest that male pipefish discriminate during each mate encounter. This behaviour could be interpreted as maladaptive if the consequence is a refusal to mate unless they do encounter a large female. Instead, we suggest that males of this species may have a memory of inhabiting an environment with no scarcity of appropriate mates (previous experience in the wild), or that males have an innate preference which may have evolved over generations with abundant mates. We do not suggest that male pipefish lack the ability to alter their strength of preference, but we do suggest that the multiple mating system in this species may be plastic. Such that, it would be maladaptive for male pipefish to have a fixed strong preference for large females regardless of the environment in which they live.

This study, therefore, strongly suggests that researchers do not generalise the application of the mechanism described by the trade-up hypothesis. Instead, we suggest caution and consideration of alternative mate-sampling tactics in an environment- and species-specific context.

Video playbackThe aim of paper V, was to investigate the application of the video playback technique to studies on mate choice in the pipefish. Specifically, I compared the response of males to video images of females to that of live females.

The experiment consisted of exposing males to three mate choice treatments: live females behind clear glass, enabling visual interactions between sexes, live females behind one-way glass, preventing females from responding to male behaviours, and video playback of females (similar design as used in a guppy study, Kodric-Brown & Nicoletto 1997).

The preference which males exhibited for the larger female differed significantly between treatments (one-way ANOVA: F 2,51 = 3.45, p<0.05). Results revealed that male pipefish showed a stronger preference for the larger female in the video playback treatment than in the clear glass (two-way interaction) or live female treatment. The two latter treatments did not differ.

The video playback treatment confirms previous results of a male preference for large females (Berglund et al. 1986a), and, therefore, demonstrates that the video playback technique is a most powerful method for studying mate preferences in this pipefish.

21

Conclusions

The pattern of phenotypic differentiation without genetic divergence revealed among populations of pipefish (Paper I) has provided the basis for further investigation into local adaptation and phenotypic plasticity. As such, the studies described in papers II and III, suggest that pipefish, both males and females, have the behavioural flexibility to adjust to the environment in which they live. Taken together, these studies support the idea that the fitness of an individual will be greater when it matches the environment in which is occurs via a process of phenotypic fine-tuning (Kingslover et al. 2002).

Plasticity in mate choosiness is another aspect which is discussed in this thesis (Paper IV). The study presented in paper IV revealed that, contrary to predicted, even if a male was unmated he refused to show interest unless the female was ‘large’. This behaviour could be interpreted as maladaptive if the consequence is a refusal to mate unless they do encounter a large female. Instead, paper IV suggests that males of this species may have a memory of inhabiting an environment with no scarcity of appropriate mates (previous experience in the wild), or that males have an innate preference which may have evolved over generations with abundant mates. This study does not suggest that male pipefish do not have the ability to alter their strength of preference, but does suggest that the multiple mating system in this species may be highly plastic. It would be, as mentioned previously, maladaptive for male pipefish to have a fixed strong preference for large females regardless of the environment in which they live. In fact, I suggest that the plasticity in this species which is indicated in papers I, II and III, is a consequence of, for example, their characteristic cryptic phenotype (Vincent et al. 1994) and possible changes in their experienced population density, risk of predation or operational sex ratio over time. Overall, if male pipefish have a strong preference, and that preference lacks plasticity population-wide, reproductive output could be severely reduced (see Kokko & Mappes 2005).

Taken together, papers I, II, III and IV open an exciting avenue for future research to investigate the fitness costs and benefits of plasticity in this sex-role reversed species. Paper V, which demonstrates the appropriateness of the video playback technique, also opens up new areas for research.

22

Sammanfattning på svenska

PopulationsdifferentieringOlika djurpopulationer kan skilja sig åt, genetiskt såväl som utseende- och beteendemässigt, trots att de tillhör samma art. Sådana skillnader kan ha flera orsaker: genetisk drift (slumpmässiga förändringar i framförallt små populationer), olika grad eller typ av naturlig selektion, fenotypisk plasticitet (skillnader i utseende eller beteende orsakade av miljön) samt olika grad eller typ av sexuell selektion. Detta medför att populationerna kan vara lokalt anpassade, antingen genom selektion eller plasticitet eller, vanligen, någon kombination därav.

