The three-spined stickleback (Gasterosteus aculeatus) has repeatedly and independently adapted to freshwater habitats from standing genetic variation (SGV) following colonization from the sea. However, in the Mediterranean Sea G. aculeatus is believed to have gone extinct, and thus the spread of locally adapted alleles between different freshwater populations via the sea since then has been highly unlikely. This is expected to limit parallel evolution, that is the extent to which phylogenetically related alleles can be shared among independently colonized freshwater populations. Using whole genome and 2b-RAD sequencing data, we compared levels of genetic differentiation and genetic parallelism of 15 Adriatic stickleback populations to 19 Pacific, Atlantic and Caspian populations, where gene flow between freshwater populations across extant marine populations is still possible. Our findings support previous studies suggesting that Adriatic populations are highly differentiated (average F-ST approximate to 0.45), of low genetic diversity and connectivity, and likely to stem from multiple independent colonizations during the Pleistocene. Linkage disequilibrium network analyses in combination with linear mixed models nevertheless revealed several parallel marine-freshwater differentiated genomic regions, although still not to the extent observed elsewhere in the world. We hypothesize that current levels of genetic parallelism in the Adriatic lineages are a relic of freshwater adaptation from SGV prior to the extinction of marine sticklebacks in the Mediterranean that has persisted despite substantial genetic drift experienced by the Adriatic stickleback isolates.

Adaptive radiation is an important mechanism of evolution, which can lead to emergence of sympatric species or morphotypes. Among other biological interactions, parasitic pressure can have significant evolutionary implications for host populations by reducing the fitness of the host individuals. Parasite community structure of fishes is typically strongly dependent on both host ecology (e.g. habitat and feeding behaviour) and environmental factors (e.g. water quality and temperature). However, the relative importance of these factors for parasite-mediated speciation is not known. Also, host gender-specific differences in parasite communities can have an effect on the differentiation of host morphs. In this Master's thesis, I focused on differences in parasite communities of sympatric morphs of three-spine stickleback (Gasterosteus aculeatus) in two large Icelandic lakes, Thingvallavatn and Mývatn. In these lakes, the habitats of sympatric mud and lava morphs are the same (soft/hard bottom), but the habitat water temperatures are opposite between the lakes. In this unique system, it is thus possible to compare between the effects of host ecology and water temperature on parasite community structure and strength of parasite-mediated selection. Additionally, I studied the effect of host sex on the parasitic pressure. There is also a third stickleback morph, Nitella morph, inhabiting the cold limnetic habitats in Lake Thingvallavatn. I sampled the stickleback morphs from both lakes and identified their parasite species. I discovered a total of five parasite genera: trematodes Apatemon and Diplostomum, and cestodes Diphyllobothrium, Proteocephalus and Schistocephalus. Most of the observed parasites have negative effects on health and fitness of the host. I found more parasites in sticklebacks living in higher temperature in both lakes regardless of the morph. When comparing the cold water morphs in Thingvallavatn, parasite abundance was higher in the limnetic Nitella morph than in the shallow water lava morph. Fish gender had an effect on parasitism only in Thingvallavatn as males of both lava and Nitella morphs were more heavily infected with cestodes. Similarities in parasite communities with water temperature indicate that water temperature mainly determines parasite infections in this system instead of host ecology. As similar fish morphotypes exist in different lakes under opposite parasitic pressure, parasitism has unlikely initiated host differentiation, but differences in infection probably have emerged secondary to the ecological specialization of the morphs to different habitats. These results are among the first to tackle the key question in parasite-mediated divergent selection: at which point of the speciation process parasite communities become differentiated and thus can have an effect on speciation. However, the comparison between the cold water morphs (lava and Nitella) indicates that although water temperature seems to be the main factor controlling infections in this system, its effect may still be over ridden by host ecology. Sex-depended differences in parasitic pressure, on the other hand, are likely to reflect specific characteristics of each fish population and lake. These results suggest complex interactions between host ecology and abiotic environment, such as water temperature, in determining the parasite community structure. Hence both factors have to take into consideration when studying the role of parasites in speciation processes. In future, it is necessary to pinpoint the stage of the host speciation process when parasite infections become differentiated in replicated systems to gain comprehensive understanding of the role of parasites in adaptive radiations.

