Abstract Sepsis is a serious clinical condition that is characterized by a systemic inflammatory response syndrome resulting from a known or suspected infection. The major clinical symptoms involve an abnormal WBC count, elevated body temperature, respiration and pulse rate. Reported cases with high mortality rate range between 13 - 20 million. In order to treat Sepsis, the detection of bacteria in blood culture is extremely crucial. Treating patients with broad spectrum antibiotics is usually related to adverse effects, drug resistance, increased mortality, and high cost. In the past decades, researches had detected that E. coli and S. aureus are the major role players that cause sepsis. These microbes are molecularly tested by methods like MALDI TOF, FISH and Microarrays.   In this analysis, DNA fluorescence in situ hybridization (FISH) assessment for the identification of S. aureus, one of the most frequent blood pathogens, was optimized in the labs of Högskolan i Skövde. As a result, the growth of S. aureus was observed very carefully, optimizing the FISH procedure for gram positive bacteria was done and the sensitivity, stability and specificity of the DNA probe were examined under variant conditions like the continuous decrease in the bacteria cells number and utilizing a mixture of different types of bacteria cells. / Data mining strategies and molecular techniques for microbial detection in severe sepsis

A standardised procedure was established for the laboratory maintenance of C.irritans in thick-lipped mullet. Nine isolates of C.irritans were obtained of which eight were successfully established for up to 48 weeks. Studies on individual life cycle stages with regards to longevity, viability, and susceptibility to a chemotherapeutic agent, revealed the difficulties in eradicating the cysts. Transmission of the parasite both to and from the host correlated with darkness. High infection levels resulted in the death of host fish within five days following exposure to theronts. An acquired protective immune response developed in host mullet within 14 days after exposure to sub-lethal infection. The degree of immunity appeared to relate to infection dose, and was not fully protective in all fish. Protection persisted for six months after infection and appeared specific to C.irritans. Specific antibodies to trophont antigen were identified in mullet serum but not epithelial mucus following either natural exposure to theronts or intraperitoneal immunisation with trophont antigens. Serum from intraperitoneally immunised fish caused theront immobilisation and agglutination in vitro; however no evidence was found for a protective role for specific antibody. Major polypeptides were identified and characterised by molecular weight for both trophont and theront stages using SDS-PAGE. Significant homology in major polypeptide profiles was found between C.irritans and I.multifiliis, in respect to trophonts and particularly theronts. Murine monoclonal antibodies raised to trophonts identified two polypeptide components of molecular weights 20-21kDa and 68-69kDa, the latter being homologous with host immunoglobulin heavy chain. These results are discussed in relation to future management and control strategies for cryptocaryosis in warmwater mariculture systems and aquaria.

Peripheral blood leucocytes of the turbot, Scophthalmus maximus, were characterised into 4 distinct groups following morphological, morphometric and histochemical examination. Total and differential cell counts were determined. Thrombocytes, the most abundant leucocyte type (52%), were highly mobile and encountered in several morphological forms. Granulocytes, representing 5.6% of the leucocyte population, histochemically most resembled the mammalian neutrophil. Both large and small lymphocytes (40.8%), were encountered. Monocytes were rarely observed (1.6%). Thrombocytes and monocytes were phagocytic in vitro at 12oc and 22oc, showing increased phagocytic activity at the higher temperature. The thymus was paired and consisted of a well developed outer cortex and an inner meduallary region. The spleen was bounded by a fibrous tissue capsule and contained a large volume of blood. Diffuse areas of red and white pulp, ellipsoids and melanomacrophage centres were apparent. Lymphocytes, thrombocytes and mature erythrocytes made up the cellular components. The kidney, located beneath the vertebral column contained haemopoietic tissue throughout. Excretory tubules were evident posteriorly. Cellular elements included developing granulocytes, large and small lymphocytes and melanomacrophages. Investigation of ontogenic development of the lymphoid tissue, from 24h post-hatch to the completion of metamorphosis (Day 63) revealed thymic, splenic and kidney rudiments all present at Day 4 with the first lymphoid cells appearing in thymus and