Beteendens differentiering och ”personligheter” Om populationer skiljer sig utseendemässigt kan vi också förvänta oss bete-endeskillnader dem emellan. Vidare kan vissa beteenden tendera att före-komma tillsammans hos en individ, vilket vi kan kalla för denna individs ”personlighet”. Statistiskt kan en sådan personlighet beskrivas av korrelatio-ner mellan olika beteenden. Självfallet kan också sådana korrelationer skilja mellan olika populationer. Exempelvis kan en personlighet som karakterise-ras av djärvhet kopplat till aggression klara sig utmärkt i en konkurrenspräg-lad miljö med lågt tryck från rovdjur men dåligt i en miljö där rovdjurstryck-et är högt. En viss slags personlighet kan alltså vara välanpassad i en popula-tion men inte i en annan.

Multipla parningar och partnersök Tjänar en individ på att para sig flera gånger, och hur mycket ansträngning skall den lägga på att hitta en högkvalitativ partner? En alltför kräsen individ riskerar att bli helt utan partner, så en möjlig strategi för att undvika detta är att para sig första gången utan att vara kräsen. Detta säkerställer reproduk-tion över huvud taget, men att följande gånger vara betydligt mer noggrann i partnervalet för att säkerställa en så bra reproduktion som möjligt. Denna s.k. ”uppköps”-hypotes relevans bör dessutom bero på partnertillgång, den fara från rovdjur som en parning innebär med flera omvärldsfaktorer. Själv-

23

fallet är kanske denna hypotes inte den enda förklaringen till multipla par-ningar, men justerar individer sin kräsenhet på det predikterade sättet kan hypotesen utgöra en del av en adaptiv förklaring.

SyfteI denna avhandling har jag undersökt olika populationer av tångsnälla (Syngnathus typhle, en kantnålsfisk släkt med sjöhästar) på svenska västkus-ten med avsikt att kartlägga och förklara skillnader mellan närliggande popu-lationer.

Jag har därvid haft som mål att kartlägga följande:

genetisk populationsstruktur utseendemässiga populationsskillnader populationsskillnader i parningsmönster populationsskillnader i beteende och personlighet, samt, framför allt, att förklara sambandet mellan sådana skillnader.

Vidare har jag undersökt orsakerna till ett specifikt beteende, upprepad parning hos såväl hanar som honor, för att se om ”uppköps”-hypotesen kan förklara detta. Slutligen har jag utvärderat en metod för att manipulera bete-endemönster, nämligen video-uppspelning av olika stimuli. Sammantaget vill jag med detta förklara de processer som ligger bakom de populations-skillnader jag upptäckt.

TångsnällanTångsnällan lever i grunda ålgräsängar utefter Europas kuster. Den ställer alla begrepp om hanliga och honliga roller på huvudet: här blir hanen gravid när honan med en ”penis” för över äggen till honom, varpå hanen bär på ungarna några veckor i en yngelficka innan han föder fram dem. Honorna konkurrerar om hanarnas gunst, och när det gäller att välja rätt partner är tångsnällehanarna mer kräsna än honorna, något vi här kallar omvända köns-roller. Honorna kan producera ägg i en raskare takt än vad hanarna kan klara av att ta hand om, vilket är orsaken till de omvända könsrollerna: honorna konkurrerar helt enkelt om en begränsad resurs, nämligen hanarna. Hanar föredrar stora honor som producerar stora, energirika ägg som ger upphov till överlevnadsdugliga ungar. Detta är dock inte hela sanningen: även ho-norna kan vara kräsna i sitt val av hane och väljer, om chansen ges, större fäder till sina ägg. En stor hane kan nämligen ta emot fler ägg. Hanar och honor utvärderar varandra under en uppvaktningsfas, där framförallt honorna visar upp sig för hanarna. Uppstår tycke dansar sedan en hona och en hane

24

tillsammans, ibland i timtal, innan de eventuellt parar sig. En hane behöver oftast para sig flera gånger för att fylla sin yngelficka.