Experiments are often carried out in the laboratory under arti®cial conditions. Although this can control for confounding factors, it may eliminate important factors that under natural conditions mediate the interaction under investigation. Here, we show that different results can be gained in the ®eld and in the laboratory regarding host±parasite interaction. In the ®eld, courting three-spined stickleback males, Gasterosteus aculeatus, were less often infected with plerocercoids of a cestode tapeworm, Schistocephalus solidus, than shoaling males. However, when a random sample of males was allowed to nest and court females in individual aquaria in the laboratory, both uninfected and infected males built nests and courted females. Moreover, while the few infected males that courted females in the ®eld expressed less red nuptial coloration than uninfected courting males, there was no difference in redness between infected and uninfected males in the laboratory. We argue that the different results gained in the ®eld and in the laboratory are due to differences in the cost of reproduction, due to differences in the resource pool of the males. The favourable conditions in the laboratory exclude factors such as predation risk, social interactions, and ¯uctuating environmental conditions that may use up resources in the ®eld and mediate the effect of the parasite.

The three-spined stickleback (Gasterosteus aculeatus) is a model organism for studies of parallel evolution in the wild; marine stickleback populations have repeatedly colonized and successfully adapted to different brackish and freshwater habitats. During Pleistocene glaciations, three-spined stickleback populations inhabiting high-latitude areas of Europe were eradicated, whereas populations residing in (or moving to) the south persisted in refugia. After the retreat of the ice sheets covering northern Europe, the high-latitude areas became recolonized by migration from south, and hence, today’s northern European populations are relatively young. Population genetic studies of European three-spined sticklebacks have usually been conducted at high-latitude areas where freshwater populations are typically less than 10.000 years old. Few studies have focused on southern populations, where more of the ancestral diversity is likely to reside. These studies have utilized a limited number of microsatellite markers and mtDNA sequence fragments, whereas studies of southern populations focusing on genome-wide diversity, in particular from the edge of the southern distribution limit in the Iberian Peninsula, are still missing. Here, I wanted to cover this gap in knowledge by carrying out an empirical and statistical study with RAD-seq data from southern and northern European populations of three-spined stickleback. The main aims of this study were two-fold. First, to investigate whether the southern European freshwater populations of the three-spined stickleback – which currently lack or have limited connection to ancestral marine populations carrying most of the standing genetic variation (SGV) – have lost genetic diversity due to population bottlenecks and inbreeding as compared to their northern European counterparts. Second, to compare the degree of genetic parallelism in southern vs. northern European populations in genomic regions which have been shown to be consistently associated with freshwater colonization in earlier studies. Under the assumption that the lack of continued access to SGV in the ancestral marine population reduces the likelihood of parallel evolution, I hypothesized that the degree of genetic parallelism in genomic regions subject to positive selection in freshwater environments is lower in the southern than in northern European populations. However, if a reduction in genetic diversity and/or cessation of gene flow between southern European freshwater and marine populations occurred following freshwater adaptation, the opposite pattern could be expected. I paid particular attention to chromosomal inversions associated with marine-freshwater adaptations identified in previous studies. The results confirmed my expectation of reduced genetic diversity in southern as compared to northern European stickleback populations. On the other hand, and contrary to what I expected, analysis of clusters of global parallelism involved in freshwater adaptation revealed that southern European populations exhibit a higher degree of genetic parallelism in response to freshwater colonisation than those from northern Europe. This suggests that the loss of genetic diversity in southern populations has occurred after they had adapted to freshwater environments, explaining the high degree of genetic parallelism in spite of the current low levels of genetic diversity. In addition, it could be that selection pressures in south are more homogenous than in north, which would also explain the higher degree of genetic parallelism observed in southern Europe. The findings presented here, challenge the current paradigm that parallel evolution is unlikely in populations with low genetic diversity and that have experienced recent bottlenecks.

