Open Collections

UBC Theses and Dissertations

UBC Theses Logo

UBC Theses and Dissertations

Astaxanthin in juvenile farmed Chinook salmon (Oncorhynchus tshawytscha) : effective dietary levels for… Thomas, Alison Carolyn 1999

Your browser doesn't seem to have a PDF viewer, please download the PDF to view this item.

Item Metadata

Download

Media
831-ubc_1999-0135.pdf [ 4.48MB ]
Metadata
JSON: 831-1.0089140.json
JSON-LD: 831-1.0089140-ld.json
RDF/XML (Pretty): 831-1.0089140-rdf.xml
RDF/JSON: 831-1.0089140-rdf.json
Turtle: 831-1.0089140-turtle.txt
N-Triples: 831-1.0089140-rdf-ntriples.txt
Original Record: 831-1.0089140-source.json
Full Text
831-1.0089140-fulltext.txt
Citation
831-1.0089140.ris

Full Text

ASTAXANTHIN IN JUVENILE FARMED CHINOOK SALMON (Oncorhynchus tshawytscha):  EFFECTIVE DIETARY LEVELS FOR FLESH PIGMENTATION AND INFLUENCE ON FATTY ACID PROFILE DURING COLD TEMPERATURE STORAGE OF FILLETS by Alison Carolyn Thomas B.Sc.(Agr.) The University of British Columbia, 1996 A THESIS SUBMITTED IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE in THE FACULTY OF GRADUATE STUDIES (Food Science) We accept this thesis as conforming to the required standard  THE UNIVERSITY OF BRITISH COLUMBIA January, 1999 © Alison Thomas, 1999  In  presenting  degree  at  this  the  thesis  in  partial  fulfilment  of  University  of  British  Columbia,  I agree  freely available for copying  of  department publication  this or of  reference  thesis by  this  for  his  and study. scholarly  or  thesis  her  for  purposes  financial  of  VoccX  The University of British Vancouver, Canada  Date  DE-6 (2/88)  Vcb  glVSt^  Scig,i/\rV  Columbia  gain  requirements that  agree  may  representatives.  permission.  Department  I further  the  It  shall not  be is  that  the  Library  permission  granted  by  understood be  for  allowed  an  advanced  shall for  the that  without  make  it  extensive  head  of  my  copying  or  my  written  ABSTRACT Astaxanthin (Ax) is the pigment responsible for the characteristic flesh colour of salmonids and must be present in the diet of farmed salmon in order for them to attain a colour acceptable to consumers. These studies were conducted to investigate the potential for early pigmentation in chinook salmon, to determine an effective dietary concentration of Ax for juvenile chinook, and to examine possible antioxidative activity of Ax in salmon fillets stored under aerobic and anaerobic conditions at -3°C. In Experiment 1, all-female chinook salmon smolts with a mean initial body weight of ~40g were fed a diet containing no added Ax (control), or one of three diets containing low (3 mm pellet, 29.56 ppm Ax; 4 mm pellet, 25.73 ppm Ax), medium (3 mm pellet, 40.11 ppm Ax; 4 mm pellet, 39.60 ppm Ax) or high (3 mm pellet, 76.42 ppm Ax; 4 mm pellet, 59.82 ppm Ax) synthetic Ax. Fish were fed the 3 mm pellet for the first three monihs and the 4 mm pellet for the remainder of the study. At three and seven months, 24 fish per treatment were sacrificed and analyzed for colour using the Roche Colour Card for Salmonids, a Hunter Lab Labscan, and quantification of flesh Ax concentrations by high performance liquid chromatography. ResultsfromExperiment 1 showed that chinook salmon weighing < 120 g were capable of attaining a high degree of visible pigmentation. In general, there was not a good direct relationship between fish size andfleshlevel of Ax. The concentration of Ax in the diet had a significant effect onfleshcolour and Ax content, but not the proximate composition of the fillet. Time was not a significant factor in pigmentation, however the time*diet interaction was significant. Good correlations (r > 0.8) were found between flesh levels of Ax and the Hunter Lab a, chroma, and Roche Colour Card scores.  In Experiment 2, pigmented and non-pigmentedfilletswere stored at -3°C for up to 6 weeks under aerobic and anaerobic conditions. Methyl esterified lipid extracts from allfilletswere analyzed via gas chromatography to assess changes in fatty acid composition attributed to oxidation. Analysis of the ratios of unsaturated to saturated fatty acids and polyunsaturated to saturated fatty acids from Experiment 2 revealed no significant difference due to pigmentation of fillets. This indicates that astaxanthin had neither antioxidative nor prooxidative activity under the conditions used in the present study.  iii  T A B L E OF CONTENTS  Abstract  ii  Table of Contents  iv  List of Tables  vii  List of Figures  viii  Dedication  xi  1 Introduction  1  2 Literature Review  3  2.1 Chemistry of Astaxanthin  3  2.2 Pigment Sources  5  2.3 Factors Affecting Pigmentation  6  2.3.1 2.3.2 2.3.3 2.3.4 2.3.5  Dietary Astaxanthin Dietary Lipid Level Genetic Influence Fish Size Other Factors Influencing Pigmentation  6 7 7 8 9  2.4 Flesh Colour Determination 2.4.1 Roche Colour Card for Salmonids 2.4.2 Colourimetric Analysis 2.4.3 Quantification of Astaxanthin  9 10 10 13  2.5 Antioxidative Functions of Astaxanthin 2.5.1 In Vitro Studies 2.5.2 In Vivo Studies  14 15 17  3 Experiment 1 3.1 Materials and Methods 3.1.1 Fish History and Pre-Experimental Period 3.1.2 Experimental Design and Fish Distribution 3.1.3 Diets 3.1.4 Feeding 3.1.5 Sampling  19 19 19 21 23 24 24  iv  3.1.6 Analysis of Fillets 3.1.6.1 Qualification of Colour Using the Roche Colour Card for Salmonids 3.1.6.2 Colourimetric Measurements 3.1.6.3 Quantification of Astaxanthin 3.1.6.4 Proximate Composition of Fillets and Diets 3.2 Calculation of Performance Parameters and Statistical Procedures 3.3 Results 3.3.1 Proximate Analysis and Astaxanthin Content of Diets 3.3.2 Fish Mortality 3.3.2.1 Period 1 3.3.2.2 Period 2 3.3.3 Feed Intake Growth and Feed Efficiency 3.3.3.1 Period 1 3.3.3.2 Period 2 3.3.4 Fillet Composition 3.3.5 Colour Measurements  25 25 26 28 30 30 32 32 34 34 34 35 35 35 36 36  3.4 Discussion  65  3.5 Conclusions  76  4 Experiment 2  78  4.1 Materials and Methods 4.1.1 Experimental Design 4.1.2 Analysis of Fillets 4.1.2.1 Colourimetreic Measurements 4.1.2.2 Lipid Analysis 4.1.2.3 Methyl Ester Preparation and Analysis 4.2 Statistical Analysis  78 78 79 79 79 80 81  4.3 Results 4.3.1 Colour Measurements 4.3.2 Lipid Analysis  83 83 83  4.4 Discussion  86  v  4.5 Conclusions  88  5 General Conclusions  90  6 References  92  Appendix 1 - Raw Data for Fillets from Experiment 1  98  Appendix 2 - Statistical Analyses for Experiment 1  103  Appendix 3 - Fillet Fatty Acid Profiles for Experiment 2  106  Appendix 4 - Raw Data for Fillets from Experiment 2  112  Appendix 5 - GC Profile Examples for Experiment 2  113  Appendix 6 - Statistical Analyses for Experiment 2  114  vi  List of Tables  Page  Table 1- Time Line for Experiment  20  Table 2 - FfPLC Conditions Used for Quantification of Astaxanthin  29  Table 3 - Levels of Moisture, Ash, Crude Lipid (CL), Crude Protein (CP), All-£astaxanthin [Ax] and Corrected Astaxanthin in the Test Diets (dry weight basis except moisture)  33  Table 4 - Feed Intake, Growth and Feed Efficiency of Chinook Salmon Fed Diets Containing Low, Medium (Med.), High, or No (Ctrl) Supplemental Astaxanthin for Three Months (Period One)  37  Table 5 - Partial Proximate Analyses of Fillets from Chinook Salmon Fed Diets Containing Low, Medium (Med.), High, or No (Ctrl) Supplemental Astaxanthin for Three or Seven Months  37  Table 6 - Measures of Hunter Lightness (L), Hunter Redness (a), Hunter Yellowness (b), the Roche Colour Card for Salmonids (RCC), Flesh Concentration of Astaxanthin ([Ax] ppm), Hue Angle, and Chroma of Fillets from Chinook Salmon Fed Diets Containing Low, Medium (Med.), High, or No (Control) Supplemental Astaxanthin for Three or Seven Months  44  Table 7 - Correlations Between Hunter Lightness (L), Hunter Redness (a), Hunter Yellowness (b), the Roche Colour Card for Salmonids (RCC), Flesh Concentration of Astaxanthin ([Ax] ppm), Hue Angle, and Chroma as Colour Scoring Methods Used to Assess Chinook Salmon Fillets  45  Table 8 - Correlations between Fish Weight and Concentration of Flesh Astaxanthin for Chinook Salmon Fed Diets Containing Low, Medium (Med.), High, or No (Ctrl) Supplemental Astaxanthin for Three or Seven Months  45  Table 9 - GC Conditions Used for Analysis of Methyl Esters  82  Table 10 - Hunter L (lightness), a (redness), and b (yellowness) Values of Pigmented and Unpigmented Chinook Salmon Fillets stored in Oxygen Permeable (aerobic) or Impermeable (barrier) Bags for Zero to Six Weeks at -3°C  84  Table 11 - Percentages of Crude Lipid and Ratios of Unsaturated to Saturated Fatty Acids (U:S), and polyunsaturated to Saturated Fatty Acids (P:S) of Pigmented and Unpigmented Chinook Salmon Fillets stored in Oxygen Permeable (aerobic) or Impermeable (barrier) Bags for Zero to Six Weeks at -3°C  85  vii  List of Figures Figure 1 - Chemical Forms of Astaxanthin  Page 4  Figure 2 - Mechanism for Free Radical Trapping by Carotenoids  16  Figure 3 - Experimental Tank Setup  22  Figure 4 - Schematic Representation of Hue Angle  27  Figure 5 - Hunter L, a, b, Colour Solid Showing Fillets from Chinook Salmon Fed Diets Containing Low (square), Medium (diamond), High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  46  Figure 6a - Relationship Between Hunter L and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low (square), Medium (diamond), High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  47  Figure 6b - Regression Plot for Hunter L and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  48  Figure 6c - Regression Plot Using Spearman's Rank Order Correlation for Hunter L and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  49  Figure 7a - Relationship Between Hunter a and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low (square), Medium (diamond), High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  50  Figure 7b - Regression Plot for Hunter a and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  51  Figure 7c - Regression Plot Using Spearman's Rank Order Correlation for Hunter a and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  52  viii  Figure 8a - Relationship Between Hunter b and Concentration of Flesh Astaxanthin (Ax (ppm)) for FilletsfromChinook Salmon Fed Diets Containing Low (square), Medium (diamond), High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  53  Figure 8b - Regression Plot for Hunter b and Concentration of Flesh Astaxanthin (Ax (ppm)) for FilletsfromChinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  54  Figure 8c - Regression Plot Using Spearman's Rank Order Correlation for Hunter b and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  55  Figure 9a - Relationship Between Hue Angle and Concentration of Flesh Astaxanthin (Ax (ppm)) for FilletsfromChinook Salmon Fed Diets Containing Low (square), Medium (diamond), High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  56  Figure 9b - Regression Plot for Hue Angle and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  57  Figure 9c - Regression Plot Using Spearman's Rank Order Correlation for Hue Angle and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  58  Figure 10a - Relationship Between Chroma and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low (square), Medium (diamond), High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  59  Figure 10b - Regression Plot for Chroma and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  60  Figure 10c - Regression Plot Using Spearman's Rank Order Correlation for Chroma and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  61  ix  Figure 1 la - Relationship Between Roche Colour Card Scores (RCC) and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low (square), Medium (diamond), High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  62  Figure 1 lb - Regression Plot for Roche Colour Card Scores (RCC) and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  63  Figure 11c- Regression Plot Using Spearman's Rank Order Correlation for Roche Colour Card Scores (RCC) and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  64  Figure 12 - HPLC Profile Typical of Diet Extract (Medium Ax Diet)  66  Figure 13 - HPLC Profile Typical of Chinook Salmon Flesh Extract (High Ax Diet)  68  Figure 14 - Astaxanthin-actomyosin Complex Model  69  x  DEDICATION  In loving memory of Christopher Wayne Dallas, February 19, 1971 - June 29, 1998 An avid fisherman. My 'other brother'. May you rest with the greatest fisherman of all.  1 INTRODUCTION  Adequate pigmentation of cultured salmon flesh is of great concern to the aquaculture industry. The distinctive colour of salmon is one of the most important factors in consumer identification and acceptance of the product, along with other traits such as fillet size and shape. Flesh colour is considered indicative of the quality of salmon products and is, therefore, a determinant of marketability and price. Salmonid (salmon, trout and charr) flesh colour is a result of the deposition of specific carotenoids in the tissues of the fish. Scientists have identified sixteen carotenoids present in the flesh, skin, gonads and other tissues of salmonids, most being distributed in the muscle. More than 90% of the carotenoids in the muscle during the pre-spawning period is astaxanthin (Kitahara, 1983). Since salmonids are incapable of de novo synthesis and interconversion of carotenoids, farmed salmon must be fed a diet containing astaxanthin and/or canthaxanthin (the only other carotenoid which has been demonstrated to impart a similar flesh colour) in order to achieve the same desirable colouration as wild counterparts. Astaxanthin can be *  added to the diet through inclusion of ingredients where it is a naturally occurring compound, such as crustacean products, certain strains of algae, bacteria, and Phaffia yeast, or by addition of commercially produced synthetic astaxanthin (Carophyll® Pink, Hoffmann-La Roche, Basle, Switzerland). Due to the high cost of incorporating astaxanthin into salmonid feeds, various factors affecting the pigmentation of salmonids have been, and continue to be, studied. Such factors include; dietary source, level of inclusion and chemical form of astaxanthin, digestion, absorption and deposition of astaxanthin in the salmon, genetics, size, and environmental  1  influences. Post harvest chemical and colour stability of astaxanthin are also of interest to the food industry. Preventing oxidation of the pigment during storage by controlling environmental conditions such as exposure to light, heat and oxygen as well as the potential antioxidant activity of astaxanthin are active areas of research. The following study was undertaken to investigate several of these areas of interest as they apply to juvenile chinook salmon. In experiment 1, it was hypothesized that with supplemental astaxanthin in the diet, chinook salmon would be capable of early pigmentation and that resultant flesh levels of astaxanthin would reflect the dietary concentration. A further objective was to determine an effective dietary concentration of astaxanthin for colouring the flesh of juvenile chinook salmon. In experiment 2 it was postulated that astaxanthin would be effective as an antioxidant in chinook salmonfilletsstored at -3° C .  2  2 LITERATURE REVIEW  2.1 Chemistry of Astaxanthin Astaxanthin (3,3'-dihydroxy-4,4'-diketo-P-carotene) can be present in various chemical forms (Figure 1), (ie., free astaxanthin, astaxanthin mono- and diesters, and astacene which is an artifact, or breakdown product, of astaxanthin (Torrissen et al, 1989)). Free astaxanthin is deposited in the flesh of salmonids (Schiedt etal, 1981; Storebakken etal., 1985). Sincefreeastaxanthin has two chiral centers, at C-3 and C-3', it occurs in three stereoisomeric forms; the enantiomers 3S, 3'S and 3R,3'R, and the meso form 3'R, 3S. No epimerization occurs at the chiral centers and all three stereoisomers are equally effective at pigmenting salmonid flesh (Foss et al., 1984). All forms of free astaxanthin are deposited as consumed. Therefore, the ratio of deposited free astaxanthin enantiomers reflects that of the ingested enantiomers (Foss et al., 1984; Whyte et al., 1998). Astaxanthin contains 13 conjugated double bonds and, therefore, may be present in a variety of geometrical stereoisomers including an all-E-form, and several Z-isomeric forms. The Z-isomeric forms of astaxanthin have a lower bioavailability than the all-£-form (Bjerkeng et al, 1997b). Synthetic astaxanthin contains -75% all-£-astaxanthin and 25% Z-isomers of astaxanthin (Bjerkeng et al, 1997b). Astacene, in the free and esterified forms, is incorporated into the flesh and skin with an efficiency similar to that of astaxanthin (Simpson et al, 1981). Natural fatty acid esters of astaxanthin can also be used as a pigment source, but have been shown to be less effective at pigmenting rainbow trout (Torrissen & Brakken, 1979). Hydrolysis of the ester bond in the digestive tract is indicated by the finding that only free astaxanthin is found in the flesh and in the plasma (Sivtseva, 1982; Torrissen et al, 1989). Cleavage of astaxanthin esters  3  Fieure 1 - Chemical Forms of Astaxanthin  to free astaxanthin is a rate limiting step for pigmentation (Storebakken et al, 1987). This may account for the lower efficiency of absorption of astaxanthin esters compared to free astaxanthin. Astaxanthin dipalmitate, with an apparent digestibility of about 47%, was shown to be used less efficiently than free astaxanthin which had an apparent digestibility of about 64% (Storebakken et al, 1987).  