The populations of Atlantic cod (Gadus morhua) in the Baltic Sea has been heavily exploited for decades, with fishing mortalities close to, and for several occasions, even above one. The larger part of the spawning stock biomass is consequently being removed each year. The issue of fisheries induced evolution (FIE) has been gaining attention from researchers lately. The selection pressure driving this evolution is powered by a connection between high mortality rates and heritable traits. The fishing in the Baltic Sea is mainly performed with size selective gear that can impose selection on traits like size-at-age or size-at-maturity. In this review I show how FIE may affect the Baltic cod towards decreased size-at-age/maturity and how size selective fishing on stocks at low levels can increase the inherent instability of the population as well as deprave the Baltic Sea of ecosystemic services from cod.

Sexual maturation during commercial culture of Atlantic cod (Gadus morhua L.) represents a significant production bottleneck restricting the profitability of the industry. Such problems in other species have traditionally been addressed by artificial manipulation of photoperiod cycles, however little research exists in this field in cod. This thesis therefore investigates the interactions between artificial photoperiod manipulation, sexual maturation and somatic growth in this species. In the first experiment, populations of Atlantic cod (hatched, spring 1999) were maintained on either a simulated natural photoperiod (SNP) or continuous illumination (LL) from approximately 15 months post hatch (MPH) (July 2000) in an enclosed tank system. Growth performance was recorded monthly along with observations of reproductive activity over the subsequent 2 years (up to July 2002). At both 2 and 3 years of age the entire population raised under SNP matured and spawned, during which time mean weight reduced by 13% and 24% respectively. No spawning individuals were recorded at 2 years of age in the LL population and only 18% were observed to spawn at 3. However, observations of both changes in gonadal morphology (observed via ultrasound scanning) and a suppression in growth rate at 2 years of age in the LL population alluded to a maturation “dummy run” regulated by an endogenous clock. Despite this phenomenon, the LL treatment realised a 39% and 43% improvement in wet weight following 1 and 2 years of exposure to LL respectively. When the diel cycle of plasma melatonin was compared between the treatments in February 2001 (23MPH) the SNP population displayed an A-profile diel rhythm ranging between 20 and 50 pg/ml while the LL treatment did not display any rhythm. In the second experiment of this work, two populations of cod (hatched, spring 2001) were reared in commercial open cage systems, one of which experienced continuous additional artificial illumination between July 2002 (15MPH) and October 2003 (30MPH) provided by four, 400W submerged lighting units. Growth and maturation were assessed in both populations throughout. In March 2003 (24MPH) it was apparent that spawning individuals were present in both the SNP and LL populations though a significantly lower number of spawning individuals in the LL treatment suggested that the peak in spawning activity was delayed by about 1 to 2 months. With both populations apparently maturing at 2 years of age, there was no significant difference in weight between the populations at the end of the trial. In comparison to experiments I and IV of this work, these results would suggest that in comparison to salmonids for example, Atlantic cod appear to have a heightened sensitivity to light allowing individuals to differentiate the ambient photoperiod signal from the application of continuous artificial light. In the third experiment, 6 populations of approximately 20 tagged individuals (hatched spring 1999) were maintained, from December 2000 to July 2002, under either SNP, LL or one of four, out of season “square wave” photoperiod regimes (repeating cycles with a 12 month period, consisting of a 6 month window of LL followed by six months of short day lengths [SD, 7L:17D] which had been staggered to start over a six month period). Each individual was monitored monthly for maturation status. Out of season “square wave” photoperiods were demonstrated to successfully entrain maturation and hence significantly alter the spawning profiles in