Conspecifics inhabiting divergent environments frequently differ in morphology, physiology, and performance, but the interrelationships amongst traits and with Darwinian fitness remains poorly understood. We investigated population differentiation in morphology, metabolic rate, and swimming performance in three-spined sticklebacks (Gasterosteus aculeatus L.), contrasting a marine/ancestral population with two distinct freshwater morphotypes derived from it: the typical low-plated morph, and a unique small-plated morph. We test the hypothesis that similar to plate loss in other freshwater populations, reduction in lateral plate size also evolved in response to selection. Additionally, we test how morphology, physiology, and performance have evolved in concert as a response to differences in selection between marine and freshwater environments. We raised pure-bred second-generation fish originating from three populations and quantified their lateral plate coverage, burst- and critical swimming speeds, as well as standard and active metabolic rates. Using a multivariate Q(ST)-F-ST framework, we detected signals of directional selection on metabolic physiology and lateral plate coverage, notably demonstrating that selection is responsible for the reduction in lateral plate coverage in a small-plated stickleback population. We also uncovered signals of multivariate selection amongst all bivariate trait combinations except the two metrics of swimming performance. Divergence between the freshwater and marine populations exceeded neutral expectation in morphology and in most physiological and performance traits, indicating that adaptation to freshwater habitats has occurred, but through different combinations of traits in different populations. These results highlight both the complex interplay between morphology, physiology and performance in local adaptation, and a framework for their investigation.

Planktivorous fish can exert strong top-down control on zooplankton communities. By incorporating different feeding strategies, from selective particulate feeding to cruising filter feeding, fish species target distinct prey. In this study, we investigated the effects of two species with different feeding strategies, the three-spined stickleback (Gasterosteus aculeatus (L.)) and roach (Rutilus rutilus (L.)), on a low-diversity brackish water zooplankton community using a 16-day mesocosm experiment. The experiment was conducted on a small-bodied spring zooplankton community in high-nutrient conditions, as well as a large-bodied summer community in low-nutrient conditions. Effects were highly dependent on the initial zooplankton community structure and hence seasonal variation. In a small-bodied community with high predation pressure and no dispersal or migration, the selective particulate-feeding stickleback depleted the zooplankton community and decreased its diversity more radically than the cruising filter-feeding roach. Cladocerans rather than copepods were efficiently removed by predation, and their removal caused altered patterns in rotifer abundance. In a large-bodied summer community with initial high taxonomic and functional diversity, predation pressure was lower and resource availability was high for omnivorous crustaceans preying on other zooplankton. In this community, predation maintained diversity, regardless of predator species. During both experimental periods, predation influenced the competitive relationship between the dominant calanoid copepods, and altered species composition and size structure of the zooplankton community. Changes also occurred to an extent at the level of nontarget prey, such as microzooplankton and rotifers, emphasizing the importance of subtle predation effects. We discuss our results in the context of the adaptive foraging mechanism and relate them to the natural littoral community.

Snapshot analyses have demonstrated dramatic intraspecific variation in the degree of brain sexual size dimorphism (SSD). Although brain SSDis believed to be generated by the sex-specific cognitive demands of reproduction, the relative roles of developmental and population-specific contributions to variation in brain SSD remain little studied. Using a common garden experiment, we tested for sex-specific changes in brain anatomy over the breeding cycle in three-spined stickleback (Gasterosteus aculeatus) sampled from four locations in northern Europe. We found that the male brain increased in size (ca. 24%) significantly more than the female brain towards breeding, and that the resulting brain SSD was similar (ca. 20%) for all populations over the breeding cycle. Our findings support the notion that the stickleback brain is highly plastic and changes over the breeding cycle, especially in males, likely as an adaptive response to the cognitive demands of reproduction (e.g. nest construction and parental care). The results also provide evidence to suggest that breeding-related changes in brain size may be the reason for the widely varying estimates of brain SSD across studies of this species, cautioning against interpreting brain size measurements from a single time point as fixed/static.