2.2 Figment Sources Various sources of astaxanthin are used in the production of commercial salmonid diets. Use of crustacean waste products in cultured salmonid diets dates back to the early days of pisciculture in Norway (Torrissen et al., 1989). Various species, including shrimp (Torrissen etal., 1981; Choubert & Luquet, 1983), euphausids (Kotik etal, 1979), and crab (Spinelli et al. 1974; Spinelli & Mahnken, 1978), have been found to be useful as pigment sources. Crustacean waste seems an ideal source considering the large quantities available, however, without extensive processing, the low protein and high chitin content typical of these products necessitates restrictions on the amount which can be included in diet formulations. Astaxanthin in crustacean products is found in the free and esterified form as well as in protein complexes (Torrissen et al, 1989). Phaffia rhodozyma is a red yeast which naturally synthesizes astaxanthin during its life cycle and has been used successfully to colour salmonid flesh (Johnson & An, 1991; Johnson etal, 1980; Johnson etal, 1977; Sanderson & Jolly, 1994). Astaxanthin in Phaffia is present only in the 3R,3'R configuration (Andrewes & Starr, 1976) and, if handled properly, has a bioavailability comparable to that of synthetic astaxanthin (Torrissen & Christiansen, 1994). Several types of algae have also been shown to produce astaxanthin (Torrissen et al, 1989). Since pure crystalline astaxanthin is very  sensitive, deteriorating in the presence of light, heat or oxygen, synthetic astaxanthin (Carophyll® Pink) is stabilized in the form of a water soluble beadlet (8% astaxanthin emulsified in ascorbyl palmitate and ethoxyquin, surrounded by a protective matrix with a gelatin and maize starch outer layer). Synthetic astaxanthin is a racemic mixture of the three configurational stereoisomers (3S, 3'S: 3'R, 3S: 3R, 3'R) (Whyte etal, 1998).  2.3 Factors Affecting Pigmentation 2.3.1 Dietary Astaxanthin  The amount of available astaxanthin in the diet of cultured salmonids has been shown to be one of the most important determinants of final flesh colour. Increasing levels of astaxanthin in the diet leads to increased levels of astaxanthin in the flesh (Torrissen, 1985; Storebakken etal, 1987; Choubert & Storebakken, 1989; March & MacMillan, 1996), however, the efficiency of astaxanthin deposition decreases with increasing concentrations in the diet (Torrissen, 1985; Choubert & Storebakken, 1989; Storebakken & Choubert, 1991; Smith et al,,1992). Currently, commercial salmonid diets contain astaxanthin in the range of 30-100 ppm. Research into the optimal dietary dosage required has been conducted on various species, however, an absolute value is difficult to assign because of the influence of other factors (March & MacMillan, 1996) including, but not limited to, genetics, diet composition, length of feeding period, rearing facilities, fish health and environment (Torrissen etal, 1989).  6  2.3.2 Dietary Lipid Level  The effects of the fat content of salmonid diets on astaxanthin deposition have received some attention since astaxanthin is fat soluble and its absorption is postulated to be enhanced by increased levels of fat (Choubert & Luquet, 1983). One must consider, however, the potential for decreased consumption of high fat diets due to their increased caloric density which would, presumably, necessitate higher astaxanthin concentrations in the diet in order to achieve the same overall astaxanthin intake. The reported effects of fat content of the diet on astaxanthin deposition in salmonids are varied. While Spinelli (1979), Torrissen (1985), and Bjerkeng etal. (1997a) found that increasing lipid levels improved absorption, Choubert & Luquet (1983) found that astaxanthin absorption was not enhanced.  2.3.3 Genetic Influence  Genetically determined differences in pigmentation between salmonid species have been long recognized, and are well documented (Torrissen etal., 1989). Differences also occur between different strains within a species, and genetic variation has been described for rainbow trout (Gjerde & Gjedrem, 1984; Torrissen & Naevdal, 1984; Torrissen, 1985), Atlantic salmon (Gjerde & Gjedrem, 1984; Torrissen & Naevdal, 1988), coho salmon (Iwamoto et al., 1990), and chinook salmon ( McCallum et al, 1987). Chinook salmon show differences in pigmentation among different sections of muscle, the tissue being consistently more highly pigmented in the midsection as compared to the anterior and distal sections, with the distal section being the least pigmented (March & MacMillan, 1996). Chinook salmon are unique because the fish develop one of two distinct muscle colours, becoming either redfleshed or white-fleshed; a trait under strong genetic control (Withler, 1987). Refsgaard et al.  7  (1998) demonstrated the extent of biological variation within a relatively large subpopulation of farmed Atlantic salmon as it applied to the concentration of lipid and lipid components including astaxanthin. A very wide range was found and the biological variation coefficients were determined to be -25%, demonstrating a surprising lack of uniformity in what would be expected to be a nominally homogeneous population.  2.3.4 F i s h Size  Fish size is thought to have an influence on salmonid pigmentation. Torrissen (1985) reported a direct linear relationship between fish weight and astaxanthin levels in the flesh, but concluded that fish size had no effect on astaxanthin deposition. The positive relationship was instead ascribed to the greater amount of pigment ingested by the larger fish and not their weight per se. It seems that fish below a certain weight do not deposit astaxanthin in the flesh although it is present in the skin. The reason for this observation is unknown but is more likely related to physiological status than body weight (Torrissen, 1985). The weight for commencement of pigment deposition in the flesh has been reported as 80-90 g and even up to 150 g for chinook salmon and rainbow trout, and 200-400 g for Atlantic salmon (Torrissen et al., 1989). In a study on the pigmentation of rainbow trout from start feeding to sexual maturation, Bjerkeng et al. (1992) noted that the fish deposited 'little' astaxanthin in the flesh (< 3 ppm) prior to reaching -100 g (50 weeks of feeding, 100 ppm astaxanthin in the diet), while deposition in the skin was preferred. Between body weights of 100-200 g (49-58 weeks), there was a strong increase in astaxanthin deposition in the flesh, with efficiency decreasing after that weight range until maturation.  8  2.3.5 Other Factors Influencing Pigmentation  Rearing and environmental conditions may also have effects on salmonid pigmentation. Rainbow trout raised infreshwater showed higher astaxanthin content in the flesh than those raised in salt water (Storebakken & Choubert, 1991). The effects of water temperature were studied by No & Storebakken (1991a). Although fish raised at a higher temperature grew and pigmented faster than those raised at a lower temperature, no difference was seen when fish of the same weight were compared. This information supports the proposal of a linear relationship between fish weight and flesh pigmentation. It is most likely that temperature effects on pigmentation are a result of effects on feed consumption. Health-related factors, or anything resulting in reduced feed intake and/or conversion, would be expected to have similar results. Season of the year also has an influence on pigmentation; the astaxanthin content of salmonid flesh decreasing in the spring (Torrissen & Christiansen, 1994). Refsgaard et al. (1998), for example, found that salmon harvested from the same population, had significantly higher astaxanthin contents in January than those sampled in May.  2.4 Flesh Colour Determination  Since humans have relative difficulty in objectively discerning between shades of a colour, it is necessary to employ standard methods of colour analysis to qualify or quantify the degree of pigmentation of salmonid flesh. Various methods, including colour cards, spectral characteristics and tristimulus colour measurements, are used for quantification of colour. Solvent extraction coupled with high performance liquid chromatography or  (  spectrophotometry are employed to quantify astaxanthin levels.  9  2.4.1 Roche Colour Card for Salmonids.  The Roche Colour Card for Salmonids has become an industrial standard for use in assessing fish flesh colour at slaughter. A colour score of 16 is generally accepted in most markets, however a 17 is required for acceptability in Japan (Mann, 1999). Though easy to use, this method has numerous shortcomings. For instance, the colour card is a subjective measure, and both the extent and type of background illumination, as well as the colour perception of the grader, can influence the results. While it is accepted that a wide range of pigment levels can be matched to a particular colour tile (King, 1996), a study by Christiansen etal. (1995) on Atlantic salmon, found a good linear regression between the Roche Colour Card score and meanfleshastaxanthin, the coefficient of variation decreasing for the higher scores. It must be noted that the Roche Colour Card for Salmonids was developed for use on Atlantic salmon. Hence, due to species differences, it is possible that the colour card score may not correlate as well with the astaxanthin concentration in the flesh of Pacific salmon species.  2.4.2 Colourimetric Analysis  The following information on colour vision and the evolution of colour scales is a highly abridged summary of pertinent points covered in the first four chapters of Measuring Colour (Hunt, 1987). Since colour vision at normal levels of illumination is essentially a function of three types of cone in the retina, various systems used in industry for describing colour are based on trichromatic (three colour) matching. The p, y, and P cones differ in their sensitivity to different wavelengths of light, the p cones having maximum sensitivity in the yellow-orange  region of the spectrum, the y cones in the green region, and the P cones in the blue-violet region. Unfortunately, knowledge of the spectral sensitivity curves of the three types of cone is not sufficiently precise to allow for their direct use in evaluating colour. As a result, experimental matching is done by adjusting amounts of precise colours of red, green, and blue light in one half of a field of view, to match the colour of a test sample in the other half of the field of view. Matching can be achieved in this fashion owing to the presence of only the three spectrally different cones in the human eye. Several systems used to describe colour have evolved from trichromatic colour matching. The initial system, the tristimulus R, G, B values, represents the amounts of the three colour matching stimuli as expressed in their respective units. This method occasionally results in negative values, describing the situation where a colour is added to the test sample in order to create a colour match (ie. if the test sample is yellow). The X,Y, Z tristimulus values were developed to avoid the need to use negative numbers, and are based on the R, G, B values; X = 0.49R + 0.31G+0.20B Y = 0.17697R + 0.81240G + 0.01063B Z = 0.00R + 0.01G + 0.99B. Since the chromaticity diagram developed from X, Y, Z tristimulus values does not give an indication as to the magnitude of the values and is only applicable to colours having the same luminance (brightness), the CIELAB colour space was developed. This colour space uses the parameters (L*) lightness, (a*) red/green chromaticity, and (b*) yellow/blue chromaticity. These parameters are as follows:  L*  116(Y/Y„)  L*  903.3(Y/Y„)  a* = 500[(X/X )  - 16  m  n  for Y/Y„ > 0.008856 for Y/Y„ < 0.008856  - (Y/Y„)  1/3J  b* = 200[(Y/Y„) - (Z/Z„) ] 1/3  1/3  These points are plotted on three axes at right angles to one another, where X„, Y„, Z„ are those values for the chosen reference white, and if any of (X/X„), (Y/Y„), or (Z/Z„) is less than or equal to 0.008856 it is replaced by 7.787F + 16/116, F being (X/X„), (Y/Y„), or (Z/Z„). The CJJ2LAB system is designed to be used for comparing differences in objects of the same size and shape under identical viewing conditions. The Hunter Lab system uses the variables L (lightness), a (red/green chromaticity), and b (yellow/blue chromaticity) and is also based on a relationship to X, Y, Z, tristimulus values where, L = 10Y  1/2  a= 17.5(1.02X-Y)*Y  _1/2  b = 7.0(Y - 0.847Z)*Y"  1/2  for standard daylight (S ) c  Since both the L*, a*, b*, and L, a, b, systems are based on the tristimulus values, but are derived using different formulas, we can only compare general trends from the two systems but not the actual values. The use of objective instrumental systems of analysis in assessing the colour of salmonidfleshhas been explored. Both the CJJELAB (L*, a*, b*) and Hunter Lab (L, a, b,) systems measure the colour of the sample in terms of the primary colours, red-green (+a, -a) and yellow-blue (+b, -b), as well as of lightness (0 = black, 100 = white). The a* (a) value has been identified as a good measure of colour preference in sockeye salmon before and after canning (Schmidt & Idler, 1958), as well as in raw and cooked Atlantic, and canned Pacific  12  salmon (Saito, 1969). It has also been shown to correlate well with the actual pigment concentration in raw coho salmon (Smith et al, 1992), raw rainbow trout (No & . Storebakken, 1991b), and with the flesh colour of Atlantic salmon before and after processing (Skrede & Storebakken, 1986a, 1986b; Christiansen etal, 1995). Although hue, or relation of yellowness to redness (H° = tan' (b*/a*) or tan^b/a); H° = 0 for red, H° = 90 for 1  ab  ab  ab  yellow), has been reported to be the best predictor of processed flesh colour from raw flesh (Schmidt & Idler, 1958; Skrede & Storebakken, 1986b), Christiansen etal (1995) reported that the hue of raw flesh had a low correlation with the actual astaxanthin concentration. The L* (L) value decreases with increasing astaxanthin concentration (No & Storebakken, 1991b; Skrede & Storebakken, 1986a; Smith et al, 1992; Christiansen et al, 1995). Skrede & Storebakken (1986a) found that the b* (b) value is not significantly influenced by the astaxanthin concentration, however, No & Storebakken (1991b), Smith et al. (1992), and Christiansen et al. (1995) found that it increased with increasing concentrations of astaxanthin. Little et al. (1979) showed that multiple regression analyses were necessary in order for colour system parameters to match visual assessments of raw flesh colour in rainbow trout. King (1996) also found that the use of regressions resulted in better correlations of L, a, b and hue values, with actual carotenoid content in rainbow trout.  2.4.3 Quantification of Astaxanthin  Chemical methods for quantifying astaxanthin vary considerably, but are generally based on a solvent extraction with acetone (Bjerkeng et al, 1997a; March & MacMillan, 1996; Ando et al, 1992; Choubert et al, 1992; No & Storebakken, 1991; Skrede & Storebakken, 1986b; Foss etal, 1984; Torrissen & Naevdal, 1984; Kotikera/., 1979;  13  Johnson et al., 1977) or chloroform/ methanol (Kiessling et al., 1995; Bjerkeng et al., 1997b) followed by chromatographic (HPLC, TLC, GC) or spectrophotometric identification and quantification. Standardization of the method(s) used would be of benefit to research in this area. In the case of spectrophotometric quantification, extinction coefficients used to quantify astaxanthin in acetone range from as low as 1600 (Johnson et al., 1977), to as high as 2500 (Torrissen & Naevdal, 1984), in addition to the use of an average extinction coefficient (1900) for both astaxanthin and canthaxanthin when the two are being quantified simultaneously (Skrede & Storebakken, 1986b). In the case of chromatographic quantification, the identification of astaxanthin isomers, esters, and artifacts are seldom mentioned and, presumably, are not taken into account. This is a result of the difficulty in obtaining pure standards for these compounds.  2.5 Antioxidative Functions Of Astaxanthin  Lipid oxidation is a major cause of deterioration and loss of quality in salmon products and is, therefore, of concern to the industry. Autooxidation of fats occurs via a free radical chain reaction which is commonly represented in a simplified three step mechanism as follows, 1) Initiation Initiator —» Free radicals (R., ROO.) 2) Propagation r—• R. + 0 - » R O O . 2  ROO. + RH -> ROOH + R.  3) Terrnination R. + R. —» Nonradical products R. + ROO. —> Nonradical products ROO. + ROO. -> Nonradical products (Adapted from Nawar, 1996) An antioxidant preserves food by retarding deterioration, rancidity, and/or discolouration resulting from oxidation reactions. As an antioxidant, astaxanthin is thought to function via two mechanisms; 1) as a quencher of singlet oxygen, preventing initiation of oxidation, and 2) as a trapper of radicals, preventing propagation of lipid oxidation by breaking the chain reaction.  2.5.1 In Vitro Studies  The potential effectiveness of carotenoids, including astaxanthin, as antioxidants is an active area of research. Burton and Ingold (1984) first described a free radical chain breaking reaction where P-carotene functions as a radical trapping antioxidant. P-carotene seems to trap the hydroxyl radical by adding it to the conjugated double bond (CDB) system (Figure 2). Since the unpaired electron is delocalized in the CDB system, the resultant radical is resonance stabilized and will not participate in the free radical chain reaction. The chain-breaking antioxidant capability of P-carotene should, theoretically, be exhibited by other compounds, such as astaxanthin, with extensive systems of CDB's. The protection afforded by carotenoids in vivo is a function of the number of CDB's. The quenching ability of carotenoids increases with the number of CDB's, but varies with chain structure and functional group (Hirayama et  15  Figure 2 - Mechanism for Free Radical Trappine by Carotenoids  1  CR,  ROO.  AdaptedfromBurton & Ingold (1984)  al, 1994). Carotenoids with 9, 10, and 11 CDB's have about a threefold higher level o f activity compared to those with 8 or less CDB's, (Bermond, 1990). Astaxanthin has thirteen CDB's. I t has been shown that those carotenoids having oxo groups at the 4 and 4' positions on the 3-ionone ring, such as astaxanthin, are more effective as antioxidants and more resistant to autocatalytic radical chain reaction than other carotenoids containing the 3-ionone ring system (Palozza & Krinsky, 1992; Terao, 1989). It is postulated that the electron withdrawing character o f the additional oxygen atoms decreases the unpaired electron density on the CDB system resulting in decreased reactivity o f the carotenoid radical (Terao, 1989). The antioxidative effectiveness o f astaxanthin, compared to other known antioxidants, appears to be in question. I n 1991, M i k i reported that, as a free radical scavenger in model systems, astaxanthin has five times the cellular antioxidative activity o f P-carotene, and 15 times that o f vitamin E. As a quencher o f singlet oxygen, astaxanthin is reported to have about 100 times the activity o f vitamin E. Palozza & Krinsky (1992) found astaxanthin to be as effective as vitamin E , and much more effective than P-carotene, as a radical-trapping antioxidant in a membrane model. I n two more recent works, however, astaxanthin was reported to be less reactive as an antioxidant compared to vitamin E (Mortensen & Skibsted, 1997) and p-carotene (Mortensen & Skibsted, 1997; Woodall et al, 1997) and was therefore considered less effective.  