kidney by Day 8. Splenic lymphoid cells and the development of areas of white pulp were apparent by Day 28. Differentiation of the thymus had occurred and melanomacrophage centres were seen in the spleen, completing structural lymphoid development by Day 63. Critical stages of lymphoid ontogeny were correlated with easily recognisable external morphological features. A study of the kinetics of carbon clearance by the reticuloendothelial system, revealed a phagocytic capacity in the spleen, kidney and heart. Splenic carbon was seen at 20min post injection, accumulating around ellipsoids and rising to a maximum level at 24h. By Day 5 carbon levels within phagocytes, by now more distant from the ellipsoids, had begun to decrease and carbon was seen within melanomacrophages. Levels of kidney carbon, present within large macrophage-like cells which increased in size forming larger aggregations, increased to a maximum at Day 3. Clearance appeared more rapid in the posterior kidney. Low level uptake was seen within the epicardium. Carbon uptake was not observed in the liver or gill. Kidney leucocyte migration in vitro was examined to a range of chemoattractants using a number of assays. 24h bacterial culture supernatants of Vibrio alginolyticus induced significant cellular responses. The under agarose assay demonstrated migration inhibition to 100%, 50% and 40% supernatant dilutions. Enhanced migration was detected to dilutions of 5-50% in the microchemotaxis chamber, being optimal at 20%. The leucocyte polarisation assay demonstrated cell orientation in response to I 00% culture filtrate and the capillary tube migration assay revealed cellular inhibition at concentrations of 10% & SO%. Leukotriene B4 (LTB4) also induced migration in the filter-based assay, being optimal at to-7M. Cellular migration and orientation were observed in filter and polarisation assays to turbot serum, with normal and activated serum inducing elevated responses in the filter based assay. No response was detected by any of the assay systems to n-formylmethionyl-leucyl- phenylalanine (FMLP) or casein at any concentration tested. Results are discussed in relation to the cellular defence mechanisms of fish.

The frequency and severity of fish diseases is; increasingly, being reported as a limiting factor in the future development of aquaculture. The control of fish disease is largely performed in retrospect through curative chemotherapy. However, the development of resistant strains of fish places the emphasis on preventative, rather than the curative, control of disease. The freshwater parasite Icthyophthirius multifiliis causes the disease known as ichthyophthiriasis or white spot. Losses of fish due to this parasite are believed to total over one million dollars per annum worldwide. The objective of this study was to examine the genetics of resistance to I. multifiliis, Comparisons were made between four stocks of the tropical livebearing fish Xiphophorus maculatus. One of these stocks, the blue platyfish, was found to be less susceptible to white spot than the yellow comet tail, red or red wag tail platyfish, Comparisons with four other tropical spades of fish found significant differences between X. maculatus, X. variatus, Ameca splendens and Ilyodon xanthusi in levels of susceptibility. A. splendens was the most susceptible species, with the blue platy (X. maculatus) and sunset platy (X. variatus) forming the most resistant group. All the remaining stocks and species formed an intermediate group. Examination of resistance to white spot infection in four scale types of related common carp (Cyprinus auratus) also found variation in resistance, with the fully scaled carp being the most resistant phenotype (scattered mirror, linear mirror and leather carp being similar in infection level). More detailed analysis of the genetics of disease resistance was performed with heritability determinations in stocks of A. splendens, X. maculatus (yellow comet tail) and X. maculatus (Vera Cruz). The highest heritability value, based on sire components only, was that for Vera Cruz platyfish of 0.75, with a value of 0.23 for X. maculatus (yellow comet tail) and 0.00 for A. splendens. A breeding programme was also performed between X. maculatus (red platy) and X. variatus (sunset platy) to evaluate the presence of any heterosis. Such was observed, with a heterosis value, based on actual parasite counts, of 16.2%. Several factors of the infection process are also discussed, especially the fluctuating yearly trends in infection levels and parasite strain differences. Finally, the future potential of genetic manipulation of fish stocks for increased disease resistance is discussed in the light of this study. It is considered that a useful foundation has been laid for the further development of this approach to disease prevention in aquaculture.