Tångsnällan är långt ifrån en snabbsimmare och tycks mycket bunden till ålgräsängar, där den är perfekt skyddsfärgad och –formad. Då den också saknar planktoniska larver, som kan driva iväg med strömmar, syns potentia-len för utveckling av lokalt anpassade bestånd stor. Därför är tångsnällan till en ideal art för att studera populations-differentiering. De tre populationer jag studerat mellan 2002 och 2005 finns alla inom 3 km från Kristinebergs Marina Forskningsstation, Fiskebäckskil, där också alla experiment utförts.

Resultat och diskussion Den genetiska analysen av populationsskillnader, utförd med DNA-teknik och mikrosatelliter, visade överraskande på inga som helst skillnader mellan populationerna. Troligen har ungarnas eller de vuxnas rörlighet underskattats totalt tidigare, då en avsaknad av genetisk differentiering tyder på ett stort genflöde mellan populationerna. Däremot skilde sig populationerna utseen-demässigt åt: honorna i de tre populationerna var olikstora och hade olika grad av kontrast i sin färgteckning, medan hanar inte skilde sig åt beträffande detta. Miljöerna för de tre populationerna skilde sig åt vad gäller vattendjup, täthet av ålgräs, siktdjup, artrikedom av andra arter och brunalgspåväxt av ålgräset. Jag drar slutsatsen att fenotypisk plasticitet möjliggör lokala an-passningar hos tångsnällorna, och att det kön som konkurrerar om partners och därmed är mer aktivt och risktagande, honorna, troligen har en större nytta av plasticitet än de mindre konkurrensbenägna hanarna. Alternativt kan honorna genom sitt aktiva levnadssätt vara mer utsatta för rovdjur än hanar-na, och ofördelaktiga storlekar och färgteckningar rensas bort genom rovfis-kars och –fåglars försorg. Vad än orsaken är så manar bristen på överens-stämmelse mellan genetiska och utseendemässiga skillnader mellan popula-tionerna till vidare forskning.

Även beteendemässigt fanns skillnader mellan populationerna: i ett expe-riment mätte vi aktivitet, aggression, kräsenhet och grad av uppvaktning i partnerval, samt risktagande inför hot från ett rovdjur, hos hanar såväl som honor från alla tre populationerna. Visserligen skilde sig inte hela personlig-heter åt mellan vare sig populationerna eller mellan könen, d.v.s. olika bete-enden förekom tillsammans i samma mönster hos alla populationer och hos hanar och honor, men enskilda beteenden kunde vara mer eller mindre fram-trädande: individer från en population kännetecknades av en betydligt lägre aktivitet i uppvaktandet av partners än individer från de andra populationer-na. Den avvikande populationen karakteriseras också av stora honor med en färgteckning av låg kontrast. De bebor dessutom en miljö med god sikt. Möj-ligen tjänar honor i denna population mer på att lägga sina resurser på att

25

uppnå en stor kroppsstorlek än på uppvaktings-beteenden och grann färg-teckning.

Ett kanske oväntat resultat var att könen inte skilde sig åt i personlighet. Även om tångsnällan i grunden har omvända könsroller finns dock konkur-rensbeteenden även hos hanar liksom kräsenhet finns hos honor, och denna relativa ömsesidighet i könsroller kan förklara att när hanar och honor sätts i exakt samma experimentella situation så beter de sig inte så olika. Detta betyder inte att deras beteenden i naturen inte skiljer sig åt: här har honor mer att vinna på att konkurrera om partners eftersom lediga hanar utgör en bristvara, och hanar kan kosta på sig att vara kräsna då villiga honor finns i överflöd.