these populations. Application of LL from December 2000 failed to inhibit maturation in the spring of 2001 and, in fact, advanced the spawning season by 1 month while those that experienced SD from the same date showed significant extension of the subsequent spawning season. Interestingly, the males maintained on LL throughout the experiment matured both in the spring of 2001 and one year later in the spring of 2002 while females under the same treatment only matured and spawned in 2001. In the fourth experiment, a total of 830 tagged individuals were raised either under SNP or one of 7 photoperiod treatments, consisting of 5 groups transferred from SNP to LL at 3 monthly intervals between 6 and 18 MPH where they remained and a further two groups maintained on LL from 6 to 15MPH and 6 to 21 MPH respectively before being returned to SNP. Both the gonadic and somatic axes were monitored at the physiological and endocrinological level at three monthly intervals from 6 to 27 MPH. The results demonstrated that it is the falling autumnal photoperiod signal after the summer solstice, more specifically after October, that is responsible for recruiting individuals to enter the sexual maturation cycle. Furthermore, in all treatments where this signal was masked i.e. those which experienced LL starting at or prior to 15MPH, except for some restricted spermatogenic activity in the males testis observed at 27MPH, there was no significant reproductive activity and growth was improved by up to 60% at 27 MPH. While providing evidence for direct photic stimulation of somatic growth, the growth results were also correlated with the measurement of plasma IGF-I and demonstrated its potential as a tool to assess growth rates in the species. Plasma melatonin measured at 15MPH, as in experiment I, was suppressed in all populations which were under LL photoperiods. By identifying the photoperiod “window of opportunity” which recruits individuals into the sexual maturation cycle, this work was able to conclude that the application of LL from the summer solstice prior to maturation is the most efficient photoperiod strategy to be adopted by the aquaculture industry to realise maximum growth potential from their cultured stocks.

In Scotland, environmental regulation restricts commercial cod culture to the equivalent of 66 % of that granted for commercial Atlantic salmon (Salmo salar L.) farms. This calculation is based on estimations of nitrogen discharge from the difference in protein content between salmon and cod diets, with the higher levels of protein in cod diets suggesting a higher nitrogen discharge compared to that observed for salmon diets. In turn, this could potentially result in increased nitrogen enrichment of a marine ecosystem. The aims of this study (quantifying and modeling of nitrogenous wastes associated with the commercial culture of Atlantic cod (Gadus morhua L.) were achieved through a series of tank and cage investigations, each of which studied juvenile and adult Atlantic cod. The study provided data with respect to nitrogen excretion from juvenile and adult fish in both systems. This would allow the development of dispersion models and the calculation of nitrogen budgets for commercial cod culture, thus providing environmental regulators data independent of salmon models to create regulations that would be specifically applied to cod farming. The tank - based studies investigated three diet formulations produced by EWOS® Innovation in Norway, as a 4 mm pellet (juvenile study) and as a 7 mm pellet (adult study). The three iso -energetic diets varied primarily in protein content (40%, 50% and 60%). Two tank studies, one on juvenile and one on adult cod, investigated growth, condition and tissue composition, and the production of dissolved nitrogenous wastes over a 5 and 7 month period respectively. At the beginning of the acclimation period prior to the adult tank study commencing, the fish had a mean weight of approximately 1275 g. The difference in the final weight promoted by each diet was not significant (with an approximate final weight of 2400g), suggesting that a low protein diet (40 % protein) promoted similar growth to a high protein diet (60 % protein). Other growth and condition parameters were also similar for all diets with the condition improving over the course of the study. As the fish completed spawning immediately prior to the commencement of the