Given the widespread concern that chemical contaminants may be associated with infectious disease outbreaks in marine fish populations, work has been undertaken with the aim of developing a suite of non-specific and specific assays of marine fish immune function and the application of these techniques in a variety of field and laboratory investigations. Most of the work focused on dab, Limanda limanda (L.) in view of the importance of this species in several North Sea fish disease monitoring programmes, and was also supplemented with investigations of specific immune function in turbot, Scopthalmus maximus (L.). Initial field studies examined non-specific immune function in terms of lymphoid organ morphology in dab sampled along a North Sea gradient of chemical contamination during the March 1990 ICES/IOC Bremerhaven workshop. Significant differences were observed in the kidney and spleen cell populations from dab, and these observations were considered in view of the various other physico-chemical and biological results generated during the Bremerhaven workshop. Following the valuable experience gained of the practical aspects of the field monitoring approach, laboratory investigations were initiated with the aim of developing a suite of immune function assays for deployment in either laboratory or field studies of marine fish health. Assays for non-specific immune functions were considered, including serum protein and lysozyme levels, methods of phagocyte collection, phagocyte chemiluminescence, calorimetric detection of individual reactive oxygen species and in vitro cell migration assays. Additional field work was undertaken, with the monitoring of serum total protein levels and lysozyme activity in dab sampled from Lyme Bay, UK This study provided evidence of a marked seasonal variation in non-specific immune function, which appeared to be associated with environmental factors (e.g. water temperature) and the reproductive cycle. Selected non-specific assays were applied to dab and turbot exposed under controlled laboratory conditions to a variety of important marine contaminants, including cadmium, polycyclic aromatic hydrocarbons (PAHs) and polychlorinated biphenyls (PCBs), and the biological data integrated with appropriate chemical characterisation of the exposure matrix. Using cultured turbot as a test species, the non-specific assays were also supplemented with the assay of specific immune function in fish exposed to PCB contaminated sediments. In brief, there was evidence of significant impairment of immune function in fish exposed to either individual contaminants (viz. cadmium and PAHs) or contaminant mixtures (viz. PAHs and PCBs) under the laboratory conditions described. In summary, the project was successful in its primary aims of developing a suite of techniques for evaluating both cellular and humoral immune functions in marine flatfish, and applying these techniques in the laboratory to assess the impact of important classes of environmental contaminants on fish health. Selected techniques were also used in field monitoring studies of marine fish immune function, illustrating the potential of such techniques for use in future laboratory and field studies of fish health.

Studies were undertaken to examine the extent of molecular variation among isolates of the fish pathogen Renibacterium salmoninarum. As many isolates as possible were gathered from diagnostic laboratories within the UK, and checked for viability and contamination. The isolates were derived mainly from infected rainbow trout and Atlantic salmon that were farmed in England, Scotland and Wales, subject to the requirements of statutory legislation. Incomplete histories were available for the sources of the isolates, which spanned a time scale of 36 years, from 1962-1998. Molecular variation between the isolates was examined using two strategies. Firstly, defined regions of the genome were examined for polymorphisms. PCR analysis of previously characterised genes, msa, hly, and rsh, revealed no length polymorphisms among 43 UK isolates of R. salmoninarum. The 23S and 5S rRNA genes were sequenced and sequence analysis of the 16S- 23S (ITS I) and 23S-SS (ITS2) rRNA regions was performed. Sequence analysis confirmed the correct taxonomic placement of R. salmoninarum within the Micrococcus/Arthrobacter subdivision of the Actinomycetes. Some isolates possessed small sequence variations within ITS1 that can provide an indication of their geographical origins. Sequence variation also exists in the ITS2 region but was found only within a single isolate from an outlying region of Canada. Ribotyping was found to be of limited use for discriminating among isolates of R. salmoninarum probably because R. salmoninarum possesses only two copies of the rRNA operon with no length polymorphisms in the ITS regions. Additionally, the discovery and analysis of a genetic locus containing a 51 bp exact tandem repeat unit (designated ETR-A), revealed that some isolates of R. salmoninarum could be distinguished by the possession of one rather than two copies of this repeat unit. Finally, PCR amplification of length polymorph isms in the tRNA gene spacer regions (tDNA-PCR) was developed using consensus tRNA gene primers to enable tRNA genes and spacer regions to be cloned and sequenced. Subsequently, specific tRNA gene primers were designed and enabled the discrimination of 43 UK isolates into I 5 groupings. tDNA-PCR proved to be one of the most powerful typing methods applied to this organism. Secondly, typing methods that analysed the whole genome for the presence of molecular variation were employed. A putative insertion sequence IS994, was used as a probe in a RFLP-based study to discriminate between 52 isolates of R. salmoninarum from different countries. This method showed great potential for distinguishing between isolates and separated the 52 isolates that were examined into 12 groupings. Randomly amplified polymorphic DNA (RAPD) was also used to examine 28 isolates of UK origin. This method was found to be highly discriminatory, with 28 isolates generating 12 different banding patterns, which appeared to reflect geographical source. Pulsed field gel electrophoresis was also used to investigate the genomic diversity of isolates. Due to technical difficulties in obtaining pure, undamaged DNA the preparation of suitable macro restriction profiles was never achieved. However, preliminary findings suggested that the R. salmoninarum chromosome was linear and approximately 4.5-6Mb in size. Finally, repetitive element PCR was evaluated using 3 different approaches (ERIC, REP and BOXA-2) but proved to have a limited capacity for defining molecular variation between different isolates. Ultimately, RAPD, tDNA-PCR and 1S994 RFLP profiling proved to be most useful for the molecular differentiation of R. salmoninarum. A comparison of the groupings that resulted from each of these methods revealed substantial areas of agreement. The use of a multifactorial approach not only resulted in a greater discrimination of isolates but also provided increased confidence in the outcome. It is recommended that for typing purposes such an approach should be adopted. Epizootiological interpretations of groupings were hampered by the general lack of background information attached to each isolate. However, the application of multiple typing methods reveal that, despite the highly conserved nature of this bacterium, UK isolates of R. salmoninarum possess genetic diversity, with geographically related isolates often being grouped together. Overall there was little evidence to suggest the regular introduction of genetically distinct R. salmoninarum into the UK and the results suggest that some isolates may be relatively localised despite the international trade in fish stocks.

