The mathematical modelling of fish behaviour in relation to towed fishing gear based on quantitative analysis of stimuli of the gear and fish response, was investigated in order to increase the selectivity and efficiency of the gear. The brightness contrast of the netting and other components of the fishing gear such as floats and bobbins was studied by comparing their luminance to that of their visual background of water, sand-cloud or sea bed. Luminance of each component was measured with a luminance meter in the tank and by the video image processing of nets and gears in the sea with light intensity and grey levels. The brightness contrasts of the gear are increased with diameter of twine, light level and viewing zenith angle. A number of contrast modules are developed, based on the measurements, for predicting the visual stimulus for reactions of fish to towed fishing gear. The underwater luminance against which a fish sees the net components is calculated for the prediction process. The visual stimuli of trawl were represented through the predicted light field and methods are developed for calculating the visibility of the trawl components and sand clouds as viewed from any fish position within the net. Fish behaviour in relation to towed fishing gear was observed and analysed using a net panel towed in the gantry tank and bottom trawls at sea observed by a SIT video camera system. When fish displayed stable swimming as in an optomotor response, the variations of relative swimming speed and angular velocity were relatively low and regular in their amplitude and period. However, when fish moved suddenly in panic reaction to escape or rise up, the variations of these measured components were high, irregular and nonperiodic.

This thesis was focused on the study of (i) individual feeding behaviour and growth performance of juvenile rainbow trout fed on normal and medicated diets and (ii) the responses of hepatic detoxification mechanisms in rainbow trout to dietary oxolinic acid and flumequine. During the first experiment, the development of dorsal and caudal fin damage over time was compared between 4 groups of rainbow trout fed on different ration levels (0.25, 0.5, 1.0 and 1.5% B.W. day^-1) and between individuals of different feeding rank within each group. The feeding and growth data indicated that the strength of the social hierarchy became more relaxed within increasing ration. The severity of both dorsal and caudal fin damage was significantly dependent on the ration size fed to the group, with lower ration groups sustaining more fin damage, whereas subordinate fish suffered the most dorsal fin damage within the two lower ration groups. Subsequently, the inter-individual variability in consumption and growth was studied in two groups of rainbow trout fed diets containing flumequine in different concentrations. Flumequine, at either concentration, had a deleterious effect on growth performance, even when palatability problems were not apparent, and resulted in dose-related increases in inter-individual variability in consumption and growth and changes in the pre-established feeding hierarchies. Cytochrome P450 biomarkers and ultrastructural alterations in hepatocytes were used to assess the toxicological responses to rainbow trout following oral administration of oxolinic acid or flumequine. Both antibiotics induced and/or activated CYP1A isoforms of P450 as measured using immunoblotting for P450 protein and a combination of selective substrates and inhibitors for P450 activity. The increase in the P450 activity was correlated with a significant increase, almost 2.5-fold, in the volume of rough endoplasmic reticulum. In addition, severe depletion of glycogen deposits, formation of lipid droplets, increased lysosomal activity as well as mitochondrial hyperplasia and high nuclear-cytoplasmic ratio in flumequine- and oxolinic acid-treated individuals respectively, were considered compensatory rather than degenerative phenomena, indicative of increased biotransformation activity.

This study focuses on fish discards generated by demersal vessels fishing in Scottish waters (North Sea and West of Scotland). The analysis is based on data collected by scientific observers onboard commercial vessels during the past two decades (1975 - 1993). Chapter 1 introduces the problem of discarding, describes its implications for fisheries and the marine environment, and gives a brief overview of the sampling scheme. Chapter 2 uses generalised additive models (GAMs) and regression trees to investigate the main reasons that lead Scottish fishers to discard commercially important species (haddock - <I>Melanogrammus aeglefinus, </I>whiting - <I>Merlangius merlangus </I>and cod - <I>Gadus morhua</I>). Different discarding patterns are observed in inshore and offshore areas, with vessels fishing offshore discarding larger fish, irrespective of gear type. In inshore areas, discarding decisions are mainly driven by legal landing size, whereas offshore, high grading decisions due to market pressures and, possibly, quota limitations prevail. Chapter 3 explores the performance of several estimators of total discards and their variances in a simulated population of fishing trips. The simulated population is generated from the residuals of temporal models fitted to the catch and discards observations from the trips sampled under the Scottish discards monitoring scheme. The estimator currently applied to obtain annual estimates of total discards in Scotland is shown to be biased and imprecise, whereas several other estimators perform better under limited sampling coverage. The best estimators from chapter 3 are used in chapter 4 to estimate the total discards and their variances for species that were regularly discarded by Scottish vessels in the period 1988 - 1993. Overall, it is estimated that Scottish vessels generate annually around 100,000 tonnes of demersal fish discards in the North Sea and 30,000 tonnes in the West of Scotland. This suggests that close to 40% of the total demersal catch of Scottish vessels is annually wasted through discarding.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.

No description available.