Det är osannolikt att mortaliteten är lika hos de tre populationerna. Om skillnader i mortalitet finns förväntar vi oss skillnader i hur mycket resurser som satsas på omedelbar respektive reproduktion senare i livet. Hög mortali-tet bör gynna satsning på omedelbar reproduktion, låg på att i stället använda resurser tidigt i livet till tillväxt. En större kroppsstorlek ökar attraktivitet inför partners eftersom det indikerar hur många ägg/ungar en individ kan producera. Följaktligen kan individer som tillämpar denna strategi använda resurser senare i livet till en mer framgångsrik reproduktion. I ett experiment undersökte jag om honor från de olika populationerna reagerade olika på stress och mortalitetsrisk vad gäller deras investering i reproduktion, här mätt som tid ägnat åt uppvaktning och dans med hanar. Jag fann att honor från de olika populationerna skilde sig åt: honor från en av populationerna var obenägna att uppvakta samt undvek närhet till en rovfisk, medan honor-na från de andra populationerna var betydligt mer riskbenägna. Det sist-nämnda kan bero antingen på att dessa honor var naiva inför rovdjur och helt enkelt inte betedde sig adaptivt, eller att de inför ett uppenbart hot beslöt att satsa allt på omedelbar reproduktion, oavsett risk.

Ökar kräsenhet i partnerval med antal parningar? ”Uppköps”-hypotesen testades genom att hanar från en population i videoövervakade experiment först fick para sig med en hona (stor eller liten) och därefter med en annan hona (också stor eller liten). Kräsenhet mättes som tid spenderad på upp-vaktning. Vi fann, tvärtemot vad uppköps-hypotesen förutsade, att hanar var kräsna redan vid första parningstillfället (stora honor väckte betydligt större intresse), samt att kräsenhet inte förändrades vid den andra parningen. Upp-köps-hypotesen tycktes alltså inte tillämplig på tångsnälla: hanar hade förut-fattade preferenser som de uttryckte oavsett parningsordning, så orsaken till upprepade parningar hos denna art får sökas annorstädes.

Jag har i mina experiment använt videoövervakning för att kunna analyse-ra beteenden i detalj. Detta kan tas ett steg längre genom att använda upp-spelning av stimuli via video för att manipulera studieobjektets upplevda omgivning. Innan detta görs bör man dock studera om videouppspelade filmsekvenser uppfattas som relevanta stimuli, lämpligen genom att jämföra samma individers reaktioner på verkliga stimuli med reaktioner på samma

26

stimuli återgivna via video. Detta har jag gjort i ett experiment där en hane fick välja mellan två olikstora honor. Honorna presenterades antingen bakom en glasvägg där hanen och honorna kunde se varandra, bakom en envägs-spegel där endast hanen kunde se honorna, eller som videofilm där samma hona presenterades i liten och stor upplaga. Starkast preferens för den större honan erhölls i videopresentationen, vilket alltså visar teknikens användbar-het. Fördelarna med denna teknik är enkelheten att manipulera stimuli, här storlek, medan allt annat, som beteende, färg, etc., hålls konstant då samma hona kan användas.

Sammanfattningsvis har jag visat att såväl honor som hanar hos tångsnäl-la besitter en betydande fenotypisk flexibilitet vad gäller storlek, färgteck-ning och beteende, som trots avsaknaden av genetiska skillnader mellan olika populationer tillåter lokala anpassningar hos dessa populationer. Sär-skilt honor tycks flexibla och möjligen gör deras aktivare, mer konkurrensin-riktade och risktagande levnadssätt att de i högre grad än hanar drar nytta av en stor fenotypisk plasticitet. Denna avhandling visar också på nödvändighe-ten av mer forskning över fördelar och kostnader med fenotypisk plasticitet.

27

Acknowledgements

Firstly, my supervisor, Anders Berglund. You have provided me with the tools to become confident with my research. The unwavering encouragement and support, on both a work and personal level, you have given me has re-sulted in this thesis, thank you. Gunilla Rosenqvist, thank you for being you, and being a great friend and colleague. Göran Arnqvist, a special thanks for the jokes, scientific discussions and the statistical advice.

From both before and during my PhD, I want to thank Ian Owens and Mark Blows. It really goes without saying that both of you have inspired me to enter and stay in the field of research. Your encouragement means more to me that you can know. Thanks for fantastic scientific discussions. Mark, thanks for the Friday beers and ‘tissue box reach’.