Consumer preference for commercially reared fish products that resemble their wild counterparts has resulted in the supplementation of pigments called carotenoids into aquafeeds to promote a pink-red colour in the flesh of salmonid fish. To date synthetic forms of these pigments have been commonly utilised to achieve this desired colouration, with the carotenoid astaxanthin being the regular choice for the feed manufacturer. However, increase in consumer demand for farmed fish products reared on natural feed additives has evoked an interest in natural sources of astaxanthin that could be successfully used to pigment salmonid fish efficiently. In the current study, the microalga Haemalococcus pluvialis has been assessed as a potential feed supplement to pigment the flesh of rainbow trout (Oncorhynchus mykiss). More specifically, those natural characteristics that may well limit the absorption and utilisation of astaxanthin from this source have been assessed individually and discussed from a physiological standpoint. The cell wall of Haemalococcus pluvialis when cracked efficiently presents no limitation to the absorption and utilisation of astaxanthin from this source. Indeed, the cell wall remnants help to prevent oxidation of astaxanthin in the feed compared to cell wall free extracts of carotenoid from the same source. However, esterified astaxanthin (which this algae predominantly contains) is not absorbed as efficiently as unesterified synthetic astaxanthin. Furthermore, the extent of esterification is negatively related to the absorption of astaxanthin. Regional variation in ester hydrolysis along the gastrointestinal tract combined with gut transit time of the ingested feed may explain these limitations. However, despite limitations in absorption, the muscle deposition of astaxanthin supplied as esters does not significantly differ from the unesterified form. The optical purity of astaxanthin esters from this source does not prejudice the final deposition of astaxanthin in fish tissues. An in vitro model has been developed to assess the absorption of astaxanthin at the intestinal level in salmonid fish in order to define absorption characteristics of carotenoids under different abiotic and biotic conditions. The absorption of astaxanthin seems to occur in a linear passive manner into the intestinal tissue. Although size of the fish does not affect the absorption of astaxanthin, temperature does have a significant effect. Although there were no significant differences in absorption between Atlantic salmon (Salmo salar) and rainbow trout, absorption tended to be greater in the latter species and merits further study.