2.5.2 In Vivo Studies Several papers have reported on the effectiveness o f astaxanthin in delaying in vivo lipid peroxidation o f fishery products during storage, with varying conclusions. Bjerkeng & Johnsen (1995) concluded that astaxanthin did indeed possess antioxidative activity based on  17  thiobarbituric acid reactive substance (TBARS) analysis of rainbow trout (filleted and deskinned) representing three different levels of astaxanthin (high, medium, and low) stored for 17, 29, and 36 weeks at -18°C, in a variety of packaging materials, in both illuminated and dark cabinets. By contrast, Sigurgisladottir et al. (1994) reported that astaxanthin was ineffective as an antioxidant in the flesh of Atlantic salmon fed diets with synthetic astaxanthin at 88.6 ppm, or 84.2 ppm with additional tocopherols, for 4, 7, 10, and 15 weeks, as determined by forced oxidation of the muscle. Astaxanthin in the flesh of rainbow trout fed a diet with 60 ppm synthetic astaxanthin at 2% body weight/ day for ten weeks, was also reported to be ineffective as an antioxidant, during storage of the trout flesh for 16 and 33 weeks at -15°C (Ingemansson et al. 1993). Astaxanthin in salmonid eggs is thought to protect against light, elevated temperatures, low oxygen tension and metabolic products from adult fish (Craik, 1985). Torrissen (1984), however, found no effects of astaxanthin levels on Atlantic salmon egg or alevin survival. In addition, he noted that light sensitivity increased with increasing astaxanthin in the eggs.  18  3 EXPERIMENT 1 Experiment 1 proceeded as per the time line shown in Table 1.  3.1 Materials and Methods 3.1.1 Fish History and Pre-Experimental Period Approximately two thousand female Big Qualicum chinook salmon smolts (Onchorhynchus tshawytscha) were acquired on May 14, 1996fromOmega Pacific Hatcheries, Port Alberni, B.C., and subsequently maintained at Fisheries and Oceans Canada, West Vancouver laboratory (WVL). Two outdoor 4000 L circular fiberglass tanks, supplied with ambient temperature seawater pumped from Burrard Inlet, were used to house the fish. Water circulated around the tank and out through a central standpipe in a non-recirculating system. Additional oxygenation of the seawater was provided by passing the water through a one meter long, 12.5 cm PVC pipe filled with Koch rings prior to cascading into the tank. Air bars were also used in order to keep the dissolved oxygen levels as close to saturation as possible. At the hatchery, the stock fish (~ lOg) had been fed a commercial salmon diet (EWOS Vextra Smolt) which was supplemented with 60 ppm astaxanthin. Upon arrival at WVL, the fish were hand fed to observed satiety twice daily with the same diet. Terramycin Aqua® (Oxytetracycline, or, OTC) was administered at 1% body weight per day in the feed over a 10 day period (starting ~ June 5, 1996) as a prophylactic treatment against Bacterial Kidney Disease (BKD).  Table 1- Time Line for Experiment 1 Pretrial Period  Vextra Smolt  May 14, 1996 -June 5, 1996  - -2000 fish arrive at W V L - OTC treatment for 10 days  July 2, 1996  • Start of Experiment 1: -62 fish in each of 12 tanks -10 fish sacrificed (day 0 fish) • Fish anaesthetized with Marinil®, fiberglass inserts removed • First sampling, three months: - all fish anaesthetized with Marinil® & MS222®, weighed & measured - all tanks cleaned & fitted with fiberglass raceways - 12fish/tanksacrificed & filleted for analysis - all other fish returned to tanks  1  2  Period 1  3 mm diet  July 8, 1996 Oct. 1 & 2 , 1996  Period 2  4 mm diet  1  OTC treatment for 10 days Onset of tail rot outbreak One hour static bath in 8.5 ppm chloramine-T Nov. 8, 1996 High mortalities, one tank/treatment removed from experiment Nov. 18 -27, 1996 Vit. C & E supplementation for 10 days Dec. 13, 1996 One hour static bath in 6.0 ppm chloramine-T Jan. 23 & 24, 1997 Second sampling, seven months: - all fish killed with MS222®, weighed & measured - 12fish/tankfilletedfor analysis - remaining fish from control & high Ax diet groups filleted & packaged for Experiment 2 WVL = West Vancouver Laboratory, Fisheries and Oceans Canada, West Vancouver, B.C. OTC = Oxytetracycline Oct. 19 - 28, 1996 End of Oct., 1996 Nov. 7, 1996  20  3.1.2 Experimental Design and Fish Distribution  On July 2, 1996 all stock fish were anaesthetized according to the protocol of Krieberg and Powell (1991) viz., 0.25 ppm Marinil® for 15 minutes and, subsequently, 60 ppm MS222 ® for complete anaesthesia. Fish with weights in the range of 25 to 55g were selected randomly and distributed ten fish at a time (twelve in the last round) to each of twelve tanks, resulting in 62 fish per tank. Selected fish were rolled on a moistened absorbent cloth prior to weighing to the nearest 0.1 g and measuring forklengths (fish length from the tip of the nose to the fork of the tail) to the nearest 1 mm. The groups of fish (mean weight -40 g) were held in 12 outdoor 4000 L fiberglass tanks, the protocol for which is described previously, with the exception that the incoming water was not routed through the pipe containing the Koch rings but instead through three jets that were positioned to create a slow circulating current in each tank. Water velocity was increased until the fish started to school (~0.5 body lengths/second). Each of the four dietary treatments (control, low, medium, and high Ax diets) were randomly assigned to triplicate groups of fish. Of the remaining stock fish, 10 were sacrificed as representative of day 0, while the rest were maintained on the control diet for the duration of the experiment. Swimming channels (-45 cm wide, 55 cm deep; Figure 3) were created by means of a fiberglass insert (Kiessling et al., 1994) to minimize aggression between fish by forcing them to school. These inserts were removed on July 8, 1996, since the fish were small enough to swim through the notches at the bottom of the insert, and into the center of the tank. The swimming channels were placed back into the tanks for period 2. During period 1, water temperature and salinity ranged from 9.9 to 13.5°C , and 28 to 33 ppt respectively, whereas during period 2, water temperature and salinity ranged from 6.3 to 11.7°C , and 29 to  21  Figure 3 - Experimental Tank Setup  1  34 ppt respectively. Dissolved oxygen levels were kept near 100% throughout the experiment. The natural photoperiod was maintained (49° 15' N , 123° 10' W).  3.1.3 Diets  Each of the four extruded diets were manufactured by a commercial feed manufacturer twice during the study. The diets were formulated to differ only in their supplemental astaxanthin content, 0, 35, 55 or 75 ppm added level, designated as the control, low, medium, and high Ax diets, respectively. Carophyll® Pink (Hoffman-La Roche, Basel, Switzerland) was used as the source of supplemental astaxanthin. Thefirstbatch of each diet (3 mm pellet) was fed during period 1, whereas the second batch (4 mm pellet) was fed during period 2 to accommodate the larger size of the fish. The larger pellet size was introduced gradually over a six day period following the first sampling. Diets were kept in paper feed bags in a 4°C walk-in cooler for the duration of the period in which they were fed, then at -25°C pending analysis. Over the period of Oct. 19-28, 1996, Terramycin Aqua® was added to the feed (22.7 g premix/kg feed), and the ration was adjusted to - 1 % body weight as a prophylactic treatment against BKD. During the period of Nov. 18-27, 1996, vitamins C and E, at supplemental levels of 5 and 3 g/kg feed respectively, were mixed in with the feed in an attempt to enhance the immune system of the fish. Oil (2 g/kg feed), along with the palatability enhancer betaine (20 g /kg feed) was applied topically to the feed, the oil serving as a carrier for the supplemental antibiotic and the vitamins. Canola oil was used as the carrier for the Terramycin Aqua® so as not to introduce any additional astaxanthin to the diets, however, diet palatability was compromised and feed acceptance was poor. As a result, menhaden oil was used as the carrier for administering the vitamins.  23  3.1.4 Feeding During the experimental period, fish were hand fed to satiation twice daily following each weighing until normal feeding resumed. Each group was then fed to excess over the daylight hours (6-8 hours) using mechanical belt feeders (Ziegler Bros. Inc., USA). In order to more accurately determine feed intake per group, uneaten pellets were removed by siphon and counted every 1-2 days. Subsequently, their estimated as fed weight (number of pellets multiplied by the mean pellet weight before feeding) was subtracted from the total amount fed to obtain estimates of the actual rations consumed. On the day before each weighing and the actual weighing days themselves, feed was withheld to minimize fish stress and prevent overestimation of fish weight. Feed was also withheld on July 8, 1996 due to administration of Marinil® for the purpose of removing the swimming channels, as well as the day of and the day after the Chloramine-T treatment in November (Nov. 7 & 8, 1996), and the day of the Chloramine-T treatment in December (Dec. 13, 1996). Further information on the Chloramine-T treatments is covered in section 3.3.2.2 (Fish Mortality, Period 2).  3.1.5 Sampling The first sampling (October 1 and 2, 1996), was conducted after three months of feeding. Allfishwere again anaesthetised according to the protocol of Krieberg and Powell (1991). Individualfishweights and fork lengths were recorded. For each group, every fifth fish was sacrificed (for a total of twelvefish),and collected for flesh astaxanthin and colour analyses. Deskinnedfilletswere colour scored using the Roche Colour Card for Salmonids under natural lighting conditions. The left and rightfilletswere packaged separately under vacuum and stored between -35 and -40°C pending analysis. The left fillet was analyzed for  flesh astaxanthin, therightwas used for colourimetric measurements. All remaining fish were returned to their respective tanks for recovery. The tanks were cleaned by brushing followed by rinsing with clean sea water before refilling. Thefiberglassinserts were installed during the weighing procedure. At the final sampling (January 23 and 24, 1997), after seven months of feeding, the fish in each group were anaesthetised with Marinil® and then killed using a lethal dose of MS222®. Body weights, lengths, and fillet colour scores were recorded as described previously. Twelve fish from each tank were again randomly collected for flesh astaxanthin and colour determinations. The remaining fish from the tanks on the control and high Ax dietary treatments were packaged for Experiment 2.  3.1.6 Analysis of Fillets 3.1.6.1 Qualification of Colour Using the Roche Colour Card for Salmonids Immediately after filleting, allfilletswere assigned a colour score using the Roche Colour Card for Salmonids under natural lighting conditions. The weather was overcast for the first sampling but clear and sunny for the second. Therefore, the colour scores were judged in the shade to minimize the influence of the weather. At each sampling, a single judge was used to score all fillets, although the judge was not the same for both samplings. A scale of 11-18 was used, however, most fish on the control diet, as well as some of the smaller fish on the pigmented diets, were too pale to be considered an 11. For statistical purposes, those fish were assigned a score of 10. It is recognized that a variety of colours from palest green to pale peach were included in this group and, therefore, the assigned 10 should not be regarded as a colour score in itself. Assigning a colour score to fish with striping down the center of  25  the fillet (ie. colour along the lateral line but none towards the dorsal or ventral edges of the fillet) presented an obvious problem. Such fish were scored one score below the colour of the central stripe. It is noted that this problem arose primarily in smaller fish within the low pigment group at the first sampling, and was less frequently observed at the second sampling.  3.1.6.2 Colourimetric Measurements Hunter L, a, b scores were recorded on one fillet from each sacrificed fish. Fillets were thawed at room temperature for one hour prior to scanning with a HunterLab LabScan using a 2.54 cm aperture and standard illuminant D 5 (simulating daylight, correlated colour 6  temperature -6500 K). Eachfilletwas placed on an optically clear plastic plate and scanned four times at the midsection, being rotated 90 degrees between scans, and the average L, a, b values were recorded. Values for quantitative hue and quantitative chroma were calculated using the Hunter L, a, b scores, based on the equations given for use with CfELAB values (Hunt, 1987). Hue is defined by Hunt, as being an "attribute of a visual sensation according to which an area appears to be similar to one, or to proportions of two, of the perceived colours red, yellow, green, and blue", and is calculated as the angle having tangent to b/a (H° b = tan (b/a)). In _I  a  the case of salmon, hue generally describes the relationship between yellowness and redness (0a in Figure 4, reading counterclockwise from the a axis). However, in several of the +  +  control fish for which the Hunter a was negative, hue is describing the relationship of yellowness to greenness (0a' in Figure 4, reading clockwise from the a" axis). In order to ensure that comparable angles were being considered for fish in the two quadrants of the L, a, b colour solid, negative hue angle values were corrected to read in the positive sense  Figure 4 - Schematic Representation of Hue Angle  -b  27  (180° + H° ) such that 0a' = -85, would read 0a = 95. +  ab  The equation C b = (a + b ) 2  a  2  m  was used to calculate chroma, which is defined by  Hunt as "the colourfulness of an area judged in proportion to the brightness of a similarly illuminated area that appears to be white or highly transmitting". The chroma gives no information as to the actual colour as defined by hue.  3.1.6.3 Quantification of Astaxanthin  Extraction of astaxanthin from individualfilletswas performed as described by Kiessling et al. (1995). Wholefilletswere homogenized in a Waring blender until a paste of even consistency was produced. The blending time varied from fillet to fillet due to the way in which flesh adhered to the sides and under the blades of the blender, necessitating manual mixing on occasion. A one (± 0.005 ) gram sample of fillet homogenate was then extracted three times in a separatory funnel using a 2:2:1.8, chloroform: methanol: water mixture. The chloroform layers from each subsequent extraction were pooled in a 100 ml graduated cylinder, the volume recorded, and a 20 ml aliquot transferred via pipette to a round bottom flask for rotary evaporation under vacuum on a water bath (< 50°C). Following evaporation, the sample was dissolved in n-hexane and re-evaporated, then redissolved in 1ml n-hexane and transferred to a 1.0 ml vial (Wheaton, USA) for analysis. Vials were flushed with medical grade nitrogen gas (Praxair) prior to capping. Samples were stored overnight at -35°C then shipped to WVL, for analysis via high-performance liquid chromatography (HPLC) using the conditions outlined in Table 2.  Table 2 - HPLC Conditions Used for Quantification of Astaxanthin  HPLC Pump Auto injector U V Detector Wavelength Column Column manufacturer Column dimensions Pore size Mobile phase Flow rate Temperature Volume of sample injected  Waters LC Module 1 Waters 600E PowerLine™ Controller Waters 715 Ultra WISP™ Waters 486 Tunable Absorbance Detector 472 nm uPorasil™ Waters 3.9 mm by 150 mm 125A 82% Hexane, 18% Acetone (isocratic) 1 ml/min. 20°C 50 pi  29  Astaxanthin in the diets was quantified in a similar manner. One gram of feed, ground in a coffee grinder (Braun, type 4014), was extracted as per the Bligh & Dyer (1959) method of lipid extraction, and a 1 ml aliquot of the chloroform layer was prepared for HPLC analysis in the manner described above for analysis of the fillets. Only all-.E-astaxanthin was identified and subsequently quantified due to the unavailability of standards for Z-isomers, astacene etc. All chemicals used for the extractions were from Burdich & Jackson (double distilled, HPLC grade). Red glassware was used for bothfleshand diet extractions to minimize sample exposure to overhead light.  3.1.6.4 Proximate Composition of Fillets and Diets Four fillets from each group offishwere randomly selected and 5 g of homogenate from each was pooled for proximate analysis. The pooled samples were homogenized in a Waring blender as described previously. Analyses of diets and pooled fillet homogenates were performed in duplicate using the methods described by Higgs et al. (1979).  3.2 Calculation of Performance Parameters and Statistical Procedures Performance parameters were calculated for fish in all groups and averaged for each dietary treatment at the end of period 1, using the following formulae: 1) Weight gained (g) = average wet weight (g) of fish in the group at the end of period 1 - average wet weight (g) of fish in a group at the start of the experiment 2) Total feed intake (g/fish) = total weight of feed ingested by a group of fish (g, dry weight) / the number of fish in the group (adjusted accordingly to accommodate deaths)  30  3) Daily Intake (g/fish/day) = total feed intake (g/fish) / total number of days fed over period 1 4) Feed efficiency = weight gained (g) / total feed intake (g/fish) Minitab Statistical Software, version 12.0, was used to perform analysis of variance (ANOVA) using the General Linear Model (diet, time, diet*time) on all measures of fillet colour. Significance was determined by Tukey's test (p = 0.05). Correlations (Pearson's product moment and Spearman's rank order) were also performed using Minitab Statistical Software, version 12.0.  