study, an increase in condition was not observed until approximately day 90 of the 210-day investigation. Over the course of the 5-month juvenile study, growth was approximately 224g for the 40% protein diet and approximately 275g for the 50% protein and 60% protein diets. This suggests that a higher protein diet is required for optimal growth of juvenile cod and that 50 % and 60% protein diets promote similar growth and condition, potentially reducing the protein requirement of juvenile diets. In the juvenile investigation, condition increased over the full range of the study. In both tank studies, nitrogen digestion was directly related to protein (and associated nitrogen) content of the diets in the juvenile study expressed as a percentage of the nitrogen content of the diet as 59.19% (40% protein), 56.90% (50% protein) and 52.23% (60 % protein) suggesting that nitrogen digestion is more efficient at lower protein content in the diet. When expressed as a percentage of the nitrogen content of the diet, nitrogen digestion observed in the adult study was 60.55%, (40% protein) 60.92% (50% protein) and 60.60% (60% protein) respectively, suggesting protein digestion is similar regardless of protein content in adult cod. In the adult tank study, under a manual feeding regime, a post-prandial-peak is observed at 105 min. following the cessation of feeding. Thereafter, ammonia levels drop over the course of the sampling period. Following the afternoon meal commencing at 420 min., ammonia levels rise at least until the final samples are collected at 450 min. Under an automated regime, a lesser post-prandial-peak is observed but the ammonia concentration is lower over the sampling period compared to the respective profile under a manual regime. The ammonia profile produced throughout the juvenile tank study follows a very similar trend to that observed in the adult study under the manual feeding regime. Two cage - based investigations took place at the No Catch® Ltd. commercial organic cod farm in Vidlin Voe on the east coast of Shetland. Both studies investigated growth, condition and tissue composition, as well as the production of dissolved nitrogenous and particulate wastes associated with the culture of juvenile and adult Atlantic cod in cage systems. Sampling for the adult study occurred over three days during three sampling trips (September 2005, November 2005 and February 2006). Sampling for the juvenile study took place over three days on a single trip to Vidlin in late April 2006. The diets used at No Catch® Ltd. were produced by Biomar® in Grangemouth. A relationship between feeding and ammonia concentration is less evident in the cage studies than in the tank studies, and similarly, the relationship between feeding and ammonia concentration is less evident in juvenile fish than in adult fish. As ammonia values were converted to (µg/L/tonne biomass), the ammonia concentration recorded is largely dependent upon the biomass of the sampled cages at both the nursery site and production site. Deposition rates of organic carbon and nitrogen around the production and nursery cage sites in Vidlin Voe are related to the position of the sediment trap relative to its location and proximity of the trap to the specific cage site. Weather condition also had an impact on deposition rates with calmer weather producing lesser deposition rates. Around the production site, deposition rates of organic carbon and nitrogen are greatest in the direction of the prevailing current. Deposition rates decrease with an increasing distance from the cages. Although sediment trap results were inconsistent, a similar trend is observed for each of the three sampling trips, although actual deposition values were different. Differences between the deposition rates at the highly dynamic production site and the low energy nursery site indicated that sedimentation of waste from cod culture is highly dependent on water currents. Models of particulate waste deposition associated with the production cage site in Vidlin Voe were produced using the spreadsheet - based Cage Aquaculture Particulate Output and Transport (CAPOT) model, developed at the Institute of Aquaculture. The models were parameterised using the data collected and tested against an established regulatory model, DEPOMOD. The similarity in results illustrated the robustness of the highly flexible spreadsheet waste model for cod culture.