The lucinid bivalve Codakia orbiculata, whose gills contain sulphur-oxidizing symbiotic bacteria, occurs in high densities (500/m2) in the rhizosphere of shallow-water Thalassia testudinum sediments in Bermuda. Both sulphide and thiosulphate stimulate aerobic respiration in the isolated bacterial symbionts of C. orbiculata. Sulphide and thiosulphate stimulate anaerobic ^^COa fixation in bacteria isolated from sulphur-starved bivalves. Interstitial water sulphide concentrations in the bivalves' habitat reach 300 pM, and sulphate-reduction rates are high, but thiosulphate concentrations are low (0.66-32.27 pM) . Thiosulphate supplied to the symbionts in vivo must be produced by sulphide oxidation, possibly by the host bivalve. Isolated symbionts also respire aerobically and fix i * C02 in the absence of exogenous reduced sulphur, suggesting utilization of intracellular elemental sulphur stores. Codakia orbiculata symbiotic bacteria are able to respire nitrate. Nitrate concentrations in the interstitial water of C. orbiculata habitat can reach 36 pM. Thiosulphate stimulates nitrate respiration in the intact symbiosis, incubated in oxic and anoxic conditions, and in anoxic incubations of isolated symbionts. Intracellular elemental sulphur is also used by the •bacteria as a substrate in nitrate respiration. Nitrate respiration in the absence of exogenous nitrate suggests that the sjnnbionts may have a limited ability to store nitrate. There is no direct evidence that sulphide stimulates nitrate respiration in either the isolated symbionts or the intact symbiosis, incubated in anoxic conditions. Nitrite respiration in the symbionts is stimulated by sulphide (only), however. Because nitrate respiration was measured by nitrite accumulation, complete denitrification would explain the apparent failure of sulphide to stimulate nitrate respiration. High nitrate respiration rates in the intact symbiosis, incubated with sulphide in oxic conditions, may be in response to thiosulphate, supplied to the bacterial symbionts after host oxidation of sulphide. Nitrite respiration in the intact symbiosis, even when incubated in oxic conditions, demonstrates that the symbionts have access to some sulphide in vivo, however, and that host sulphide-oxidation may not be- 100% efficient. Nitrate and nitrite respiration in the intact symbiosis, even when incubated in oxic conditions, suggests that the bacteria may be exposed to low oxygen levels in vivo and may require the ability to utilize an alternate electron acceptor. Like some free - living bacteria , Codakia orbiculata bacterial symbionts may co-respire, or alternately respire , oxygen and nitrate . The Thalassia testudinum sediments in Bermuda may be ideal for this bacteria-bivalve symbiosis due to the availability of oxygen, nitrate and sulphide.

The distribution system in the fishing industry has attracted little attention from academics in comparison to the catching sector. This realisation came in the early 1980's with questions being raised about system efficiency and the identification of problems. A lack of understanding of the distribution system led to various calls for an investigation. This research builds on existing understanding by identifying the distribution system in Devon and Cornwall and develops a modelling framework with forecasting properties within which distribution system problems can be analysed. A review of the marketing and distribution literature reveals that system efficiency and the impacts of changes in external conditions are important areas for study. The literature also reveals that few attempts have been made to develop analytical and conceptual modelling frameworks that enable the study of these phenomena in existing distribution systems. Attempts that have been made suffer from conceptual and data problems. The research identities the major features of the fresh fish distribution system, especially the structural determinants including lack of standardisation and variability of supply. Studies of the system in the United Kingdom and overseas are found to be descriptive, concentrating upon trends and past problems. Few modelling approaches have been used or developed. The system in Devon and Cornwall is identified through interview techniques and presented in quantitative terms. From an identification of the system, it is determined that a model should focus on the port merchant sector and have the ability to explain and forecast the effects of supply variation. Econometric methods are used to develop a model with exogenous supply and seasonality characteristics which when combined with sample data on costs and pricing provides a method of analysis and forecasting. The research discusses the wide range of applications possible with the model developed.

This research programme aimed to review the nutritional requirements for the main minerals, (namely calcium, phosphorus, magnesium and zinc) formulated in commercial diets that are essential for the health and growth performance for salmonid fish. This was undertaken with the aim of improving our knowledge of their physiology, metabolism and fate in the rainbow trout, Oncorhynchus mykiss. Phosphorous (P) featured strongly in this work due to its adverse role in pollution and the environmental impact of intensive fish farming. The first chapter surveyed the gross nutritional requirements of fish and focused on the mineral requirements in particular. Typically the P requirements for trout were found to range from 0.5-0.8% of the diet. The problems of P loading as a consequence of dietary loses was addressed and the physiological and metabolic roles of both calcium and phosphorous were especially noted in relation to fish health and for phosphorous its environmental implications were addressed. Experimental approaches were evaluated and it was decided to conduct both standard growth trial studies as well as digestibility trials to provide the basis of most investigations with the rainbow trout. Novel approaches and strategies were used in relation to specific experiments such as examining the mineral levels in blood, and various tissues and the testing of different feed ingredients, dietary supplements and mineral sources in successive investigations. Initial