The pipefish group, thank you. Firstly, the field assistants, especially, a big thank you to Tanja, Camilla, and Ronny: your help and friendship has been invaluable. Ingrid, thank you for reading through my manuscripts. Do-minique, thanks for the encouragement and getting me through that ‘tough’ first field season and the following winter. Anna, thanks for being a great friend both during and outside of the field season. You are a wonderful per-son.

All the people here at the department thank you for making this depart-ment a great place to work. Especially, Anna Q. (for proof-reading most of my manuscripts), Niclas (no more IOU), Nina, Ted, Mari, Joanna R., Olivia, Alex, Urban, Johanna A., Joanna S., Emma, Olivia, Marnie, Mats, Lars and Ingela. Claudia, over the past four years we have become best of friends. Thank you for your friendship and support – I will miss you more than you know. A very special thanks to Katherine, Mårten, Chris and Sara who have been fantastic people to share an office with: the light-hearted fun chats and the great scientific discussions have been invaluable. Sara, thanks for being a supportive friend and for reading through my work at the very end. Kathe-rine, you have known me right from the start, hmmm…too many things to say, so THANK YOU, your friendship kept me here. As promised, a thanks to Damian for keeping me in line. Also, to the other three Australians (Damian, Chris and Sandra) in the department, it is great to have you here – a refreshing sarcastic sense of humour is the best way to get through the day.

28

Back in Australia, Susan, Yvette and Katrina, thank you for standing by me even though my communication has wavered at times. I miss you all and cannot wait to live near you again, hopefully, soon.

Tash, I will never forget your encouragement and support when I have needed it the most. I am so happy to have you as a very special friend.

Throughout my PhD time I gained another very special family all of who have been invaluable support for me while I have been living and studying here in Sweden. Most influential, thank you to Birgitta and Carl Johan for the unwavering and unconditional support (even though my pipefish have no clear economic value). Anders, your light-heartedness has also provided me with the ‘time-out’ to have a good laugh! You are a very special friend. Jonas, what can I say. You were there when Kram and I first met, you were there within hours of Lilja’s birth, and overall, you are one of my best friends here in Sweden, and finally, thank you for the invite to join you in eating cookies, sorry I could not join you!

Grandma and grandpa, you have never ceased being a source of encour-agement, love and support for me. You both have a very special place in my heart. I think of you often and will always love you both so very much.

A HUGE thanks to poppa, dad and Anne for making the trip all the way over here to see me cross this milestone, it means a lot to me. I hope you gain some insight into what I have been doing over here in Sweden for the last 4.5 years. Dad, thank you for your ongoing support – our friendship means more to me than you know.

Alex, Katie and Clare – Each of you put a HUGE smile on my face. You make me so happy. Thank you.

JJ, you are my source of inspiration in so many ways. I am who I am to-day, largely thanks to you! Thank you for your encouragement, support, and friendship – you are one of my best friends.

Lilja, you are sunshine in my life and make me laugh everyday. You put my life in wonderful perspective. Kram, words cannot possibly come close to thanking and acknowledging you for what you have helped me achieve in life. You give me butterflies in my stomach and put a smile on my face. Jag älskar dig / I love you!

29

References

Arendt, J. D. & Reznick, D. N. 2005. Evolution of juvenile growth rates in female guppies (Poecilia reticulata): predator regime or resource level? Proceedings of the Royal Society of London Series B: Biological Sciences, 272, 333-337.

Bell, A. M. 2005. Behavioural differences between individuals and two populations of stickleback (Gasterosteus aculeatus). Journal of Evolutionary Biology, 18, 464-473.

Bell, A. M. & Stamps, J. A. 2004. Development of behavioural differences between individuals of populations of sticklebacks, Gasterosteus aculeatus. Animal Behaviour, 68, 1339-1348.

Berglund, A. & Rosenqvist, G. 1990. Male limitation of female reproductive success in a pipefish effects body-size differences. Behavioral Ecology and Sociobiology, 27, 129-134.

Berglund, A. & Rosenqvist, G. 2003. Sex role reversal in pipefish. Advances in the Study of Behavior, 32, 131-167.

Berglund, A., Rosenqvist, G. & Svensson, I. 1986a. Mate choice, fecundity and sexual dimorphism in two pipefish species (Syngnathidae). Behavioral Ecology and Sociobiology, 19,301-307.

