Marine and freshwater environments support numerous species of teleost fish with a wide and diverse range of reproductive strategies. Despite the considerable interest in fish reproduction, our knowledge regarding ejaculate traits and factors affecting them is limited. Using computer-assisted sperm analysis (CASA) I measured ejaculate traits (sperm swimming speed, motility, path trajectory, longevity and concentration) from sexually mature chinook salmon (Oncorhynchus tshawytscha) activated in freshwater and ovarian fluid. I also looked at these ejaculate traits in relation to measures of male quality (body condition) and investment into reproduction (relative testes mass). Furthermore, I determined the chemical composition of seminal and ovarian fluid and looked at the effect these fluids have on sperm behaviour. A considerable amount of intraspecific variation existed in all ejaculate traits measured, and investment into reproduction (relative testes mass) was dependent on male body condition, as males in better condition had relatively larger testes. However, these males did not have superior quality ejaculates or ejaculates with a higher density of spermatozoa; hence the potential reproductive advantage of having larger relative testes in this species remains unknown and requires further investigation. In addition, a positive relation between sperm longevity and sperm swimming speed was observed defying the expected trade-off between ejaculate traits according to theory. There was also a weak negative trend in our data between body condition and sperm swimming speed, linearity, and longevity. All sperm traits measured were greatly enhanced when activated in a solution containing ovarian fluid (a viscous fluid which is excreted with the egg batch during spawning) from female chinook salmon. Interestingly, sperm swimming speed activated in fresh water only accounted for < 12% of the observed variation in mean sperm swimming speed in ovarian fluid. This result suggests the sperm traits measured in fresh water are not relevant to those same traits measured in ovarian fluid, so caution should be applied when comparing the potential for individual males to fertilize ova when sperm traits are activated in water, especially in studies of sperm competition in an externally fertilising species. Sperm competition between males is known to strongly influence sperm and ejaculate traits, but less is known about female sperm choice after copulation via a process called cryptic female choice (CFC). In CFC, females may have the ability to favour the sperm of one male over another and bias fertilisation accordingly. To test whether ovarian fluid could act as a mechanism of CFC in an externally fertilising fish species, I measured sperm traits from each male activated in the ovarian fluids from different females. I found that mean sperm swimming speed, longevity, and path trajectory differed significantly among males, but most importantly, the pattern of within-male variation in these traits also varied significantly among males in response to different females’ ovarian fluids. This result suggests that ovarian fluid may be a potential mechanism of CFC whereby females differentially enhance the swimming speed of sperm from different males. In addition, I found that sperm longevity was negatively correlated with variation in [Ca²⁺] and [Mg²⁺] concentration in the ovarian fluid, while percent motility increased with increasing concentration of [Mg²⁺]. These observations provide a possible chemical basis for cryptic female choice whereby female ovarian fluid differentially influences the behaviour of sperm from different males and thus their fertilisation success. This finding is particularly exciting, as we may have uncovered a potential mechanism of CFC in an externally fertilising species, which is poorly understood. In addition, results from this study suggest new directions for genetic studies to provide direct evidence for CFC. For example, does sperm selection via ovarian fluid promote favoured genetic combinations that enhance male reproductive success?

Reproduction in vertebrates, including fishes, is under control of the brain-pituitary-gonad (BPG) axis. The female three-spined stickleback, Gasterosteus aculeatus, produces egg clutches at intervals of a few days and spawns them in a nest built by male. Following ovulation, eggs are stored in the ovarian cavity surrounded by the ovarian fluid (OF). If spawning or spontaneous release do not occur, the eggs can undergo overripening, a phenomenon occurring both in nature and captivity. In this PhD thesis, the changes of reproductive hormones and vitellogenesis were studied at overripening of eggs and over the natural spawning cycle. OF properties were also examined at overripening of eggs and after treatment with sex steroids. Plasma levels of steroids: testosterone (T), estradiol (E2), 17,20β-dihydroxypregn-4-en-3-one (17,20β-P) and 17,20β,21-trihydroxypregn-4-en-3-one (17,20β,21-P) were measured by radioimmunoassay, and relative mRNA levels of the pituitary gonadotropins (fsh-β/lh-β), brain gonadotropin-releasing hormones (gnrh2/gnrh3) and kisspeptin and its receptor (kiss2/gpr54) by qPCR. Overripening of eggs was accompanied with a significant reduction in most of endocrine parameters of BPG axis (T, E2, 17,20β-P; lh-β; kiss2, gpr54). Low level of hormones could be advantageous for the overripe egg-bound females, since this would reduce further ovulations giving higher chances to survive and reproduce again. Over the 3-day spawning cycle, T and E2 were highly correlated, showed cyclicity with low levels at ovulation and increasing from 24 and 6 hours post-spawning (hps), respectively. Spawning may give rise to this increase as these rises did not occur if release of the eggs does not happen (overripe females). A peak at pituitary lh-β mRNA levels appeared 48 hps, a day before the next ovulation. No significant changes were found for the other studied hormones. Vitellogenesis was studied by measurement of the vitellogenin mRNA levels in the liver by qPCR. The levels were highest at 24 and 48 hps and were positively correlated to both E2 and T over the cycle. However, changes were small suggesting a rather continuous vitellogenesis over the stickleback spawning cycle which could be an advantage for a multiple spawner with a limited spawning season. Overripening reduced vitellogenin mRNA levels but did not abolish it. OF amount was diminished in overripe females and had a lower viscocity but higher dry weight and protein levels than in non-overripe ovulated females, suggesting that changes in OF properties are related to the egg overripening. The effects of steroids were studied using Silastic capsules. T and 17,20β-P induced an increase of OF amount, but protein levels were only increased in 17,20β-P-treated females, proposing a role of this steroid in the control of OF secretion. 1-D SDS-PAGE showed that OF contained several proteins, some of them came from eggs, but no consistent differences between groups. Concluding, the knowledge of the reproductive physiological changes is important for understanding their essential roles in the production of viable eggs in this species but also in the reproductive physiology of female fishes in general. / <p>At the time of the doctoral defense, the following papers were unpublished and had a status as follows: Paper 3: Manuscript. Paper 4: Manuscript.</p>