31  3.3 Results 3.3.1 Proximate Constituents and Astaxanthin Content of Diets The levels of proximate constituents and of Ax in the experimental diets are given in Table 3. The proximate compositions of the eight diets were similar, as anticipated for this experiment. The Ax contents of all diets with added Ax were generally lower than the added levels. Some loss of Ax due to thermal decomposition during processing was expected. While data for loss of Ax (as Carophyll® Pink) is not available, Gadient and Fenster (1994) have shown that for vitamin A in extruded trout feed, 11 - 15% loss is found dependent on the extrusion temperature. Additional losses could be attributed to the length of time between production of the diets and pigment analysis two years later. Breakdown of Ax is to be expected over this time despite the diets being kept at -25°C subsequent to the end of the respective feeding periods. It has been shown in previous research that while low temperatures serve to slow deterioration of astaxanthin, it is not fully arrested (Chen et al, 1984; No & Storebakken, 1991b; Ingemansson etal, 1993; Bjerkeng & Johnson, 1995). The 3 mm and 4 mm pellets of the control, low, and medium Ax diets had similar levels of Ax, in each case, whereas the amounts of Ax in the high Ax diet differed considerably between the two pellet sizes. Given that the 3 mm pellet of the high Ax diet had an Ax content (76.42 ppm) higher than the intended added value (75 ppm) it would seem that a larger amount of synthetic Ax, perhaps somewhere in the range of 90 - 100 ppm, was mistakenly added. This difference in Ax content between the two pellet sizes of the high Ax diet appears to have been reflected in flesh levels of Ax of the fish fed that diet (see section 3.3.5). Low levels of Ax were present in the control diet (~2ppm) and can be attributed to the use of fish oil in the diet formulation.  32  Table 3 - Levels of Moisture. Ash. Crude Lipid (CD. Crude Protein (CP). astaxanthin [Axl and Corrected Astaxanthin in the Test Diets (dry weight basis except moisture) A l l - E -  Diet  CL Pellet Moisture Ash Corrected CP size (mm) (%) [Ax] ppm [Ax] ppm (%) (%) (%) 1 3 control 9.10 10.46 18.49 50.06 1.65 n/a 2 low Ax 3 9.30 10.31 18.37 48.56 22.17 29.56 3 3 8.50 10.69 19.04 49.19 30.08 40.11 med. Ax 4 high Ax 3 9.50 10.73 19.06 47.96 57.31 76.42 5 4 9.00 control 10.40 19.17 48.80 2.56 n/a 6 low Ax 4 7.60 10.52 17.29 48.92 19.30 25.73 7 4 7.70 10.53 16.56 48.59 29.70 39.60 med. Ax 8 high Ax 4 7.50 10.27 18.59 49.31 44.86 59.82 Values are the average of duplicate samples per treatment Value corrected for content of Z-isomers of astaxanthin (based on Carophyll® Pink containing 75:25, all-is-Ax:Z-isomers) Treatment  2  A 3 1 - E -  1  2  33  3.3.2 Fish Mortality 3.3.2.1 Period 1 During period 1, only four deaths occurred in the twelve groups on trial. Of these, one death was attributed to anaesthetic (July 4, 1996), one fish was inadvertently struck by the siphon during tank cleaning (Aug. 28, 1996), one fish died from infection by skin bacteria (Sept. 19, 1996), and the last died of unknown causes (Sept. 22, 1996). Feed intake per fish per tank was adjusted accordingly.  3.3.2.2 Period 2 As a result of the first sampling, 8 fish died due to failure to recover from the anaesthetic (7 of these were from a tank which was later removed from the experiment). During the months following this sampling, all but one tank offish was plagued by disease, a Sporacytophaga infection causing tail rot (diagnosed by the Fish Health Program, Fisheries and Oceans Canada, Pacific Biological Station, Nanaimo, B.C.). As mortalities increased, an attempt was made to treat the disease via a topical therapeutant. A one hour static bath in Chloramine-T at a concentration of 8.5 ppm was recommended, to be followed the next day by a second such bath at a concentration of 17 ppm. Following the first bath, Nov.7, 1996, acute mortalities overnight (85 fish in total) in four of the tanks resulted in the removal of three of these tanks from the experiment, one each from the control, low, and medium Ax treatments. The fourth tank, which experienced relatively low mortalities (5fish),was also from the medium Ax treatment and was kept in the experiment because of the need for duplicate tanks of fish per treatment. One tank from the high Ax treatment was also removed from the experiment so that all treatments would have only duplicate groups. Under these  34  circumstances, the prescribed second bath was not conducted. Following this event, vitamins C and E were added to the diets, as described in section 3.1.3, in an effort to enhance the immunocompetence of the fish. A second static bath in Chloramine-T was given on Dec 13, 1996, for one hour at a concentration of 6.0 ppm, in a further attempt to stop mortalities. Despite attempts to control the disease, mortalities continued, and resulted in the premature termination of this experiment. Fish of all sizes were affected. Total mortalities for the remaining duplicate tanks of fish on each dietary treatment were as follows: control (15% & 24%), low Ax (30% & 22%), medium Ax (39% & 24%), high Ax, (24% & 7%). It is noteworthy that fish from the low and medium Ax groups were the most affected by the disease, whereas the control and high Ax groups experienced less mortality and morbidity. Recent research suggests that the addition of Ax to the diet can improve fish health (Nakano et al., 1995), however, the lack of a trend in the present study does not support this claim.  3.3.3 Feed Intake, Growth and Feed Efficiency 3.3.3.1 Period 1  The average feed intakes, weight gains and feed efficiencies of the duplicate groups on each of the dietary treatments during this period are reported in Table 4. The diets which varied only in astaxanthin content did not support different weight gains and feed efficiencies among the groups.  3.3.3.2 Period 2  Due to the high fish mortality rates and likelihood of moribund fish being anorexic, it was felt that data pertaining to feed consumption over this period would be erroneous.  35  Furthermore, analysis of the weights of fish fed the control, medium Ax, and high Ax dietary treatments, revealed that they did not follow a normal distribution pattern, almost certainly a result of morbidity. Therefore, the performance data have not been reported for period 2.  3.3.4  Fillet Composition  Partial proximate analyses of the pooled fillets for duplicate groups on each dietary treatment are reported in Table 5. These results affirm that feeding of different levels of astaxanthin did not affect fillet composition. Statistical analysis via ANOVA was not done on proximate composition results because only duplicate values for all measures (ie: crude lipid, crude protein, ash, moisture) were obtained. While an ANOVA is possible, it would be inappropriate due to the small number of observations for each treatment, in addition to these values being obtained from pooled fillets. Variance due to fish - fish difference would not be reflected in these data.  3.3.5 Colour Measurements  Due to their small size (-25 - 55 g), nine of the fish sampled at time 0 (July 2, 1996), were analyzed for flesh astaxanthin only (X - 0.05, o = 0.04, n = 9) and are therefore not included in the table outlining colour data obtained for fish sampled at three and seven months. One of the day 0 fish was too small for analysis. With respect to flesh astaxanthin levels, the day 0 fish were significantly different (p < 0.05) from all fish being fed diets with supplemental astaxanthin for 3 & 7 months, but not significantly different from those fed the control diet for 3 & 7 months. Results of colour scores for all fillets at three and seven months using the Roche Colour Card for Salmonids (RCC), Hunter Lab L, a, b scores,  36  Table 4 - Feed Intake, Growth and Feed Efficiency of Chinook Salmon Fed Diets Containing Low, Medium (Med.), High, or No (Ctrl) Supplemental Astaxanthin for Three Months '(Period One) Diet  Initial Final Weight Total Feed Weieht(e) Weight (g) Gained (g) Intake (g/fish) Ctrl 39.14 115.50 76.32 75.09 (7.99) (24.55) low 38.17 116.58 78.40 74.49 (7.73) (24.98) med. 38.19 116.12 77.94 71.83 (7.55) (30.73) high 39.47 111.82 71.60 73.65 (7.23) (24.64) Values are the average of duplicate groups per treatment Standard deviations are shown in parentheses  Daily Intake (g/fish/dav) 0.89  Feed Efficiency 1.02  0.89  1.05  0.87  1.09  0.88  0.97  2  1 2  Table 5 - Partial Proximate Analyses of Fillets from Chinook Salmon Fed Diets Containing Low, Medium (Med.), High, or No (Ctrl) Supplemental Astaxanthin for Three or Seven Months 1  Diet  Months Rep %Moisture 2  %Ash  %CL  3  %CP  4  1 72.96 2.01 4.55 20.58 2 73.86 1.93 3.85 21.00 low Ax 3 1 74.92 1.77 3.62 21.04 2 73.25 1.87 4.24 21.28 med. Ax 3 1 74.21 1.89 4.42 20.22 2 73.44 1.86 4.29 20.76 high Ax 3 1 73.95 4.23 20.97 1.91 2 73.81 1.96 4.65 20.74 Ctrl 7 1 74.44 1.90 4.11 21.45 2 74.76 1.76 3.66 20.47 low Ax 7 1 73.16 1.87 5.12 21.39 2 73.49 1.77 4.15 21.20 med. Ax 7 1 74.00 1.79 5.09 20.26 2 72.11 2.45 4.82 20.95 high Ax 7 1 73.50 1.80 4.29 20.74 2 73.44 1.86 4.36 21.00 Values are the average of duplicate analyses on flesh samples comprised of 4 pooled fillets Rep = replicate group CL = crude lipid CP = crude protein Ctrl  3  2  3 4  37  hue angle, chroma andfleshconcentrations of astaxanthin via HPLC ([Ax]), are presented in Table 6. Dietary [Ax], time, and the interaction between them (time on the diet), all significantly affected fillet lightness (Hunter L) (p < 0.05), and the score given using the RCC. Lightness decreased with an increase in either dietary [Ax] and/or time. The exception to this trend were thefilletsfromfishfed the low Ax diet, where lightness increased with increasing time on the diet, though not significantly (p < 0.05). Considering Hunter L, the effect of the control diet was significantly different from the effects of all diets with supplemental astaxanthin, and the effect of the low Ax diet was different from that of the high Ax diet (p < 0.05). Roche Colour Card scores increased with an increase in either dietary [Ax] and/or time. An exception here were the fillets from fish fed the control diet where RCC decreased with time, however, this decrease was not significant. Based on RCC results, the effect of the control diet was significantly different (p < 0.05) from the effects of all diets with supplemental astaxanthin, and the effects of both the low and medium Ax diets were different from that of the high Ax diet (p < 0.05). Dietary [Ax] and the diet*time interaction both had a significant effect on fillet redness (Hunter a), yellowness (Hunter b), hue angle, chroma, and [Ax], whereas time alone did not. Fillet redness, yellowness, and chroma tended to increase with increased dietary [Ax] as well as with increasing time on the diet. Exceptions to these trends are the fillets from fish fed the high Ax diet and those fed the control diet, where a decrease in the above measures, though not significant, was observed with increasing time on the diet. Based on either Hunter a or chroma, the effect of the control diet was significantly different from the effects of all diets with supplemental astaxanthin, and the effect of the low Ax diet was different from that of the  38  high Ax diet (p < 0.05). Based on Hunter b, only the effect of the control diet was significantly different (p < 0.05). The [Ax] in the fillets also tended to increase with increasing dietary [Ax] and with time on the diet. The only exception were thefilletsfromfishfed the high Ax diet where a decrease in fillet [Ax], again not significant, was observed. Considering the [Ax] in the fillets, the effect of the control diet was significantly different (p < 0.05) from the effects of all diets with supplemental astaxanthin, and the effect of the low Ax diet was different from that of the high Ax diet (p < 0.05). Fillet hue angle decreased with increasing dietary [Ax] and with increasing time on the diet, exceptions once again being the fillets from fish fed the control and high Ax diets where an increase was observed with increasing time on the diet. With respect to hue angle, the effect of the control diet was significantly different from the effects of all diets with supplemental astaxanthin, and the effect of the low Ax diet was different from that of the high Ax diet (p < 0.05). Correlation coefficients indicating the strength of the linear relationship between each of the measures of colour are provided in Table 7. All treatment groups were combined since relationships between the different colour measurements are considered independent of the dietary treatments. All correlations were significant (p = 0.000) with the exception of that between Hunter L and Hunter b (r = -0.025, p = 0.727). The strength of the relationship between the actual flesh [Ax] (response) and the other measures of colour (predictor) is of interest for the development of a rapid, non-destructive, and less expensive method of closely estimating flesh [Ax]. The Roche Colour Card score, a popular industry method of assessing flesh pigmentation, had a high correlation (r = 0.847) with the actual concentration of  39  astaxanthin in the flesh. However, a higher correlation (r = 0.901) was obtained between the flesh concentration of astaxanthin and the corresponding Hunter a value. This indicates that Hunter a values would be a more accurate predictor offlesh[Ax]. The correlation coefficients for fish weight and flesh [Ax] for individual dietary treatment groups at both the three and seven month time intervals are given in Table 8. All groups show a significant relationship between fish weight and flesh astaxanthin concentrations, the correlations generally becoming stronger with increasing time on the diet. However, the relationships are not good enough for fish weight to be considered as a predictor offlesh[Ax]. Figure 5 represents the Hunter L, a, b colour solid for fillets from all dietary treatments. This diagram is a spatial arrangement of the fillets with respect to their Hunter L, a, and b values, and is a good illustration of how the fillets from fish fed the control diet differ fromfilletsoffish fed diets with different supplemental levels of astaxanthin, with respect to their Hunter parameters. Allfilletsfromthe three month (October, 1996) sampling, are represented with open symbols, while those from the seven month (January, 1997) sampling are represented by closed symbols. Fillets offishfromthe control group are denoted with circles, the low Ax group with squares, the medium Ax group with diamonds, and the high Ax group with triangles. Allfilletsfromfishfed the control diet are represented in the upper half (high L) of the front (low b), left side (low a) of the colour solid, whereas fillets from fish fed the diets supplemented with Ax are better described as being observed to the rear (medium b), right side (medium to high a) of the colour solid, any differences between dietary treatment groups with supplemental astaxanthin being less obvious.  The relationships between flesh [Ax] and the other measures of colour; Hunter L, a, b, hue angle, chroma and RCC, are illustrated in Figures 6 through 11. Each relationship has been represented three times (a, b, and c) for the purpose of drawing attention to; a) the relationship of the eight groups offish(4 dietary treatments, 2 sampling times), b) the linear regression, and c) the linear regression using Spearman's rank order correlation. In a), all fillets from the three month (October, 1996) sampling, are represented with open symbols, while thosefromthe seven month (January, 1997) sampling are represented by closed symbols. The control group is again denoted with circles, the low Ax group with squares, the medium Ax group with diamonds, and the high Ax group with triangles. The groups of fillets fromfishfed the different dietary treatments were shown in this way so as to demonstrate the amount of variance in each group. These illustrations also serve to highlight the degree of overlap between groups fed varying amounts of supplemental astaxanthin, in contrast to the control group which is quite obviously different in all of the figures. The latter representation of eachfigure,c), is included because several of the relationships between various colour evaluation methods appear, graphically, to be curvilinear, despite having a good linear relation (r > 0.80). In order to assess the curvilinearity of all of the relationships, the data were reanalyzed using Spearman's rank order correlation to observe if linearization of the data would improve the correlation coefficient. It is recognized that the ranked data is of no use as a predictor equation since the magnitude of difference between values is eliminated by ranking, it serves only to determine curvilinearity. In all cases there was no marked improvement in the correlation coefficient, indicating that none of the relationships are more curvilinear than linear. It is suggested that the visual appearance of curvilinearity on such graphs, is a floor effect due to at least one quarter of the observations (fillets from fish fed the  41  control diet, receiving no supplemental pigmentation) all falling on, or near, the lower limit of the Y-axis (concentration offleshastaxanthin = 0). Use of a quadratic equation was also investigated, however, no significant increase in r was observed for any of the relationships, 2  indicating that the linear model was the best fit for the data collected in this study. The relationship between Hunter L and fillet [Ax] is shown in Figure 6a and b. Hunter L could be generally described as increasing with respect to decreasing flesh [Ax]. However, the relationship, while significant (r = -0.597, p = 0.000), is poor. Rank order linearization of the data (Figure 6c) did not markedly improve the correlation coefficient (r = -0.638). Figure 7a and b illustrate the monotonically increasing relationship between Hunter a and actual [Ax] of each fillet. In Figure 7a, the difference between fillets from fish fed the control diet and those from fish fed diets with supplemental astaxanthin is very obvious. Figure 7 (a and b) is an excellent example of the visual appearance of curvilinearity, a result of the floor effect of the lower limit of the y-axis, despite there being a strong positive linear relationship (r = 0.901). The distribution of points along the regression line (Figure 7b) adds to this effect as there are fewer observations in the region which appears to curve. Figure 7c shows the data points plotted according to their ranked order on both axes. The correlation coefficient (r = 0.911) that was obtained for the linearized data was only slightly better than that reported above. The relationship between Hunter b and actual flesh [Ax] is represented in Figure 8a and b. As a general trend, Hunter b was positively correlated with flesh [Ax]. While significant, the linear relationship was poor and of no use as a predictor equation (r = 0.598). Rank order linearization of the data did not improve the correlation coefficient (r = 0.587).  