Search

This thesis examines a particular selfish genetic element (SGE), X-chromosome meiotic drive (XCMD), in the species Drosophila subobscura. XCMD is a system where the X-chromosome kills or disables Y-chromosome sperm to enhance their own transmission to the next generation. This also results in those males producing female biased broods. This selfish enhancement of their own transmission results in conflict with the rest of the genome that can be a potent force in evolution. The first chapters deal with sex and mating behaviour and how XCMD and other SGEs are linked to it. Chapter three focusses on the marking techniques and mating behaviour in three species of Drosophila. This work was completed while establishing the XCMD system from wild populations Chapter four presents case studies of how SGEs are intrinsically linked to sex. Chapter five examines XCMD in D. subobscura and reveals that this species is completely monandrous. This shows that polyandry does not play a role in preventing the spread of XCMD in this species, unlike in a number of other taxa which have XCMD. I also demonstrate weak female choice against XCMD in this chapter. In chapter six and seven I examine the XCMD phenotype when it is expressed in different population genetic backgrounds. I test for evidence of suppression and incompatibilities, when XCMD is exposed in four different populations (Tunisia, Morocco, Spain, and UK). I find evidence of suppression in North Africa, but no suppression in Europe. I also find evidence for severe incompatibilities specific to XCMD on European genetic backgrounds, which are absent in North African backgrounds. These results are consistent with genetic conflict causing rapid evolution in North Africa between XCMD and suppressors, which results in XCMD specific hybrid incompatibilities in naïve European populations. My final chapter evaluates how the testes proteomes of two species, D. subobscura and D. pseudoobscura, differ between XCMD and non-XCMD individuals. This ongoing work identifies some putative candidate genes that could be involved in the network that results in XCDM in these species. Interestingly, very few strong candidate genes overlapped in the two species, supporting the idea that separate genes and mechanisms are responsible for the two XCMD systems.

Coevolutionary interactions change across landscapes, leading to the formation of geographic selection mosaics. Analyses of coevolutionary dynamics have so far focused on interactions between two trophic levels. The tritrophic interactions between the pierid butterfly Anthocharis cardamines (the orange-tip), its brassicaceous host-plants Cardamine pratensis (lady's smock) and Alliaria petiolata (garlic mustard), and primary parasitoids Phryxe vulgaris (Tachinidae) and Cotesia saltator (Braconidae), are known to vary across mainland Britain and Continental Europe. In Britain, northern A. cardamines populations tend to utilize C. pratensis and southern ones A. petiolata; the primary parsitoid in this country is P. vulgaris. In Sweden, the butterfly does not show a strong preference for any brassicaceous hostplant, and the primary parasitoid is C. saltator. In this thesis, I investigate likely selection pressures operating on these interactions in a single study site on the Wirral peninsula in northern England. Host use affects the emergence timing and dispersal of A. cardamines males; specimens utilizing C. pratensis emerge earlier than those utilizing A. petiolata, and are also smaller. Since small size is linked with depressed dispersal, utilization of C. pratensis results in an "emerge early and wait" mate seeking strategy; conversely, utilization of A. petiolata is associated with an "emerge late and rove" strategy. These alternative strategies are likely adapted to the varying density of locally emerging females across the landscape, with high density populations tending to be associated with C. pratensis and low density ones with A. petiolata. In dense populations, late emerging males will be at a disadvantage since females mate only once; in the study population, their predicted fitness always declines to < 1 in late season. This is coupled with a 'stay-or-go' response, in which a proportion of late emerging males immediately emigrate to a low density continuum outside the study area, where early emergence is less critical for fitness. Such a response could help maintain sink populations by averting Allee effects (decreased population growth due to low mate encounter rates). Late instar A. cardamines larvae are heavily parasitized by P. vulgaris. This could select for early vacation of the host-plant at small larval size, whereas the size-fecundity relationship in females should select for prolonged growth to larger size. Mathematical models indicate that high rates of parasitization are sufficient to overturn the fecundity benefit of large size, but fall short of maintaining strong stabilizing selection for an optimal winglength. The tendency of some larvae to move off their host-plant before the final instar resting phase is probably a direct evolutionary response to parasitism risk. A. cardamines larvae are pre-dispersal seed-predators; an early flowering ecotype of C. pratensis has likely been selected to avoid egg-laying A. cardamines females. In turn, the butterfly appears to be invading this host-plant's phenological space, with selection favouring small, early emerging females which oviposit on it. This cautions against interpreting the recent advance in A. cardamines' phenology solely in terms of a response to climate change. Spatio-temporal variation in the intensity of these effects likely contributes to the ongoing coevolutionary dynamics within this tritrophic system.