Aquaculture is an essential developing sector for world food production, however the attainment of sexual maturity during commercial on-growing is a major bottleneck to industry expansion. Sexual maturation brings a commercial loss due to reduced growth performance as well as reduced immune function. Furthermore, serious concerns exist over potential genetic interaction with native stocks through broadcast spawning or spawning interaction by escapees. In the north Atlantic region, the Atlantic cod (Gadus morhua) and Atlantic halibut (Hippoglossus hippoglossus) are key aquaculture species in which industry expansion is limited by pre-harvest sexual maturation. However, through a species specific combination of modern technologies and refinement in management practices it is possible that this sexual maturation can be controlled and on-growing potential enhanced. Thus the overall aim of this thesis was to conduct novel research that will improve our understanding of the underlying mechanisms that regulate sexual maturation, whilst also advancing the optimisation of technologies for the management of maturation in cod and halibut. In Atlantic cod, owing to the inconsistent inhibition of maturation in commercial conditions, ever increasing intensities of light and in some cases narrow spectrum technologies are being used to try to combat this problem. Firstly, this PhD project investigated the potential welfare impacts of high intensity artificial lighting which have not been studied to date (Chapter 2). The work specifically investigated the effect of traditional metal halide and novel green cathode lighting on the stress response, innate immunity, retina structure, feeding activity and light perception of Atlantic cod. Results indicated that although acute responses to light were observed, there were no clear significant long term effects of any of the lighting treatments on these parameters. Regarding light perception, interestingly even when subjected to high intensity constant lighting (metal halide mean tank intensity: 16.6 watts m-2), cod still demonstrated a day/night rhythm in melatonin release which suggests perception of the overlying ambient photoperiod. The second trial of this PhD project investigated the efficacy of shading of ambient photoperiod in addition to constant lighting to inhibit maturation of cod outdoors (Chapter 3). This aimed at improving the performance of artificial lighting regimes in the open cage system during commercial on-growing by reducing the relative difference between day/night light intensities. The trial was conducted over a one year period where a low and high shade treatment were tested in outdoor tanks. Shading increased the relative night time illumination to 6.6% and 31.3% of daytime levels respectively, compared to <2% in an unshaded set-up. Both shading treatments were effective at suppressing sexual development in cod as confirmed through measurements of gonadosomatic index, histological analysis of gonadal development, oocyte diameter measurements and sex steroid profiles as well as measurements of growth. In addition to research at the applied level in Atlantic cod, this thesis has also extended to the fundamental level and explored one of the potential mechanisms relaying photoperiod signal to the endogenous regulation of sexual maturation in cod, namely the kisspeptin system (Chapter 4). Partial sequences for the signal peptide Kiss2 and its receptor Kissr4 were isolated and described showing similarity to other teleost species such as the medaka, Oryzias latipes and stickleback, Danio rerio. Novel molecular qPCR assays were designed and developed to measure the expression of both genes in male and female cod over a maturation cycle and compared to cod under constant lighting which remained immature. Interestingly, expression patterns of kiss2 and kissr4 did not reveal any clear association with season or photoperiod treatment. However, pituitary expression of gonadotropins (FSH, follicle stimulating hormone; LH, luteinising hormone) did show a differential expression in relation to treatment from early winter approximately 4-6 months after the photoperiod change. These new results are in contradiction with the hypothesis that the kisspeptin system would be involved in the initiation of gametogenesis, as shown in mammals. However, the FSH/LH data defines a window during which time kisspeptin or another GnRH stimulating mechanism must be active, this compels the need further investigation. In Atlantic halibut farming, all-female production removes the concerns of production losses through sexual maturation. Accordingly, this thesis investigated the potential/feasibility of generating monosex populations by FACS (fluorescence activated cell sorting) semen sexing based on cellular DNA content, as proven in terrestrial agriculture. Results however did not show any clear differences between the DNA of sperm in a range of species tested (Atlantic halibut, cod, sea bass, perch) suggesting that this technique may not be applicable in such species. The project also focussed on the production of a population of sex reversed halibut broodstock (neomales) that will generate, in the long term, a basis for traditional monosex population generation in the UK. Two in feed MDHT (17α-methyldihydrotestosterone) treatments were tested with the aim to reduce the use of hormone. Results were very successful with a hormone treatment of 5ppm MDHT generating a 97% phenotypic male population thus suggesting the presence of sex-reversed halibut which can be used for future monosex production. Overall, this work aimed to develop and/or refine potential remediation techniques for sexual maturation in two key commercially important farmed marine fish species, cod and halibut, as well as further our understanding on the regulation of puberty. The knowledge gained from this work provides a means to optimise the techniques employed in the industry and has the potential to increase production and profitability without compromising farmed animal welfare, thus ultimately promoting the sustainable expansion of the Atlantic cod and halibut aquaculture.