Berglund, A., Rosenqvist, G. & Svensson, I. 1986b. Rversed sex roles and parental energy investment in zygotes of two piepfish (Sygnathidae) species. Marine Ecology Progress Series, 29,209-215.

Berglund, A., Rosenqvist, G. & Svensson, I. 1989. Reproductive success of females limited by males in two pipefish species. American Naturalist, 133, 506-516.

Berglund, A., Sandvik Widemo, M. & Rosenqvist, G. 2005. Sex role reversal revisited: choosy females and ornamented, competitive males in a pipefish. Behavioural Ecology, 16,649-655.

Borevitz, J. O., Maloof, N., Lutes, J., Dabi, T., Redfern, J. L. & Trainer, G. T. 2002. Quantitative trait loci controlling light and hormone response in twoaccessions of Arabidopsis thaliana .Genetics, 160, 683-696.

30

Borgerhoff Mulder, M. 2004. Are men and women really so different? Trends in Ecology and Evolution, 19, 3-6.

de Jong, M. C. M. & Sabelis, M. W. 1991. Limits to runaway sexual selection: the wallflower paradox. Journal of Evolutionary Biology, 4, 637-655.

DeWitt, T. J., Sih, A. & Sloan Wilson, D. 1998. Costs and limits of phenotypic plasticity. Trends in Ecology and Evolution, 13,77-81.

Dingemanse, N. & Reale, D. 2005. Natural selection and animal personality. Behaviour, 142, 1165-1190.

Evans, J. P., Kelley, J. L., Ramnarine, I. W. & Pilastro, A. 2002. Female behaviour mediated male courtship under predation risk in the guppy (Poecilia reticulata). Behavioral Ecology and Sociobiology, 52, 496-502.

Grant, P. R. 1998. Evolution on Islands New York: Oxford University Press.

Hedrick, A. V. 2000. Crickets with extravagant mating songs compensate for predation risk with extra caution. Proceedings of the Royal Society of London Series B: Biological Sciences, 267, 671-675.

Jennions, M. D. & Petrie, M. 1997. Variation in mate choice and mating preferences: a review of causes and consequences. Biological Review, 72, 283-327.

Jennions, M. D. & Petrie, M. 2000. Why do females mate multiply? A review of the genetic benefits. Biological Review, 75, 21-64.

Johnstone, R. A., Reynolds, J. D. & Deutsch, J. C. 1996. Mutual mate choice and sex differences in choosiness. Evolution; International Journal of Organic Evolution, 50, 1382-1391.

Kingslover, J. G., Pfenning, D. W. & Servedio, M. R. 2002. Migration, local adaptation and the evolution of plasticity. Trends in Ecology and Evolution (Amsterdam), 17, 540-541.

Kodric-Brown, A. & Nicoletto, P. F. 1997. Repeatability of female choice in the guppy: response to live and videotaped males. AnimalBehaviour, 54, 369-376.

Kokko, H. & Mappes, J. 2005. Sexual selection when fertilisation is not guaranteed. Evolution; International Journal of Organic Evolution, 59, 1876-1885.

Kolm, N. & Berglund, A. 2004. Sex-specific territorial behaviour in the Banggai cardinalfish, Pterapogon kaunderni. EnvironmentalBiology of Fishes, 70, 375-379.

MacArthur, R. H. & Wilson, E. O. 1967. The Theory of Island Biogeography Princeton, NJ: Princeton University.

Mackay, T. F. C. 2001. The genetic architecture of quantitative traits. Annual Review of Genetics, 35, 303-339.

31

Neff, B. D. & Sherman, P. W. 2004. Behavioral syndromes versus darwinian algorithms. Trends in Ecology and Evolution, 19,621-622.

Piersma, T. & Drent, J. 2003. Phenotypic flexibility and the evolution of organismal design. Trends in Ecology and Evolution (Amsterdam), 18, 228-233.

Pomiankowski, A. N. 1987. The costs of choice in sexual selection. Journalof Theoretical Biology, 128, 195-218.