Figure 9a and b, describing the relationship between hue angle and [Ax] in the flesh, provide another excellent example of apparent curvilinearity. Once again, rank order linearization of the data in Figure 9c, while visually more linear, did not markedly improve the correlation coefficient (r = -0.892). This indicates that the good linear correlation (r = -0.832) should be accepted. Figure 10a and b illustrate the relationship between chroma and flesh [Ax]. A strong positive relationship (r = 0.880) was found. The correlation coefficient was not improved by rank order linearization of the data(Figure 10c, r = 0.868). The relationship between fillet scores using the RCC and flesh [Ax] is shown in Figure 11a and b. A good positive linear relationship was found (r = 0.847), however, it is noteworthy that a wide range of RCC scores, that represented half of the scale, were ascribed to fillets with approximately the same [Ax] and vice versa. Rank order linearization of the data (Figure 11c) did not improve the correlation coefficient (r = 0.836).  43  c  eu es o o # fc tf CU e o  od ro od r i  n  O  r-'  oo (fj o  cn oi cn  #  4=  CU  s w W CU  C  <u  I  D  es •<-< O  tf  cn cs ^  ON'  C£?  • cn  o  m CN VO  N vO °° co| 00  in  o  cn  a  IU  cu  13  es  "3  c o  m oo vo ^ co <~~ o o, CS 0-0  o ^  V  o rS.oo0 0£;"vo £ (N  i—i  ^ ,X CS w  p°* o  vo - H '  o  -J  /-sU  «  61 B B  o  es  o  co  r-I  ^f'  «n o  o  o  5 TJ V  H  B O  a  «"> ^ ' m'  o  CS  vo  ,_•  B O  CD  '•3  J3 bO  o  IS T3 C cd  'S  CN  m  CO  ,_'  r.;  °°. 001 CO  CO ~  O  TJ-CS vo; fO  P H  n  r H  (SJ  -d CU  03 tn +-»  CS p VO VO rH!  T3  co oo VO "^g co""^ O^vo ^ " O CO^OO o cs ^q oo o\ «n ^q N ^ ts ts oo ^ ^ cs co cs ^ co' cs C N co ^ m"  ON  O.  QQ  co  e>  •c o cn  o  <N  O  CS m O N o VO m v d cs^ vd  ^—- C O  CO  2  in in  CO  o  CN  vo r~ cs  ^ -  ON  i^i  in CO  CS  TJ-  co  CS  o ON CO CO  CO  CO  CO  IH  op  'cn CD  ON  O  H  in  w  rn  cs cs'  H  t~~ in)  u. cd  rC  2 S  S ,  *o o  °  E  o -S c II -o  CO  'J3 a  o  o  e  00  c o o  T3  o  s  00  rS  s ?s to a s  'S >. .  tu .SP ^  c  §  Si c o  cn CD cn  « H  (U c  fc H  '-3 - °  .8*  u  'I  CJ  Xi  H  E E § Js o 3 <h e ^ s O s .S 'S ^ s  C  ON  •  oa) "oo .SP  GO  eg  CU  E  Q. 3  O  u  &  CU  B  3  cn  IH  VH  B  D  <D  u-  <  eu  C  op  "3 es wo es  cn  + J  'In <U  fe u  *J  X>  C  ^~vo  v"  VO r - t N O N O O o o ^ r o o o o o T f o o r j - r ^ CN CO T f CN ^q vo r» co co N O ON oo p; cs C N  ^ cn o o s G W JS B CM O es u cu S cn V u E U o 3  vo  ^  B3  o  I uu s  es eu  00  v d ON v d  f~ c n ^ O O £  2 ucn  cn J : cn cn  cn  r H  B B O eu  U  B  r N  t«  > cu Vi e -MB * J uo s eu cu eu W oe© cu u JS 6U  m'  c-~  o  vq r -  <U  s  c s W «H  u  in  m  T3  CU  c  in  >—;  &  "33  cn V  CO  •i|  «  HS  cd  •a C cn M . SM P g •55 .2  OD S "S3 * c^.a-a.^  . -  CU  ~ ^3  S cd  rS ^ - H  O  -~  S  Q  U  .„  -o  (U  -o  JcJel rjS  e cd  4H  00  44  Table 7 - Correlations Between Hunter Lightness (L), Hunter Redness (a), Hunter Yellowness (b), the Roche Colour Card for Salmonids (RCC), Flesh Concentration of Astaxanthin ([Axl ppm). Hue Angle, and Chroma as Colour Scoring Methods Used to Assess Chinook Salmon Fillets L  a  b  RCC  [Ax]ppm  Hue  L  1  a  -0.637*  1  b  -0.025**  0.724*  1  RCC  -0.711*  0.930*  0.559*  1  [Ax] ppm  -0.597*  0.901*  0.598*  0.847*  1  Hue  0.725*  -0.955*  -0.579*  -0.939*  -0.832*  1  Chroma -0.489* * p = 0.000 ** p = 0.727  0.974*  0.844*  0.866*  0.880*  -0.877*  Chroma  1  Chinook Salmon Fed Diets Containing Low, Medium (Med.), High, or No (Ctrl) Supplemental Astaxanthin for Three or Seven Months Ctrl. 3  Ctrl. 7  Low. 7  Med. 3  Med. 7  High. 3  High. 7  0.412 0.826 0.543 0.430 (0.045) (0.000) (0.006) (0.036) Probabilities are shown in parentheses  0.459 (0.024)  0.647 (0.001)  0.406 (0.049)  0.596 (0.002)  1  1  Low. 3  45  Figure 5 - Hunter L, a, b. Colour Solid Showing Fillets from Chinook Salmon Fed Diets Containing Low (square). Medium (diamond). High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  Hunter L  46  Figure 6a - Relationship Between Hunter L and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low (square). Medium (diamond). High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  E  Q. Q.  0  47  Figure 6b - Regression Plot for Hunter L and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  Y = 6.39288-0.134420X r = -0.597  E  £  2  o —\  I  30  40  50  Hunter L  48  Figure 6c - Regression Plot Using Spearman's Rank Order Correlation for Hunter L and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  Y = 158.099-0.638332X r = -0.638  49  Figure 7a - Relationship Between Hunter a and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low (square). Medium (diamond). High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  E  CL Q.  0  50  Figure 7 b - Regression Plot for Hunter a and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low. Medium. High, or No Supplemental Astaxanthin for Three or Seven Months  Figure 7c - Regression Plot Using Spearman's Rank Order Correlation for Hunter a and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low. Medium. High, or No Supplemental Astaxanthin for Three or Seven Months  Y = 8.56756 + 0.911217X r = 0.911 200  —{  —  r  0  j  100  .  .  ^  200  Hunter a (rank)  52  Figure 8a - Relationship Between Hunter b and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low (square). Medium (diamond). High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  3 -4  *  .  '  D O  A  E CL  Q.  <  A  _  ^>  A  n  ^  O  •  O  •  1 -A  •  o  0  A  •  o  I  8  13  18  Hunter b  53  Figure 8b - Regression Plot for Hunter b and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  Y = -3.03324 + 0.357333X r = 0.598  4  H  8  13  18  Hunter b  54  Figure 8c - Regression Plot Using Spearman's Rank Order Correlation for Hunter b and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low. Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  55  Figure 9a - Relationship Between Hue Angle and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low (square). Medium (diamond). High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  3  -A  2  -A  E CL CL  A  • V* o,  ? l\<> * A«* a  •  0 T  30  40  50  60  70  80  i  i  90  100  Hue Angle  56  Figure 9 b - Regression Plot for Hue Angle and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  57  Figure 9c - Regression Plot Using Spearman's Rank Order Correlation for Hue Angle and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  Y = 182.506-0.891497X r = -0.832  Figure 10a - Relationship Between Chroma and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low (square). Medium (diamond). High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  •  •  E  Q. CL  * -4;: - • • 1 -A  0 10  20  30  Chroma  59  Figure 1 0 b - Regression Plot for Chroma and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  Y = -2.07498 + 0.199816X r = 0.880  4 —\  10  20  Chroma  60  Figure 10c - Regression Plot Using Spearman's Rank Order Correlation for Chroma and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low. Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  Y= 12.7437 + 0.867941 X r = 0.868  61  Figure 11a - Relationship Between Roche Colour Card Scores (RCC) and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low (square). Medium (diamond), High (triangle), or No (circle) Supplemental Astaxanthin for Three (open symbol) or Seven Months (closed symbol)  E  Q.  2  -  0  RCC  62  Figure lib - Regression Plot for Roche Colour Card Scores (RCQ and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium, High, or No Supplemental Astaxanthin for Three or Seven Months  Y =-4.19725+ 0.405663X r = 0.847  4  H  3  -A  i 10  i  i  i  i  i  r  r  11  12  13  14  15  16  17  RCC  63  Figure 11c - Regression Plot Using Spearman's Rank Order Correlation for Roche Colour Card Scores (RCC) and Concentration of Flesh Astaxanthin (Ax (ppm)) for Fillets from Chinook Salmon Fed Diets Containing Low, Medium., High, or No Supplemental Astaxanthin for Three or Seven Months  Y= 15.4382 + 0.840631 X r = 0.836 200  —I  0  100  RCC (rank)  64  3.4 Discussion  In this experiment, all but one diet with added synthetic astaxanthin contained substantially lower levels of astaxanthin than added. Allowing for -10% loss of Ax due to thermal decomposition, the Ax concentrations of the diets (corrected for content of Zisomers) were still lower than anticipated. There are several possible explanations for this finding. Two years had passed between the time at which the diets were produced and the time at which they were analyzed for astaxanthin content. While the diets had been stored at 25°C at the end of the experimental period, it is reasonable to expect that some pigment loss occurred. It has been shown in previous research that while low temperatures serve to slow deterioration of astaxanthin, it is not fully arrested (Chen et al, 1984; No & Storebakken, 1991b; Ingemansson et al, 1993; Bjerkeng & Johnson, 1995). It is also possible that digestion of the feed with Maxatase®, as per the method of Schierle & Hardi (1994), would have aided in breaking down the protective matrix which surrounds the astaxanthin in the Carophyll® Pink beadlet, giving a higher, more accurate, estimate of the astaxanthin content of the diets. Maxatase® was not used in this study due to its unavailability at the time. A large difference in Ax content was found between the 3 and 4 mm pellet sizes of the high Ax diet, the 3mm pellets containing 28% more Ax than the 4mm pellets. This difference most likely stemed from an error during production of the diets. Interestingly, the drop in the concentration of dietary astaxanthin was reflected by a decline (though not significant) in the average flesh Ax levels of the fish fed the high Ax diet during the time between thefirstand final samplings.  65  - i ;  EES'8 ce« B -  -  O  -o  GO  es  O  -o in +>  <  oio.:.*..-^^,..  C  s  o -o  ns eu  es «  -M  W •»-»  cu  o o  5  CN  CM  O  'S  > (3  C o PH  u E rs CU  "1  1  I  I  i CO  o  i i  ~i r  r CH o  o  o o  o o o  s  66  In Figure 12, (FfPLC Profile Typical of Diet Extract (Medium Ax)), the two peaks preceding, as well as the two peaks following the all-is-astaxanthin peak, have, in previous papers, been identified as Z-forms of astaxanthin (Schierle & Hardi, 1994; Bjerkeng et al. 1997b). Due to the lower specific absorption of these isomers as compared to that of the allis-form (El%, 1cm = 1750, 1310 and 2100 for 9-cis, 13-cis, and all-is respectively in nhexane) (Hoffmann-La Roche, 1992) it would be desirable to calculate a correction factor, specific to the HPLC conditions used, to account for the differences (Schierle & Hardi, 1994). Once again, the Unavailability of these standards has prevented the determination of such correction factors and positive identification of the peaks in question. Without proper identification and quantification of the Z-isomers it would be erroneous to include them in the overall astaxanthin content. However, in recent work by Bjerkeng et al. (1997b) it was reported that Carophyll® Pink consisted of-75% all-is-astaxanthin and 25% Z-isomers of astaxanthin. With this in mind, it would perhaps be more accurate to regard the area under the all-is-astaxanthin curve as representing 75% of the total astaxanthin content of the diet as provided by Carophyll® Pink. This could not be used to correct for the control diets since any astaxanthin present would not originate from Carophyll® Pink and could not be expected to be present in the aforementioned ratio. This problem of accounting for the Z-isomers of astaxanthin was not of great concern when quantifying the flesh astaxanthin concentration of the fillets. In general, Z-isomeric peaks were not present, and if so, were of negligible size (Figure 13). Recent work by Bjerkeng etal. (1997b), indicated selectivity in the metabolism of the various stereoisomers and suggested that there is substantial isomerization of the Z-isomers in the liver of rainbow  o -o 00  xssnr.  o  -o  4>  <  m  s  •P  w  3 o -o  e o C5  © o e  o -o CM  et  D  >• H  <u  C o u  e IS  c  •  PM  U  -J PH •  M M  T T  I M  M M  I  3  M M  in CN O  O CM O  If) H O  O H O  If) O  O  O  O  O  o  o  o o o  o c  3 CO  a  i-H  <U u  S  Oi  II  -  < ex  Figure 14 - Astaxanthin-actomyosin Complex Model  1  indicates hydrophobic bonds  Hydrophobic pocket in actomyosin complex  Adapted from Henmi et al. (1991)  69  trout. Based on the binding of astaxanthin to the acto-myosin complex as proposed by Henmi et al. (1991) (Figure 14), it is possible that stearic hindrance prevents binding of the Z-forms of astaxanthin to actomyosin. As seen in other studies, scores using the Roche Colour Card for Salmonids correlated well withfleshastaxanthin concentrations (Figure 11, r = 0.847). Contrary to thefindingsof Christiansen et al. (1995), in the present study the range offleshastaxanthin levels corresponding to a particular colour score increased for the higher scores. It is recognized that dissimilar weather conditions on the two sampling dates (clear vs. cloudy) could have affected the relative scores awarded to thefillets,due to the effect on illumination and subsequent colour perception of the Roche Colour Card tiles (illuminant metamerism). This effect may have been moderated by the phenomenon of colour constancy, the ability of the human visual system to compensate for changes in lighting (level and colour) and, therefore, recognize objects as having almost the same colour under a variety of conditions (Hunt, 1987). The relationship between flesh astaxanthin levels and Roche Colour Card scores in the present study, agreed with the results of the study by Sigurgisladottir et al. (1994). However, when thefindingsof this study were compared to those of Christiansen et al. (1995), the flesh astaxanthin concentrations in the former seem relatively low. In this regard, the present study on chinook salmon (weight 50 - 360 g) and that of Sigurgisladottir et al. (1994) on Atlantic salmon (weight 560 - 630 g), showed thatfleshastaxanthin concentrations of 0.3 - 0.4 ppm corresponded to an 11 on the Roche scale. By contrast, the study by Christiansen et al. (1995) on Atlantic salmon (weight, 1.1 - 6.6 kg) showed that a flesh astaxanthin level of 2.0 ppm corresponded to a value of 11 on the Roche scale. Further,fishscoring a 16 on the  70  Roche scale in this study had flesh astaxanthin levels that ranged from 1-4.3 ppm, which agreed with the values obtained by Sigurgisladottir et al. (1994; mean -2.5 ppm), but not those determined by Christiansen et al. (1995; mean ~8.6ppm). While recognizing the limitations of the Roche Colour Card, there seem to be very large discrepancies in the range of flesh astaxanthin concentrations which correspond to a particular colour tile. Several explanations for this are as follows: 1) the larger fish used in the study by Christiansen et al. (1995) may have had some degree of fat separating the myotomes of the fillets which could have led the judge to give them a lower score as compared to the fish in the present study which had no visible fat in the fillets, 2) there may have been differences between species, although this point does not apply to the two studies on Atlantic salmon that were cited, (Christiansen et al. 1995; Sigurgisladottir etal. 1994), 3) the experience of the judges to judge colour was likely not the same between studies which could have caused some discrepancy in findings, 4) the differences in illumination source, natural light in the case of the present study, artificial light (RA > 90, colour temp > 5000 K) used by Christiansen et al. (1995), and unspecified in the study by Sigurgisladottir et al. (1994) could have led to some discrepancy in colour perception between studies, 5) there may have been losses of astaxanthin in the fillets during storage prior to analysis; the fillets were kept frozen (-35°C) in foil laminate barrier films with low O2 transmission rates (0.7 cc/m /24 hr) for up to two years in the present study, in unspecified packaging 2  71  material at -20°C for no more than two months in the study by Christiansen et al. (1995), and under unspecified conditions in the study by Sigurgisladottir et al. (1994). 