investigations appraised commercial diets of varying nutritional profile with respect to mineral retention and availability for rainbow trout under controlled laboratory conditions. The effects of diets containing different fishmeal sources: i.e. brown versus white fishmeals, elevated ash content and also varying in the levels of oil were tested on juvenile rainbow trout in closed recirculated systems. Diet composition caused a significant effect on mineral retention and distribution profile in fish tissues and organs. Typically, both P and Ca were of highest concentration in vertebrae of trout (60mg/g-dry weight), compared with P concentrations for all other major organs/tissues, which were fairly even between 11-20mg/g. A small increase in dietary P level (1.08% vs. 1.22%) did not affect any growth parameters for trout for the first two commercial feeds tested but there were interesting observations with respect to the amount of P excreted in the bile with a 25% increase from 0.8mM to 1.2mM. The P levels in plasma of these fish did not reflect any dietary changes. However, there was a noticeable reduction in the digestibility of P in the diet containing the white fishmeal source (26%) compared with 49% for the higher grade fishmeal diet. High ash content feeds resulted in a marked reduction in the net mineral retention of this element (16% compared to 27% for the lower ash diet). The same was also true for Ca (12% compared with 26%). The effect of oil levels in diets on mineral utilization was investigated under farmed conditions and was of particular interest given the demand for nutrient dense feeds in the industry. There was a strong tendency for improved P and Ca digestibility coefficients at each incremental increase in oil level for juvenile production sized fish (50-l00g). This ranged from 55% to over 70% when oil levels were over 26%. However this was not observed for larger fish of over 200g in weight. Experimental investigations followed are described in (Chapter 4) where fishmeal based diet was supplemented with varying levels of inorganic phosphorous. Phosphorous, calcium and other mineral absorption characteristics in addition to retention were measured in a series of growth and digestibility trials. Interestingly, there was no apparent change in the distribution of P with increasing dietary levels ranging from 1.39-2.16%. These were above known requirements for these fish with minerals being in excess. Similar results were noted for all other minerals measured in rainbow trout. There was a significant rise in the P concentration of plasma of rainbow trout fed a diet containing over 2% P and this may infer that the homeostatic regulation of P is unable to function at this level. Other haemato-logical parameters were not affected. Although not significant, there appeared to be slight trend in elevated bile P with increasing dietary P supplementation. The faecal concentrations for each of the minerals showed that elevated P in fish meal diets led to increased faecal output from 25mg/g to over 40mg/g for the highest P diet. Overall digestibility coefficients were lower as dietary P increased above that in the fishmeal control diet. These ranged from 50% to 39% for P, Ca and Mg were not greatly affected. The net retention of P was calculated and this fell from 30% to just below 20% for the range of dietary P used in the investigation. A preliminary study, reported in chapter 5A, was useful in providing information about the relative absorption profiles for differential mineral absorption from the various regions of the gastro intestinal tract of rainbow trout. A standard commercial diet was fed to large trout (>200g) and subsequently, digesta was removed from fish and analysed. For all minerals and protein, the pyloric and mid intestinal region was the main site for digestion, release and absorption of the macro elements concerned. The protein and mineral digestibility of suitable feedstuffs commonly employed in the formulation of complete diets for trout resulting from a sequence of experimental trials are presented in chapter 5B. These included a selection of marine, animal and plant by-products which were substituted into a reference basal diet designed for salmonids. This involved the inert marker- yttrium oxide and calculations based on nutrient digestibility from diet and faecal concentrations. Mineral digestibility coefficients were found to vary considerably and a number of anomalies such as negative values were obtained for Ca and Zn in certain feedstuffs. Combined diets (reference and test ingredient) gave values that were more consistent and P digestibility ranged from 47-59% in marine and animal protein concentrates compared with plant sources (24-37%). Negative values for Ca and Zn were thought to be attributable to complex interactions with other feed components. Additionally, a group of inorganic mineral supplements were tested by inclusion into a series of diets. These included mono calcium phosphate, di-calcium phosphate (DCP), mono di-calcium phosphate and magnesium phosphate. DCP produced lower Ca and P digestibility values, 31 and 50% respectively, compared with the other sources (44-62%), indicating the importance of choice of mineral supplement in aquafeeds. A critical appraisal of this work is provided in Chapter 6 and formed a retrospective review of the results generated and integrated these findings into a foundation for further research and development. Nutritional investigations using the rainbow trout as the model salmonid species raised many more questions and possibilities and broadened the scope of the topic. The subject of mineral requirements for fish is very complex and numerous factors are involved at several physiological and biochemical levels in fish. Although the research on rainbow trout involved whole animal studies under both laboratory and commercial farm conditions, the need to explore alternative in vitro methods and to utilize larger scale farm and sea cage trials for salmon were suggested. The advent of more advanced diet formulations and feeding strategies were mentioned and the scope for more scientific investigations to improve the utilization and reduce phosphorous discharge into the environment proposed.