Preston, B. T., Stevenson, I. R., Pemberton, J. M., Coltman, D. W. & Wilson, K. 2005. Male mate choice influences female promiscuity in Soay sheep. Proceedings of the Royal Society of London Series B: Biological Sciences, 272, 365-373.

Rogers, S. M., Gagnon, V. & Bernatchez, L. 2002. Genetically based phenotype-environment association for swimming behaviour in lake whitefish ecotypes (Coregonus clupeaformis Mitchill). Evolution; International Journal of Organic Evolution, 56,2322-2329.

Sandvik, M., Rosenqvist, G. & Berglund, A. 2000. Male and female mate choice affects offspring quality in a sex-role reversed pipefish. Proceedings of the Royal Society of London Series B: Biological Sciences, 267, 2151-2155.

Sandvik Widemo, M. 2006. Male but not female pipefish copy mate choice. Behavioral Ecology, 17, 255-259.

Schoener, T. W., Losos, J. B. & Spiller, D. A. 2005. Island biogeography of populations: an introduced species transforms survival pat-terns. Science, 310, 1807-1809.

Sgro, C. M. & Hoffmann, A. A. 2004. Genetic correlations, tradeoffs and environmental variation. Heredity, 93, 241-248.

Sheldon, B. C. 1994. Male phenotype, fertility, and the pursuit of extra-pair copulations by female birds. Proceedings of the Royal Society of London Series B: Biological Sciences, 257, 25-30.

Sih, A., Bell, A. & Chadwick Johnson, J. 2004a. Behavioural syndromes: and ecological and evolutionary overview. Trends in Ecology and Evolution, 19, 372-378.

Sih, A., Bell, A. M. & Chadwick Johnson, J. 2004b. Reply to Neff and Sherman. Behavioral syndromes versus darwinian algorithms. Trends in Ecology and Evolution, 19, 622-623.

Sih, A., Bell, A. M., Chandwick Johnson, J. & Ziemba, R. E. 2004c. Behavioural syndromes: and integrative overview. QuarterlyReview of Biology, 79, 241-277.

Trivers, R. 1972. Parental investment and sexual selection. In: SexualSelection and the Descent of Man 1871-1971 (Ed. by B. Campbell), pp. 139-179. Aldine Press.

32

Via, S., Gomulkiewicz, R., De Jong, G., Scheiner, S. M., Schlichting, C. D. & Van Tienderen, P. H. 1995. Adaptive phenotypic plasticity: consensus and controversy. Trends in Ecology and Evolution, 10, 212-217.

Viera, C., Pasyukova, E. G., Zeng, Z. B., Hackett, J. B., Lyman, R. F. & Mackay, T. F. C. 2000. Genotype-environment interaction for quantitative trait loci affecting lifespan in Drosophila melanogaster. Genetics, 154, 213-227.

Vincent, A., Ahnesjö, I. & Berglund, A. 1994. Operational sex ratios and behavioural sex differences in a pipefish population. Behavioral Ecology and Sociobiology, 34, 435-442.

Vincent, A. C. J., Berglund, A. & Ahnesjö, I. 1995. Reproductive ecology of five pipefish species in one eelgrass meadow. Environmental Biology of Fishes, 44, 347-361.

Wade, M. J. & Shuster, S. M. 2002. The evolution of parental care in the context of sexual selection: a critical reassessment of parental investment theory. American Naturalist, 160, 285-292.

Acta Universitatis UpsaliensisDigital Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology 157

Editor: The Dean of the Faculty of Science and Technology

A doctoral dissertation from the Faculty of Science andTechnology, Uppsala University, is usually a summary of anumber of papers. A few copies of the complete dissertationare kept at major Swedish research libraries, while thesummary alone is distributed internationally through theseries Digital Comprehensive Summaries of UppsalaDissertations from the Faculty of Science and Technology.(Prior to January, 2005, the series was published under thetitle “Comprehensive Summaries of Uppsala Dissertationsfrom the Faculty of Science and Technology”.)

Distribution: publications.uu.seurn:nbn:se:uu:diva-6666

ACTAUNIVERSITATISUPSALIENSISUPPSALA2006