6) there may have been differences amongst the the present study and those of Christiansen et al. (1995) and Sigurgisladottir etal. (1994), as to the extraction efficiencies and quantification methods used for measuring fillet astaxanthin concentrations. Direct comparisons of the a (redness) values of the present study and those of Christiansen et al. (1995) were not possible due to differences in the colour systems and illumination sources that were used. The difference in the absolute values represented by L, a, b (Hunter Lab) and those of L*, a*, b* (International Commission on Illumination (CIE) 1976) are recognized, since these sets of values are obtained via different colour measurement systems. However, the values L and L*, a and a*, and b and b* are significantly correlated (Skrede & Storebakken, 1986a) and, therefore, will be compared throughout the discussion in terms of general trends. The HunterLab a values were shown to be the best predictor of flesh astaxanthin concentration in this study (Figure 7b, r = 0.901). Analyses of variance on Hunter a values and fillet astaxanthin concentrations (Table 6) revealed identical patterns of significance, which reaffirms the high correlation coefficient. The results here agree with previous studies by Saito (1969), Skrede & Storebakken (1986a,b), and Christiansen etal. (1995) on Atlantic salmon, together with those of No & Storebakken (1991b) on rainbow trout, and Smith et al. (1992) on coho salmon. Collectively, all of these studies have shown a* (Hunter a) to be a good predictor of actual flesh pigment concentrations. In the present study, the Hunter a value indicated a degree of redness in the control fish despite the lack of deposited astaxanthin. Thisfindingmay have been the result of myoglobin in the flesh, or there may  72  have been some residual blood in the area being scanned, although the fillets were rinsed following filleting. In either case, the Hunter a values represented something other than the pigment of interest in the control fillets. It is noteworthy that some of the unpigmented fish on the control diet had negative a values and in this situation the fillets were more green than red. The HunterLab L values were negatively correlated with flesh astaxanthin concentrations (Figure 6b, r = -0.597), and were not good predictors of flesh astaxanthin level. The negative correlation coefficient is in accord with thefindingsof Skrede & Storebakken (1986a,b) (-0.58, -0.52) for Atlantic salmon, but lower than that found by Choubert et al. (1992) (-0.84) for rainbow trout. Hunter b values were, in general, positively related to flesh astaxanthin levels (Figure 8b) and similarfindingshave been reported by No & Storebakken (1991b) for rainbow trout, Smith et al. (1992) for coho salmon, and Christiansen et al. (1995) for Atlantic salmon. Hunter b values were not found to be a good predictor of fillet astaxanthin concentration (r = 0.598). In contrast to thefindingsof Christiansen et al. (1995), hue showed a good inverse linear relationship with flesh astaxanthin (r = -0.832), suggesting that it should be a good predictor of flesh astaxanthin. Visual inspection of Figure 9 (a or b) indicates otherwise, as the data points appear to diverge into two groups showing separate trends; one (hue angle >60), a horizontal line indicative of no relationship, and the other (hue angle <60) a negative slope. This appearance of a different relation between flesh astaxanthin and hue at the higher levels of flesh astaxanthin sparks interest in the strength of the relationship at levels beyond the bounds of the present study. Attention should be drawn once again to the presence offish with negative Hunter a values. This created some difficulty in using the formula for hue angle as given by Hunt (1987) which required taking the average of the group containingfishbeing  73  described by the Hunter a value as green, in addition to those described as red. As described in section 3.1.6.2, manipulation of which angle was being measured as hue allowed for an average value for the group to be obtained. Chroma, was shown to be a relatively good predictor of flesh astaxanthin concentrations (Figure 10b, r = 0.880), as was found by Christiansen et al. (1995). However, chroma was not as good a predictor as Hunter a (Figure 7b, r = 0.901). The point must be made that the value obtained for chroma gives no indication of the actual flesh colour, moreover, the original Hunter a and b values cannot be derived from chroma due to squaring and rendering all values positive. In the present study, the effect of squaring the Hunter a values was potentially misleading in the case of control fish showing a certain degree of "greenness". Fish in this case would have the same chroma value as those with an equal degree of "redness". While this would not usually present a problem in the salmon industry, since farmed market size salmon usually contain some astaxanthin in their flesh, it is apparent that there can be pitfalls when values are obtained from formulae and the original information is not included. The results of this study indicated that, while significantly related, fish weight did not have a good linear relationship to flesh astaxanthin levels in chinook salmon fed diets supplemented with astaxanthin, regardless of dietary astaxanthin concentrations. The study by March and MacMillan (1996) also showed low correlations (no correlation coefficient reported) between fish weight and flesh carotenoid content for chinook, Atlantic salmon and rainbow trout. A study by Torrissen (1985), reported that "a linear relation was found between the weight groups and the astaxanthin level in the flesh" of rainbow trout, however, no correlation coefficient was given. In addition, the weight groups in the study by Torrissen,  74  the ranges of which were unspecified, were independent of the dietary concentrations of astaxanthin used in the study while the dietary concentration was reported to affect flesh concentration of astaxanthin. The presentfindingsindicated that chinook salmon below a weight of 120 g can indeed deposit astaxanthin in the flesh, to the extent that they score relatively high (15 - 16) on the Roche Colour Card scale provided that the level of dietary astaxanthin is sufficient. The smallest brightly pigmented fish observed in this study were the following; 1. high Ax diet  86.3 g, [Ax] = 1.32 ppm, RCC = 15.5 80.6 g, [Ax] = 2.17 ppm, RCC = 15.5 98.4 g, [Ax] = 2.25 ppm, RCC = 15  2. low Ax diet  98.7 g, [Ax] = 2.39 ppm, RCC = 14  Thesefindingswould indicate that deposition of astaxanthin in the flesh of chinook salmon begins well before 80 g body weight is attained. Considering the results from the most effective methods of assessing salmonid flesh colour, viz. actual flesh Ax levels, Hunter a, and the Roche Colour Card (as the industry standard) no difference was apparent between the effects of the low and medium Ax diet or the medium and high Ax diets. However significant differences were found for the colour of filletsfromfishfed the low and high Ax diets after seven months. Thesefindingssuggest that the addition of 55 ppm of Ax (as represented by the medium Ax diet) in the diet of chinook salmon (body weight, 50 - 360g) is adequate to ensure acceptable pigment deposition in the flesh when fed over a period in excess of seven months.  3.5 Conclusion Datafromthis experiment suggest that pigment deposition in the flesh of chinook salmon in sea water starts well before the fish reaches 80 g. It appears that no more than 55 ppm of supplemental Ax in the diet is necessary to ensure adequate pigmentation in the flesh of chinook salmon in the size range of 50 - 360 g when fed over a period of at least seven months. The results of this study further suggest that it would be worthwhile to investigate the potential for supplying lower levels of astaxanthin in the diet (-35 ppm added) starting early in the life history. Since remobilization of carotenoids from the muscle in sexually immature fish is very slow (Choubert, 1985; Foss et al, 1984), this strategy would take advantage of the higher percent retention of the lower concentration of astaxanthin in the diet (Torrissen, 1985; Choubert & Storebakken, 1989; Storebakken & Choubert, 1991; Smith etal, 1992; Whyte et al, 1998). Further, this strategy could result in less wastage of an expensive dietary supplement while simultaneously ensuring that the fish would be adequately pigmented by the time they reached market weight. In addition, the drop in flesh astaxanthin experienced by the fish on the high Ax diet which mirrored the drop in dietary Ax (22% less Ax in the second feeding period) indicates that research into the efficacy of offering high initial amounts of dietary Ax followed by lowering the amount of dietary Ax later in the production cycle or feeding unpigmented feed for the last few months before marketing, is merited. This type of production program would result in a rapid increase of flesh Ax when the fish are smaller and do not consume as much feed, then maintenance of flesh Ax levels and accommodation of new muscle as the fish grows. Care would have to be taken not to lower the Ax content of the diet  76  too early before marketing or, as indicated in this experiment, flesh colour could decrease visibly. It is recognized thatfishmortality was exceedingly high during period 2 of this study and may have influenced the degree offleshpigmentation of the salmon due to poor overall feed intake However, it is the opinion of the author that the trends found in this study are still valid.  77  4 EXPERIMENT 2  4.1 Materials and Methods 4.1.1 Experimental Design The remainingfishfromthe control and high Ax treatmentsfromExperiment 1 (unpigmented and pigmented fish respectively) were weighed, measured and filleted. The deskinned left and rightfilletswere individually weighed, colour scored with the Roche Colour Card, and packaged under vacuum in oxygen permeable ( 0 transmission; 2200 2  cc/m /24 hr, H 0 transmission; 7.8 g/m /24 hr) and oxygen barrier (Winpak Deli bags, O2 2  2  2  transmission; 2.3 cc/m /24 hr, H 0 transmission; 7.8 g/m /24 hr) bags. Fillets were stored at 2  2  2  -3°C in one of two Forma Scientific Model 3710 lower temperature incubators (Forma Scientific Mariette, OH) retrofitted with an AMF Paragon Model 8145-00 D-Frost-O-Matic Tome Control (Paragon Electric Guelph, ON) set to induce four defrost cycles per day to minimize temperaturefluctuationsin the packaged fish during defrosting. Storage of fish at -3°C is common in Japan and accepted as a high quality product. Fish muscle freezes at -5°C, therefore, storage at -3°C results in a semi-frozen product with minimal damage at the cellular level (Uchiyama, 1988). The temperature of the refrigerators was monitored every other day to ensure consistency. Fourfilletsfrom each of the pigment treatments (pigmented and unpigmented) were taken as day zero samples common to both storage treatments for that pigment group. Thesefilletswere packaged under vacuum in foil laminate oxygen barrier bags ( 0 transmission; 0.7 cc/m /24 hr, H 0 transmission; 1.2 g/m /24 hr) and stored at 2  2  2  2  -35° C to prevent further oxidation prior to analysis. At 2, 4, and 6 weeks, fourfilletsfrom each of the four treatments (pigmented, permeable & barrier; unpigmented, permeable &  78  barrier) were randomly selected and removed from storage. Upon selection, the fillets vacuum packaged in foil laminate bags and stored at -35°C pending analyses.  4.1.2 Analysis of Fillets 4.1.2.1 Colourimetric Measurements Hunter L, a, b scores were recorded, as described in section 3.1.6.2, on allfilletsprior to their homogenization for chemical analyses.  4.1.2.2 Lipid Analysis Lipid analysis offillethomogenates was performed based on the method of Bligh and Dyer (1959). Fillets were homogenized in a Waring blender as described in section 3.1.6.3. A 4 g sample of the homogenate was blended in a Sorvall® Omni-mixer (Ivan Sorvall, Inc. Norwalk, Conn. USA) at 9724 RPM with 20 ml methanol with 0.5% butylated hydroxyl toluene (BHT) (Sigma Chemical Co.), and 10 ml chloroform for 120 seconds (chloroform and methanol were from Burdich & Jackson, double distilled, HPLC grade), then with an additional 10 ml chloroform for 30 more seconds, and 8 ml distilled water for another 30 seconds. This suspension was subsequently filtered through Whatman No. 1 filter paper to remove solids. The filtrate was left to stand for one hour in a 50 ml glass graduated cylinder to allow separation into two phases. Thereafter, the volume of the lower chloroform layer was recorded and the upper methanol phase suctioned off. Five ml of the chloroform layer were pipetted into a pre-weighed aluminum weigh boat and heated in a 70-80°C water bath in a fume hood to drive off the chloroform. The weigh boat was then transferred to a 100°C drying oven for 1 hour to drive off any residual chloroform. After being allowed to cool in a  79  desiccator for 30 minutes, the weight of the boat was taken to determine, by difference, the amount of lipid in the sample.  4.1.2.3 Methyl Ester Preparation and Analysis  Base-catalyzed methyl esterification was performed based on the method described by Christie (1973). From the lipid extract, an aliquot of the chloroform layer that ensured a minimum of 0.06 g methyl ester was pipetted into a clean test tube and concentrated to ~1 ml by bubbling through with nitrogen gas (medical grade, Praxair) in a 30°C water bath. One ml benzene (OmniSolv®, distilled) and 0.2 ml sodium methoxide (0.5% sodium metal in methanol, Aldrich Chemicals) were added to the concentrated sample, and the tube was heated in a 50°C water bath for 10 minutes. Five ml of hexane (Burdich & Jackson, double distilled, HPLC grade), 5 ml distilled water and 0.1 ml glacial acetic acid (AnalaR®) were then added to the sample which, subsequently, separated into two layers. The top layer was transferred via pipette to a new test tube to which 5 ml of additional hexane were added. A small amount (tip of a spatula) of anhydrous sodium sulphate (Fisher Scientific Co.) was added to remove any residual water. The liquid was then transferred via pipette into a new test tube. The sample in the test tube was concentrated under nitrogen gas in the 30°C water bath to < 2 ml, then transferred to a 2 ml gas chromatography (GC) vial (Varian) and made up to volume with hexane. Samples were stored at -18°C until shipping to WVL for analysis via GC. The GC conditions are reported in Table 9.  80  4.2 Statistical Analysis  Minitab Statistical Software, version 12.0, was used to perform analysis of variance (ANOVA) using the General Linear Model (presence of pigment, storage time, packaging material, presence of pigment*storage time, presence of pigment*packaging material, storage time*packaging material). Significance was determined at a level of 5% (p = 0.05).  81  Table 9 - G C Conditions Used for Analysis of Methyl Esters  GC Integrator Detector type Column Column Manufacturer Internal Diameter Column Length Injection temperature Final temperature Detector temperature He flow rate Air flow rate N flow rate 0 flow rate Volume of sample injected 2  2  Varian 3400 Varian 4290 Flame ionization SP-2330 Supelco, Inc. 0.25 mm 30m 220°C 195°C 250°C 0.75 ml/min. 0.75 ml/min. 0.75 ml/min. 0.75 ml/min. 0.1 ul  82  4.3 Results 4.3.1 Colour Measurements  Hunterlab colour values are presented in Table 10. The difference in Hunter a values was significant (p < 0.05) between the pigmented and unpigmented groups offish, as was expected based on visual inspection and knowledge of the concentration of dietary Ax that had been fed to each group for seven months prior to slaughter. Neither storage time, packaging material, nor interactions between the three factors significantly influenced Hunter a. Although Hunter L and b values were recorded in Experiment 2, only Hunter a values were analyzed statistically since they correlated best with flesh astaxanthin concentrations in Experiment 1.  4.3.2 Lipid Analysis  The amount of lipid in the fillets and pertinent information regarding the degree of unsaturation derived from analysis of the methyl esters are presented in Table 11. Percentages of crude lipid in the fillets were significantly higher in the unpigmented fillets and in the fillets selected for six weeks of storage (p > 0.05). The type of packaging material did not influence crude lipid levels and there were no interactions found between the three factors Time was also found to be significant (p < 0.05) with respect to its influence on U:S (total unsaturated fatty acids: saturated fatty acids) and P:S (polyunsaturated fatty acids (20:5n3, 22:5n3, 22:6n3): saturated fatty acids) of the fillets. This trend was one of increasing degree of unsaturation over time with respect to U:S and decreasing degree of unsaturation over time with respect to P:S. Dietary treatment, packaging material and interactions between the three factors did not significantly influence the U:S or P:S in the fillets.  83  Table 1 0 - Hunter L (lightness), a (redness), and b (yellowness) Values of Pigmented and Unpigmented Chinook Salmon Fillets stored in Oxygen Permeable (aerobic) or Impermeable (barrier) Bags for Zero to Six Weeks at -3°C 1  Presence of Pigment unpigmented  Packaging Material n/a  Storage Time (weeks) 0  L  a  pigmented  1 2 3  6  barrier  6  n/a  0  aerobic  6  barrier  6  b  40.18 (1.68) 51.81 (1.39) 49.40 (3.64)  2.76 (1.48) 4.90 (2.65) 4.15 (1.78)  10.41 (0.89) 13.28 (1.45) 12.67 (0.77)  34.33 (2.74) 42.69 (4.43) 42.04 (3.03)  17.20 (2.38) 14.09 (4.56) 16.92 (5.00)  12.61 (0.49) 14.50 (0.72) 14.36 (1.30)  3  aerobic  2  Values are the average of 4filletsper treatment Dietary treatment was indicated as significant (p < 0.05) Standard deviations are shown in parentheses  84  Table 11 - Percentages of Crude Lipid and Ratios of Total Unsaturated to Saturated Fatty Acids (U:S), and Polyunsaturated to Saturated Fatty Acids (P:S) of Pigmented and Unpigmented Chinook Salmon Fillets stored in Oxygen Permeable (aerobic) or Impermeable (barrier) Bags for Zero to Six Weeks at ^"C 1  Presence of Packaging Storage Pigment Material Time (weeks) 0 unpigmented n/a  % Crude Lipid (wet basis) 4.24 (0.81) 5.99 (1.04) 6.94 (1.47) 4  pigmented  1 2 3 4  aerobic  6  barrier  6  n/a  0  aerobic  6  barrier  6  3.57 (1.51) 4.82 (0.59) 4.70 (1.60)  2,3  U:S  2  P:S  2  3.33 (0.21) 3.49 (0.11) 3.46 (0.05)  1.05 (0.19) 0.95 (0.08) 0.97 (0.06)  3.19 (0.06) 3.99 (1.03) 3.47 (0.31)  1.17 (0.24) 1.06 (0.21) 0.94 (0.18)  Values are the average of 4filletsper treatment Storage time was indicated as significant (p < 0.05) Presence of pigment was indicated as significant Standard deviations are shown in parentheses  85  Discussion Analysis of variance of the ratios of total unsaturated to saturated fatty acids (U:S) and polyunsaturated to saturated fatty acids (P:S) between the fillets at zero and six weeks storage revealed no difference due to the presence of pigment, nor any interaction between pigmentation and storage time or type of packaging material (p > 0.05), hence, the results from the two and four weekfilletswere not reported. These findings suggest that astaxanthin had neither antioxidative or prooxidative activity and are in agreement with those of Sigurgisladottir et al. (1994) and Ingemansson et al. (1993) who also found astaxanthin to be ineffective as an antioxidant in salmonid fillets. Although time was found to be significant with respect to U:S, this was most likely not a meaningful result since the trend shown was one of increasing degree of unsaturation over time. It is known that the degree of unsaturation decreases with lipid peroxidation over time owing to the relative instability of unsaturated fatty acids. As stated by Nawar (1996), "autooxidation of saturated fatty acids is extremely slow; at room temperature, they remain practically unchanged when oxidative rancidity of unsaturates becomes detectable". Time was also indicated as significant (p < 0.05) with respect to P:S. The trend shown for P:S was one of decreasing degree of unsaturation over time. Whereas this trend is in keeping with the above statement by Nawar (1996), if oxidation had indeed occurred, a difference should have been observed based on the packaging material, fillets in the oxygen permeable bags deteriorating at a faster rate as compared to fillets in the oxygen barrier packaging. Difference due to packaging material was not observed (> 0.05) , indicating that error due to the small number of observations had most likely occurred.  86  Since it is probable that there were no real differences between groups with respect to the extent of lipid peroxidation, it is possible that six weeks of storage at -3°C was not long enough for differences to be detected. Alternatively, the determination of fatty acid ratios may not be sufficiently sensitive as a measure of lipid peroxidation in the initial stages. Other methods of lipid peroxidation, such as determination of thiobarbituric acid reactive substance (TBARS) or iodine values, unfortunately could not be conducted here due to the limited amount of each sample that was available. As seen in Experiment 1, analysis of Hunter a results showed that the pigment concentration of the diet significantly affected the colour of the fillets (p < 0.05), thereby indicating differences between pigmented and unpigmented groups of fish. Neither length of storage time nor type of packaging material had any significant effect on colour as represented by Hunter a. When the lipid levels (%) of the left and right fillets of the fish in this study were compared, it was observed frequently that there were marked differences. These likely arose from inconsistencies in the filleting of the fish due to their small size. Thus, in some instances the fillets may have contained different proportions of muscle from the tail, ventral and anterior regions which are known to be different in lipid content (Refsgaard et al., 1998). Differences in the relative shape of the fillets (ie. long and thin, containing the tail section, vs. short and wide, containing more of the ventral and dorsal region, but perhaps missing the tail section) could have resulted in increased variation and less accurate results. Statistical analyses of the crude lipid levels of the fillets showed that both diet and time were significant factors (p < 0.05). The packaging material did not significantly influence the results and there were no interactions between diet, time and packaging material. The fact  87  that diet and time were significant was surprising since the two diets had similar lipid levels as shown in Experiment 1. Furthermore the stored fillets showed no signs of detectable lipid peroxidation as evidenced by the ratios of unsaturated to saturated fatty acids. Therefore, the level of lipid should not have been affected. These results might be explained by differences in the rearing conditions of the groups of fish and the dissimilar incidence of disease between the groups during the latter part of period 2 in Experiment 1. Allfishpreviously fed the unpigmented diet were from a relatively healthy group (15% mortalities), as were the day zero fish given the pigmented diet (7% mortalities). By contrast, the fillets from fish which represented the pigmented diet after six weeks of storage came from a less healthy group (24% mortalities) which ate relatively poorly throughout the experiment. This likely would have resulted in those fish having less body fat than fish from the healthier groups and consequently explains the significant effects observed for diet and time in this experiment. It was not intended for the fillets in any treatment to come from the fish of one particular tank and efforts were made to randomize the original distribution of fillets. However, the rather large down scaling of this experiment, after storage and relegation to respective groups took place, seems to have resulted in what would be considered a non-random distribution of fillets that precluded accurate treatment comparisons to be made.  4.5 Conclusions The results of this study suggest that astaxanthin acted neither as an antioxidant nor as a prooxidant infilletsduring storage at -3°C. However, since an accelerated decrease in the ratio of polyunsaturated to saturated fatty acids was not observed infilletsstored in oxygen permeable packaging as compared to fillets stored in oxygen barrier packaging, it is probable  that changes in fatty acids susceptible to oxidation had not occured after six weeks storage at -3°C. The findings also indicate that there were inconsistencies in the process o f filleting which led to differences in fillet lipid content. Accordingly, there is a need to improve the methodology o f whole fillet removal from smolt fish.  89  5 General Conclusions The results of this research indicate that juvenile chinook salmon are capable of pigment deposition in thefleshwell before reaching 80g body weight if supplemental Ax is provided in the diet. Though a direct relationship is shown between the dietary Ax level and the resultant Ax concentration and colour of the flesh, 55 ppm added synthetic Ax in the diet appears to be adequate for chinook salmon weighing between 50 and 360 g to attain an acceptable degree offleshpigmentation when fed over a period of at least seven months. As mentioned previously, the results further suggest that it would be worthwhile to investigate the potential for supplying lower levels of Ax in the diet (ie. 35 ppm added) starting early in the life history of the fish. This strategy would take advantage of the greater retention efficiency of the lower concentration of Ax in the diet resulting in less wastage of Ax while ensuring that the fish would be of desirablefleshcolour by the time they reached market weight. The Hunter a value was found to be a very good predictor offlesh[Ax] (r = 0.901 p = 0.000), and is considered preferable to the RCC (r = 0.847, p = 0.000), as a predictor of flesh [Ax]) in rating the degree of redness in chinook salmon fillets. Using Hunter a to assess the colour of salmon fillets is advantageous because it is a rapid, inexpensive, non-destructive measurement as compared to chemical quantification of Ax, and is more discriminating than the eight point ranking scale of the RCC. Astaxanthin was found to have neither antioxidative nor prooxidative activity in chinook salmonfilletsstored at -3°C for zero to six weeks in oxygen permeable and barrier bags. It is possible that six weeks of storage at -3°C was not long enough for lipid peroxidation to occur, or that the determination of fatty acid ratios may not be sufficiently  90  sensitive as a measure of lipid peroxidation in the initial stages. An alternate method of assessing lipid peroxidation, such as determination of TBARS, iodine value or headspace analysis of volatile components* should be conducted along with the determination of fatty acid ratios in order to allow for confidence in the methodology.  6  REFERENCES  Ando, S., Yamauchi, H., Hatano, M., and Heard, W.R. 1992. Comparison of muscle compositions between red- and white-fleshed chinook salmon (Onchorhynchus tshawytscha). Aquaculture, 103: 359-365 Andrewes, A., and Starr, M . 1976.(3R,3'R)-Astaxanthinfromthe yeast Phaffia rhodozyma. Phytochem., 15: 1009-1011 Bermond, P. 1990. Biological effects of food antioxidants. In: Food antioxidants. Hudson, B.J.F.(Ed.) p. 193-252. Elsevier Applied Science, London. Bjerkeng, B., Storebakken, T., and Liaaen-Jensen, S. 1992. Pigmentation of rainbow trout from start feeding to sexual maturation. Aquaculture, 108: 333-346 Bjerkeng, B., and Johnsen, G. 1995. Frozen storage quality of rainbow trout (Oncorhynchus mykiss) as affected by oxygen, illumination, and fillet pigment. J. Food Sci. 60: 285-288 Bjerkeng, B., Refstie, S., Fjalestad, K.T., Storebakken, T., Rodbotten, M., and Roem, A.J. 1997a. Quality parameters of the flesh of Atlantic salmon (Salmon salar) as affected by dietary fat content and full-fat soybean meal as a partial substitute for fish meal in the diet. Aquaculture, 157: 297-309 Bjerkeng, B., Foiling, M., Lagocki, S., Storebakken, T., Olli, J.J., and Alsted, N. 1997b. Bioavailability of all-E-astaxanthin and Z-isomers of astaxanthin in rainbow trout (Oncorhynchus mykiss). Aquaculture, 157: 63-82 Bligh, E.G., and Dyer, W.J. 1959. A rapid method of total lipid extraction and purification. Can. J. Biochem. Physiol. 37:911-917 Burton, G.W., and Ingold, K.U. 1984. P-carotene: an unusual type of lipid antioxidant. Science 224: 569-573 CIE (1976) Official Recommendations on Uniform Color Space, Color difference Equations and Metric Color Terms. Suppl. No. 2 to CIE Publication No. 15, Colorimetry. Commission International de l'Eclairage, Paris Chen, H., Meyers, S.P., Hardy, R.W. and Biede, S.L. 1984. Color stability of astaxanthin pigmented rainbow trout under various packaging conditions. J. Food Sci. 49: 1337-1340 Choubert, G., and Luquet, P. 1983. Utilization of shrimp meal for rainbow trout (Salmo gairdneri Rich.) pigmentation. Influence of fat content of the diet. Aquaculture, 32: 19-26 Choubert, G. 1985. Effects of starvation and feeding on canthaxanthin depletion in the muscle of rainbow trout (Salmo gairneri Rich.). Aquaculture, 46: 293-298  92  Choubert, G., and Storebakken, T. 1989. Dose response to astaxanthin and canthaxanthin pigmentation of rainbow trout fed various dietary carotenoid concentrations. Aquaculture, 81: 69-77 Choubert, G., Blanc, J., and Courvalin, C. 1992. Muscle carotenoid content and colour of farmed rainbow trout fed astaxanthin or canthaxanthin as affected by cooking and smokecuring procedures. Int. J. Food Sci. Technol., 27: 277-284 Christiansen, R., Struksnaes, G., Estermann, R., and Torrissen, O.J. 1995. Assessment of flesh colour in Atlantic salmon, Salmo salar L. Aquaculture Research, 26: 311-321 Christie, W.M. (Ed.) 1973. The preparation of volatile derivatives of lipids. Ch. 4. In: Lipid Analysis. P. 90. Pergamon Press, Oxford Craik, J.C.A. 1985. Egg quality and pigment content in salmonid fishes. Aquaculture, 47: 6188 Foss, P., Storebakken, T., Schiedt, K., Liaaen-Jensen, S., Austreng, E., and Streiff, K. 1984. Carotenoids in diets for salmonids. I. Pigmentation of rainbow trout with the individual optical isomers of astaxanthin in comparison with canthaxanthin. Aquaculture, 41: 213-226 Gadient, M., and Fenster, R. 1994. Stability of ascorbic acid and other vitamins in extruded fish feeds. Aquaculture, 124: 207-211 Gjerde, B., and Gjedrem, T. 1984. Estimates of phenotypic and genetic parameters for carcass traits in Atlantic salmon and rainbow trout. Aquaculture, 36: 97-110 Henmi, H., Hata, M., and Takeuchi, M. 1991. Studies on the carotenoids in the muscle of salmon - V. Combination of astaxanthin and canthaxanthin with bovine serum albumin and egg albumin. Comp. Biochem. Physiol., 99B: 609-612 Higgs, D.A., Markert, J.R., McQuarrie, D.W., McBride, JR., Dosanjh, B.S., Nichols, C , and Hoskins, G. 1979. Development of practical dry diets for coho salmon Oncorhynchus kisutch using poultry-by-product meal, feather meal, soybean meal and rapeseed meal as major protein sources. In: Finfish Nutrition andFishfeed Technology, J.E. Halver, and K. Tiews (Eds.), Vol 2, Heenemann Verlagsgesellschaft MbH., Berlin, p 191-218 Hirayama, O., Nakamura, K., Hamada, S., and Kobayasi, Y. 1994. Singlet oxygen quenching ability of naturally occurring carotenoids. Lipids, 29: 149-150 Hoffmann-La Roche. 1992. Determination of stabilized astaxanthin in carophyll pink, premixes, and fish feeds. Hoffmann-La Roche, Consumer Service for Feed and Food, Basel, Switzerland. 1pp. Hunt, R.W.G. 1987. Measuring Colour. Ellis Horwood Ltd. Chichester, England, p 17-104  93  Ingemansson, T., Pettersson, A., and Kaufmann, P. 1993. Lipid hydrolysis and oxidation related to astaxanthin content in light and dark muscle of frozen stored rainbow trout (Oncorhynchus mykiss). J. Food Sci. 58: 513-518 Iwamoto, R.N., Meyers, J.M., and Hershberger, W.K. 1990. Heritability and genetic correlations for flesh coloration in coho salmon. Aquaculture, 86: 181-190 Johnson, E.A., Conklin, D.E., and Lewis, M J . 1977. The yeast Phaffia rhodozyma as a dietary pigment source for salmonids and crustaceans. J. Fish. Res. Board Can. 34: 2417-2421 Johnson, E.A. Villa, T.G., and Lewis, M.J. 1980. Phaffia rhodozyma as an astaxanthin source in salmonid diets. Aquaculture, 20:123-134 Johnson, E.A., and An, G.H. 1991. Astaxanthin from microbial sources. Crit. Rev. Biotechnol., 11: 297-326 Kiessling, A., Higgs, D.A, Dosanjh, B.S., and Eales, J.G. 1994. Influence of sustained exercise at two ration levels on growth and thyroid function of all-female chinook salmon (Oncorhynchus tshawytscha) in seawater. Can. J. Fish. Aquat. Sci., 51: 1975-1984 Kiessling, A , Dosanjh, B., Higgs, D., Deacon, G., and Rowshandeli, N . 1995. Dorsal aorta cannulation: a method to monitor changes in blood levels of astaxanthin in voluntarily feeding Atlantic salmon, Salmo salarL. Aquaculture Nutrition, 1: 43-50 King, T.L. 1996. Use of colorimetric analysis to estimate salmonid flesh carotenoid content. Prog. Fish Cult. 58: 215-218 Kitihara, T. 1983. Behaviour of carotenoids in the chum salmon (Oncorhynchus keta) during anadromous migration. Comp. Biochem. Physiol. 76B(1): 97-101 Kotik, L.V., Tolokonnikov, G.Y., and Dubrovin, V.N. 1979. The effect of krill meal additions to feeds on muscle pigmentation in the rainbow trout (Salmo gairdneri). J. Ichthyol., 19: 114123 Krieberg, H., and Powell, J. 1991. Metomidate sedation reduces holding stress in chinook salmon. World Aquacult. 22: 58-59. Little, A.C., Marinsen, C , and Sceurman, L. 1979. Color assessment of experimentally pigmented rainbow trout. Color Research and Application, 4: 92-95 McCallum, I., Cheng, K., and March, B.E. 1987. Carotenoid pigmentation in two strains of chinook salmon (Oncorhynchus tshawytscha) and their crosses. Aquaculture, 67: 291-300 Mann, J. 1999. Personal Communication, EWOS Canada Ltd. Surrey, British Columbia  94  March, B.E., and MacMillan, C. 1996. Muscle pigmentation and plasma concentrations of astaxanthin in rainbow trout, chinook salmon, and Atlantic salmon in response to different dietary levels of astaxanthin. Prog. Fish Cult., 58: 178-186 Miki, W. 1991. Biological functions and activities of animal carotenoids. Pure Appl. Chem. 63: 141-146 Mortensen, A., and Skibsted, L.H. 1997. Relative stability of carotenoid radical cations and homologue tocopheroxyl radicals. A real time kinetic study of antioxidant hierarchy. FEBS Letters, 417): 261-266 Nakano, T., Tosa, M., and Takeuchi, M . 1995. Improvement of biochemical features in fish health by red yeast and synthetic astaxanthin. J. Agric. Food Chem. 43: 1570-1573 Nawar, W.W. 1996. Lipids. In: Food Chemistry, Third ed. p. 256 & 273, Fennema, O.R. (Ed.). Marcel Dekker, Inc. New York No, H.K., and Storebakken, T. 1991a. Pigmentation of rainbow trout with astaxanthin at different water temperatures. Aquaculture, 97: 203-216 No, H.K., and Storebakken, T. 1991b. Color stability of rainbow trout fillets during frozen storage. J. Food Sci., 56: 969-972, 984 Palozza, P., and Krinsky, N.I., 1992. Astaxanthin and canthaxanthin are potent antioxidants in a membrane model. Arch. Biochem. Biophys., 297: 291-295 Refsgaard, H.H.F., Brockhoff, P.B., and Jensen, B. 1998. Biological variation of lipid constituents and distribution of tocopherols and astaxanthin in farmed Atlantic salmon (Salmo salar). J. Agric. Food Chem., 46: 808-812 Saito, A. 1969. Color in raw and cooked Atlantic salmon {Salmo salar). J. Fish. Res. Board Can., 26: 2234-2236 Sanderson, G.W., and Jolly, S.O. 1994. The value salmonid fish. Aquaculture, 124: 193-200  Phaffia yeast as a feed ingredient for  Schierle, J., and Hardi, W. 1994. Revised supplement: Determination of stabilized astaxanthin in CAROPHYLL® pink, premixes and fish feeds. In: Analytical Methods for Vitamins and Carotenoids in Feeds. Hoffman, P., Keller, H.E., Schierle, J., and Schuep, W. Roche, Switzerland Schiedt, K., Leuenberger, F.J., and Vecchi, M . 1981. Natural occurrence of enantiomeric and weso-astaxanthin. 5. Ex wild salmon {Salmo salar and Oncorhynchus). Helv. Chim. Acta, 64: 449-457  95  Schmidt, P.J., and Idler, D.R. 1958. Predicting the color of canned sockeye salmon from the color of the raw flesh. Food Technol., 12(1): 44-48 Sigurgisladottoir, S., Parrish, C C , Lall, S.P., and Ackman, R.G. 1994. Effects of feeding natural tocopherols and astaxanthin on Atlantic salmon (Salmo salar)filletquality. Food Res. Intl. 27: 23-32 Simpson, K.L., Katayama, T., and Chichester, C O . 1981. Carotenoids in fish feeds. Ch.4 In: Carotenoids as Colourants and Vitamin A Precursors. J. C. Bauernfeind (Ed.) p. 504. Academic Press, Inc. New York, New York Sivtseva, L.V. 1982. Qualitative composition and distribution of carotenoids and vitamin A in the organs and tissues of rainbow trout, Salmo gairdneri. J. Ichthyol., 22: 96-100 Skrede,G., and Storebakken, T. 1986a. Instrumental colour analysis of farmed and wild Atlantic salmon when raw, baked and smoked. Aquaculture, 53: 279-286 Skrede, G., and Storebakken, T. 1986b. Characteristics of color in raw, baked and smoked wild and pen-reared Atlantic salmon., J. Food Sci., 51:804-808 Smith, B.E., Hardy, R.W., and Torrissen, O.J. 1992. Synthetic astaxanthin deposition in pansize coho salmon (Oncorhynchus kisutch). Aquaculture, 104: 105-119 Spinelli, J., Lehman, L., and Wieg, D. 1974. Composition, processing, and utilization of red crab (Pleuroncodesplanipes) as an aquacultural feed ingredient. J. Fish. Res. Board Can. 31: 1025-1029 Spinelli, J., and Mahnken, C. 1978. Carotenoid deposition in pen-reared salmonids fed diets containing oil extracts of red crab (Pleuroncodes planipes). Aquaculture, 13: 213-223 Spinelli, J. 1979. Preparation of salmonid diets containing zooplankton and their effect on organoleptic properties of pen-reared salmonids. In: J.E. Halver and K. Tiews (Eds.), Finfish Nutrition and Fishfeed Technology., H, Heenemann GmbH & Co, Berlin 42, 2: 383-392 Storebakken, T., Foss, P., Austreng, E., and Liaaen-Jensen, S. 1985. Carotenoids in diets for salmonids. II. Epimerization studies with astaxanthin in Atlantic salmon. Aquaculture, 44: 259-269 Storebakken, T., Foss, P., Schiedt, K., Austreng, E., Liaaen-Jensen, A., and Manz, U. 1987. Carotenoids in diets for salmonids IV. Pigmentation of Atlantic salmon with astaxanthin, astaxanthin dipalmitate and canthaxanthin. Aquaculture, 65: 279-292 Storebakken, T., and Choubert, G. 1991. Flesh pigmentation of rainbow trout fed astaxanthin or canthaxanthin at different feeding rates in freshwater and saltwater. Aquaculture, 95: 289295  96  Terao, J. 1989. Antioxidant activity of carotene-related carotenoids in solution. Lipids, 24: 659-661 Torrissen, O., and Brakken, O.R. 1979. The utilization of astaxanthin-forms by rainbow trout (Salmo gajrdneri) In: J.E. Halver and K. Tiews (Eds.), Finfish Nutrition and Fishfeed Technology., H, Heenemann GmbH & Co, Berlin 42, 2: 377-382 Torrissen, O.J., Tidemann, E., Hansen, F., and Raa, J. 1981. Ensiling in acid - A method to stabilize astaxanthin in shrimp processing by-products and improve uptake of this pigment by rainbow trout (Salmo gairdneri). Aquaculture, 26: 77-83 Torrissen, O.J. 1984. Pigmentation of salmonids - Effects of carotenoids in eggs and start feeding diet on survival and growth rate. Aquaculture 43: 185-193 Torrissen, O.J., and Naevdal, G. 1984. Pigmentation of salmonids - Genetical variation in carotenoid deposition in rainbow trout. Aquaculture, 38: 59-66 Torrissen, O.J. 1985. Pigmentation of salmonids: Factors affecting carotenoid deposition in rainbow trout (Salmo gairdneri). Aquaculture, 46: 133-142 Torrissen, O.J. and Naevdal, G. 1988. Pigmentation of salmonids - Variation in flesh carotenoids of Atlantic salmon. Aquaculture, 68: 305-310 Torrissen, O.J., Hardy, R. W., and Shearer, K.D. 1989. Pigmentation of salmonids. Carotenoid deposition and metabolism. CRC Crit. Rev. Aquat. Sci., 1: 209-225 Torrissen, O.J., and Christiansen, R. 1994. Strategies for the pigmentation of salmon. Norsk Fiskeoppdrett nr.2A-94: 19-21, 48 Uchiyama, H. 1988. Biochemical determination of fish freshness and partial freezing as a means of keeping freshness of fish and its products. Proc. Aquaculture International Congress & Exposition, Vancouver, B.C., Canada Whyte, J.N.C., Travers, D., and Sherry, K.L. 1998. Deposition of astaxanthin isomers in chinook salmon {Oncorhynchus tshawytscha) fed different sources of pigment. Can. Tech. Rep. Fish. Aquat. Sci., 2206: 33 p Withler, R.F. 1987.Genetic variation in flesh pigmentation of chinook salmon (Oncorhynchus tshawytscha). Proc. World symposium on Selection, Hybridization, and Genetic Engineering in Aquaculture, Bordeaux, France, 27-30 May 1986, 1: 421-429 Woodall, A . A , Wai-Ming Lee, Weesie, R.J., Jackson, M.J., and Britton, G. 1997. Oxidation of carotenoids by free radicals: relationship between structure and reactivity. Biochim. Biophys. Acta 1336: 33-42  97  A P P E N D I X  1- R A W D A T A  diet s a m p l i n g Ctrl.  1  a  F O R F I L L E T S F R O M  L  b  RCC  hue  1  E X P E R I M E N T  chroma  Ax ( p p m ) fish  weight (g)  2  -0.73  42.72  9.58  10.0  94.63  9.60  0.11050  155.40  Ctrl.  2  1.87  42.67  10.91  10.0  80.21  11.06  0.03200  140.60  Ctrl.  2  4.97  39.56  11.61  11.0  67.03  12.62  0.17550  203.10  Ctrl.  2  1.95  40.59  9.69  10.0  78.49  9.88  0.11050  160.10  Ctrl.  2  2.55  38.91  8.04  10.0  72.19  8.43  0.11220  136.60  Ctrl.  2  0.81  41.79  10.01  10.0  85.37  10.04  0.06600  163.10  Ctrl.  2  1.33  40.65  10.35  10.0  82.50  10.43  0.06400  83.70  2  Ctrl..  2  0.35  41.81  9.77  10.0  88.23  9.77  0.00000  115.90  Ctrl.  2  3.16  42.27  10.77  11.0  73.91  11.22  0.09100  172.40  Ctrl.  2  -0.39  40.87  9.39  10.0  92.34  9.39  0.00000  125.00  Ctrl.  2  7.37  42.87  11.55  12.5  57.29  13.70  0.21120  359.30  Ctrl.  2  3.51  38.81  10.18  11.0  71.04  10.76  0.11700  139.60  Ctrl.  2  1.30  40.53  9.00  10.0  81.93  9.09  0.08040  162.20  Ctrl.  2  5.77  38.71  11.43  12.0  63.02  12.80  0.10400  170.50  Ctrl.  2  4.43  47.37  12.81  11.5  71.04  13,55  0.12920  213.70  Ctrl.  2  2.84  46.10  12.29  10.0  76.77  12.61  0.16640  204.60  Ctrl.  2  3.58  38.21  9.80  11.0  69.90  10.43  0.07920  155.20  Ctrl.  2  -1.44  42.69  7.54  10.0  100.93  7.67  0.00000  106.80  Ctrl.  2  6.01  46.95  12.90  12.0  64.74  14.23  0.21760  258.70  Ctrl.  2  1.17  44.79  10.66  10.0  83.65  10.72  0.04550  151.30  Ctrl.  2  0.03  42.07  10.04  10.0  89.95  10.04  0.03900  150.50  Ctrl.  2  6.28  44.46  12.98  12.0  64.17  14.41  0.19800  244.20  Ctrl.  2  -0.98  39.90  8.09  10.0  96.92  8.14  0.04480  Ctrl.  2  3.18  40.79  9.86  10.0  72.19  10.36  0.10980  low  2  13.69  39.96  15.44  15.5  48.70  20.63  1.28360  185.50  low  2  7.14  40.70  11.66  14.0  58.44  13.67  0.49080  184.90  low  2  3.76  45.11  12.04  12.0  72.76  12.61  0.46170  94.30  low  2  17.67  37.97  15.80  16.5  41.82  23.70  2.99330  240.60  , ;  116.50 160.00  low  2  14.20  31.75  11.32  16.5  38.38  18.15  1.67790  230.70  low  2  12.05  35.95  12.86  14.5  46.98  17.62  1.98560  110.10  low  2  16.10  44.17  15.89  15.0  44.69  22.62  1.79900  244.30  low  2  13.80  32.25  11.62  15.5  40.10  18.04  1.68170  169.20  low  2  3.28  45.24  11.83  12.5  74.48  12.27  0.79400  126.90  low  2  18.02  40.53  15.98  15.0  41.82  24.08  1.74080  224.80  low  2  18.14  33.02  13.85  17.0  37.24  22.82  3.31890  257.70  low  2  11.00  36.04  12.73  15.0  49.27  16.82  1.07250  105.50  low  2  15.34  34.79  11.89  16.0  37.81  19.40  1.83650  167.00  low  2  17.27  32.22  12.40  16.0  35.52  21.26  2.88750  165.00  low  2  13.11  32.81  11.92  14.0  42.39  17.71  1.07970  145.40  low  2  15.72  37.31  14.29  15.5  42.39  21.24  1.72020  190.00  low  2  8.23  39.11  12.36  13.0  56.15  14.84  0.73600  166.90  98  diet sampling low low low low low low low med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. high high high high high high high high high high high high  2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2 2  a  L  10.42 18.09 10.42 15.14 15.35 14.17 14.57 17.36 13.73 17.02 19.21 9.47 23.87 15.30 14.36 12.52 19.35 12.79 6.58 14.84 17.08 19.32 12.69 20.25 17.33 16.90 18.14 15.51 14.32 14.20 18.93 14.34 19.17 17.90 20.67 16.41 17.79 19.69 4.09 15.48 14.15 14.16 13.32  34.61 31.44 42.97 38.42 37.13 38.11 37.69 33.15 33.40 31.08 32.76 35.98 37.29 37.06 30.20 33.61 32.59 32.95 36.78 37.56 33.01 31.62 33.41 34.74 32.28 35.93 33.49 30.34 35.90 35.91 32.47 47.03 31.96 36.14 32.71 33.49 36.30 33.92 42.84 34.56 35.71 35.50 41.50  b 9.84 12.85 14.05 15.02 12.99 14.51 14.58 11.69 11.43 12.70 14.54 11.51 17.17 14.01 12.29 11.99 14.27 11.71 10.92 15.43 14.76 14.03 11.29 15.30 12.49 12.00 12.65 11.30 11.50 13.25 14.38 15.63 14.17 14.53 15.11 12.49 14.19 14.97 11.47 13.36 12.00 11.62 14.51  RCC hue  chroma Ax (ppm) fish weight (g)  16.0 17.0 14.5 15.5 15.0 15.5 15.0 16.5 15.0 15.5 15.5 12.5 17.0 14.0 15.0 14.5 16.0 14.0 11.5 14.0 16.0 16.0 15.5 16.0 16.0 16.5 16.5 16.0 15.5 15.0 16.5 14.0 16.5 15.0 16.5 16.0 16.0 16.0 11.0 15.5 15.0 15.0 15.0  14.33 22.18 17.49 21.32 20.10 20.28 20.61 20.92 17.86 21.23 24.09 14.90 29.40 20.74 18.90 17.33 24.04 17.34 12.74 21.40 22.57 23.87 16.98 25.38 21.36 20.72 22.11 19.18 18.36 19.42 23.77 21.21 23.83 23.05 25.60 20.62 22.75 24.73 12.17 20.44 18.55 18.31 19.69  43.54 35.52 53.28 44.69 40.10 45.83 45.26 33.80 39.53 36.66 37.24 50.42 35.52 42.39 40.68 43.54 36.66 42.39 59.01 45.83 40.68 36.09 41.82 37.24 35.52 35.52 34.95 36.09 38.96 42.97 37.24 47.55 36.66 38.96 36.09 37.24 38.38 37.24 70.47 40.68 40.10 39.53 47.55  0.91980 2.46720 1.05600 1.37920 1.58930 1.38880 1.50480 2.61140 0.06970 2.37830 2.40170 0.85400 4.44930 1.10550 1.95840 1.14560 3.85130 1.41700 0.24700 1.48170 2.23040 1.91730 1.29360 3 .83790 2.20770 3.27680 2.47280 3.91530 2.56320 1.64480 2.49480 1.27600 3.00200 2.69760 3.28640 1.93760 2.53110 2.18080 0.29250 1.59250 0.92300 2.11250 1.39200  204.00 166.20 234.70 202.40 158.50 153.80 144.40 223.30 138.70 178.40 150.80 124.70 371.00 149.60 118.70 144.20 312.80 156.10 75.50 161.80 142.30 246.00 178.90 221.40 183.00 182.70 277.60 132.50 148.70 144.40 292.50 143.30 217.10 166.80 183.50 121.50 188.90 172.50 134.70 118.10 123.40 111.80 112.50  99  diet sampling high high high high high high high high high high high high  Ctrl. Ctrl. Ctrl. Ctrl.  Ctrl. Ctrl.  Ctrl. Ctrl. Ctrl.  Ctrl. Ctrl. Ctrl.  Ctrl. Ctrl. Ctrl. Ctrl. Ctrl. Ctrl. Ctrl. Ctrl. Ctrl.  Ctrl. Ctrl. Ctrl, low low low low low low low  2 2 2 2 2 2 2 2 2 2 2 2  a 13.29 15.67 14.45 16.33 17.02 16.36 15.77 15.14 12.02 12.64 9.53 14.05 2.62 3.88 1.80 3.12 2.89 2.94 2.03 0.36 3.23 1.13 2.81 0.93 4.41 2.88 4.22 5.65 2.89 3.86 3.65 2.61 3.84 3.70 3.97 4.10 16.07 14.51 9.30 5.18 1.49 12.78 12.41  31.23 29.15 31.13 24.31 26.21 31.75 25.66 31.03 26.91 27.64 27.97 25.42 44.71 41.24 42.52 37.34 37.00 39.60 44.09 40.92 37.35 42.15 37.79 40.08 41.45 52.20 45.34 41.21 38.87 50.02 44.35 45.99 42.80 46.86 40.79 35.59 36.47 35.41 36.45 39.08 41.59 33.64 35.09  10.94 14.43 12.40 10.55 11.52 13.84 11.01 12.49 10.05 10.79 9.92 10.89 12.54 12.09 11.79 10.49 10.01 10.37 11.55 9.46 10.31 10.99 10.97 9.84 10.93 13.50 12.78 12.40 10.95 14.24 12.69 11.29 11.32 12.02 11.73 9.37 12.64 13.60 11.88 9.83 9.74 12.84 11.99  RCC  hue  15.0 17.0 15.0 17.0 16.0 16.0 16.5 16.0 16.0 15.5 14.0 15.5 10.0 11.0 10.0 11.0 11.0 11.5 10.0 10.0 11.0 10.0 11.5 10.0 11.0 10.0 11.0 12.0 11.0 10.0 11.0 10.0 11.0 11.0 10.0 11.0 15.0 14.0 12.0 11.0 10.0 14.0 14.0  39.53 42.39 40.68 32.65 34.37 40.10 34.95 39.53 40.10 40.68 46.41 37.81 77.92 72.19 81.36 73.33 73.91 73.91 80.21 87.66 72.76 84.22 75.63 84.79 68.18 77.92 71.62 65.31 75.05 75.05 73.91 76.77 71.04 72.76 71.04 66.46 38.38 42.97 52.13 62.45 81.36 45.26 44.11  chroma Ax (ppm) fish weight (g) 17.21 21.30 19.04 19.44 20.55 21.42 19.23 19.62 15.66 16.61 13.75 17.77 12.81 12.69 11.92 10.94 10.41 10.77 11.72 9.466 10.80 11.04 11.32 9.883 11.78 13.80 13.45 13.62 11.32 14.75 13.20 11.58 11.95 12.57 12.38 10.22 20.44 19.88 15.08 11.11 9.853 18.11 17.25  1.16800 1.84930 2.02560 2.48630 3.29340 2.51550 2.26200 2.12550 1.91680 1.14660 1.60000 1.31840 0.08710 0.06030 0.04550 0.05030 0.05760 0.08910 0.03630 0.02310 0.09570 0.02970 0.07040 0.02640 0.11550 0.07150 0.04950 0.10390 0.03840 0.06600 0.10890 0.03900 0.04880 0.10080 0.07800 0.09100 2.56100 1.83630 0.62080 0.41600 0.05310 1.31630 1.05920  191.10 195.30 143.70 180.20 182.70 190.10 167.50 136.20 132.60 123.20 115.60 140.90 113.60 156.50 97.10 109.10 109.60 136.10 101.50 96.00 139.60 93.30 115.80 117.40 126.60 109.60 101.70 111.10 112.10 141.50 118.50 150.50 122.50 139.40 137.50 134.60 125.10 106.10 100.70 92.60 54.80 101.40 154.60  100  diet sampling low low low low low low low low low low low low low low low low low med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. med. high high  a 17.66 10.45 13.55 7.75 12.56 17.65 15.70 10.89 15.90 12.80 16.41 12.77 12.63 14.72 14.40 14.78 11.40 17.68 18.19 16.33 12.42 20.35 7.99 13.76 12.95 0.38 13.96 17.62 13.15 10.87 16.65 13.40 14.27 14.93 14.26 5.75 11.94 4.53 11.52 14.48 13.78 17.51 15.89  RCC 32.75 36.34 31.52 38.50 38.45 31.72 34.52 36.93 34.36 36.83 35.78 38.00 38.88 32.86 33.68 35.36 40.20 37.03 33.99 35.63 37.39 32.56 38.71 30.56 32.90 44.85 38.46 33.27 36.19 36.21 32.55 38.02 38.09 39.54 36.79 42.63 38.24 41.08 31.52 36.10 35.94 36.80 28.99  13.19 11.91 11.04 11.78 14.70 12.87 13.01 11.95 12.87 13.35 14.51 13.83 14.35 13.25 12.68 13.24 12.86 14.51 14.23 12.69 13.24 14.67 11.63 10.50 10.83 10.35 12.61 13.55 13.35 11.33 12.35 12.94 14.41 13.00 12.87 12.15 12.59 10.97 10.42 13.13 12.62 13.76 10.87  15.0 13.0 15.0 12.0 13.0 16.0 14.5 14.0 15.0 14.0 14.0 14.0 13.0 14.0 15.0 14.0 14.0 15.5 16.0 15.0 14.0 16.0 13.5 14.0 16.0 10.0 15.0 15.5 15.0 15.0 16.0 15.0 14.0 16.0 15.0 11.0 14.0 11.0 14.0 14.0 14.5 15.0 16.0  hue chroma Ax (ppm) fish weight (g) 36.66 48.70 38.96 56.72 49.27 36.09 39.53 47.55 38.96 46.41 41.25 47.55 48.70 43.54 41.25 41.82 48.70 39.53 37.81 37.81 46.98 35.52 55.57 37.24 40.10 88.23 41.82 37.81 45.26 46.41 36.66 44.11 45.26 41.25 41.82 64.74 46.41 67.60 42.39 42.39 52.71 38.38 34.37  22.04 15.84 17.47 14.10 19.33 21.84 20.39 16.16 20.45 18.49 21.90 18.82 19.11 19.80 19.18 19.84 17.18 22.87 23.09 20.68 18.15 25.08 14.11 17.30 16.88 10.35 18.81 22.22 18.73 15.70 20.73 18.62 20.28 19.79 19.20 13.44 17.35 11.86 15.53 19.54 16.01 22.26 19.25  2.89600 0.78980 1.35360 0.59520 0.89280 2.02340 2.63340 1.12130 2.89080 1.78860 2.38520 1.43380 0.88770 1.54770 2.43100 2.01960 1.40030 2.61300 1.50280 2.61800 1.21720 2.57620 0.75380 1.74240 2.4540 0.05610 2.30010 1.04980 1.23760 1.35300 2.78050 1.64900 1.00170 2.74370 1.76880 0.53150 1.90130 0.29820 1.22850 1.72520 1.71407 3.28350 1.87110  135.90 108.70 129.20 112.60 101.60 129.90 124.90 139.00 148.50 113.00 98.70 122.40 134.70 124.30 146.60 125.90 105.60 151.60 150.40 122.10 122.20 159.80 116.50 118.00 115.70 61.70 131.90 113.20 116.80 122.20 124.90 155.50 91.70 134.40 148.90 133.10 110.50 129.70 130.70 159.60 110.40 122.10 165.20  101  diet sampling  a  RCC  hue  chroma Ax (ppm) fish weight (g)  10.98 34.30 11.28 15.0 45.83 15.74 1.14840 high 18.45 34.27 13.24 16.0 35.52 22.70 3.14900 high 10.65 35.58 11.38 15.5 46.98 15.58 1.31990 high 18.28 37.98 15.04 15.0 39.53 23.67 2.65650 high 15.22 32.47 11.71 17.0 37.81 19.20 1.58730 high 12.13 37.60 12.93 16.0 46.98 17.72 0.84420 high 14.98 32.32 12.74 17.0 40.10 19.66 2.31000 high 15.00 32.81 11.79 16.0 38.38 19.07 2.96670 high 16.43 32.92 13.82 15.0 40.10 21.46 2.24580 high 13.33 34.87 12.26 15.5 42.39 18.11 2.17430 high 19.08 41.92 16.15 16.5 40.10 24.99 2.46760 high 17.82 38.44 14.77 17.0 39.53 23.14 2.08230 high 16.30 38.27 14.29 14.0 41.25 21.67 2.20110 high 21.15 36.87 16.03 17.0 37.24 26.53 2.91450 high 17.72 34.53 14.12 14.5 38.38 22.65 1.86930 high 16.45 34.33 13.52 15.0 39.53 21.29 2.13840 high 24.21 38.96 18.08 16.0 36.66 30.21 4.29990 high 20.19 42.13 16.69 14.0 39.53 26.19 2.72160 high 20.90 32.15 14.91 15.0 35.52 25.67 2.99330 high 16.65 34.41 14.23 16.0 40.68 21.90 1.99880 high 14.79 35.27 13.21 15.0 41.82 19.83 2.80640 hieh 15.76 34.90 14.30 15.0 42.39 21.28 2.08230 'Sampling 2, Jan. 23 & 24, 1997; sampling 1, Oct. 1 & 2, 1996 Control diet  high  118.80 128.20 86.30 136.60 135.80 112.10 103.30 122.80 98.40 80.60 107.70 154.00 108.60 158.00 122.50 144.60 164.20 147.10 142.70 120.90 129.20 126.90  2  102  ft)  o O O  VD LO  o  ftt O O O  o  o  o  O  o  U j rH r o ro n  rco  o  ro  o  CO  • r o ON T ) CO Ft; VD  vo CM ^ rH  Px  CO VD r o CO CO o  r - CM r~ CM  w  • r o ON. T ) VD (=C O LO  c n VD r O rH rH CM CM  CO VD r o CO 00 o  r - CM r r - CM c o  VD CO O LO  ON. VD ON CO rH CM rH CM CM r -  X  o w  < U KH  H  KH  H  < H I  cs KH  fc KH  ro CD  -P C  M  0  IM  <D 0  c  •H  UJ r o Q  rH r o  >i rH  (0  c  <  O  N  CO ON vD LO CM  S  •  •  •  •  LO CO CM vD •ro O rH O rT j vO CM <! ro  CO VD VD ^ O CO • • • • r - co r - r H •ro rH i—I O CM T i CO VD O < O "=3* •vT rH  SC  o  CO VD VD ^ O CO • • • • r - c o r— rH O rH rH O CM 0 00 vD O ^ CO o •=3" rH 1  ^ rH CO ON rH rH  (1) 0 C  (0  •H M  UJ r o  rH r o  P  VD <3* vD «C LO LO  G •H rH DJ CJN  G  g  rd • H CO a> rH -K >H rH u M +J O J - U O rd M • H iH O o •rH CO T i to T> w EH  d a) g a>  -H rH  0  OJ  W  5  >1  •H  rH  ON  M-l  •H CO  ^  CO ON  (0 >  0  (0  O  <w  > CO •H  O  Pn LO VD LO CNJ vD  w m ro cr> • • • •  W  (M O CM LO VD O  ON  C  g  CD  -H  rd CO  U  rH  -X  rl  3  4J  0)  O T - Hi CO  ft4J  g  0)  rd T - ai W  M  O  M  M W  rd O  103  o o  o  •  o  VO LO  ro o  00 o  •  o  o  CL(  o  •  fe  CO o CO  CL, o V O C O o ro o  ro o o  o  •  o  o  o  •  o  o  O LO ro 00 r - LT)  fe  •  ro  CN]  o  o  •  LO  o  •  o  CTl  o  I—1 00 LO  CN) LO  o r o LO  -ro r- o TJ CO (< 00  •  o  CO r o CN] o CO  co S  O  o  LO  rH r~ ro co r- LO  co oo vO  •r-i rH CN] TJ ro rH <J CN]  •ro ro cn  v o rH  T i ON  CN] rH 00  LO  LO  CO rH rH O o CO L O LO <C CTi  CO LO r CO L O VD  LO  •r-i r o CN] CTi CN] T ) CTi rH rH vO  • l — l CTl  <!  ro  LO  TJ  U  CO CN] rH LO CN] CO rH r o 00 r rH o T i rH < ! LO  • O  CN]  O  u  VD o  p  ro  rH r rH  0 H 0  r< LD  <D  (4 CO rH rH O o rH CO LO LO <=T CTi r o  o  ffH  Oi u 0  t y r o CN| CTi CN] CU CTi rH rH VO  u 0  CN]  Hn  CD 0  c  rO •H  fe Q  VO o  ro  r o rH r o  rH rH  C,  >1  to  <  u  U  rd  - H CO rH -* iH rH 4-) O rd 4-1 cu g 0) iH 4-)  d O - H rd • H U co T3 CO T i W  <D 0  c  (0 •rH  fe Q  r o rH r o  o  EH  a <2  ^00  rH  CTi  rH rH  u (0  CO •H CO >i H rd  CO CO  CN] CTi  CO v o  0)  C  •H  c P K  M  0 lH  <D 0  C  fd •H  0  g rd  CO •H CO  a  rH -* o M 4-) ft 4-J d cu g cu O - H rd - H CO T3 CO T i  CT cn CTl ^ r-- rH cu r - LO VD VO rH rrH CM  w  fe Q  r o rH r o  iH rH o rd  r4 4-)  iH W  o  EH  >1  H fd  c  <  rH cn rH rH CO  M  rH  CO  CO LO r— L O LO r o CO LO v o cn VD  >  •H  CT  a)  •P  rd  CT C  0  &  CU  Q  >  •H rH  0  rH  rH  00 CTi  Cn  > (0  CO  rH r - CO rH * T  M  CO •H  CTl  CN]  t-H  Xi 0  o  rH  rH  0  CU rH CO L O  CTl  < r-  CN]  CO CN] rH LO CN] rH CO rH r o oo r - o  u  cn  ^ rT ) t-~ L O CD VO  u  £•  LO  CJi  c  •H  rH  CT CU  U iH d O  CO  C  ft g  -iH Crd O rH -K 4-> ft 4-J CD g cu -H rd -rH T ) CO T l  SH O iH iH W  rH rd  4-1  O  EH  VO I D  OJ O  o O o  o CM  o o  o  o  c n CM  o  O  LO  o  [n O H CO VO VO  ^  "3" v o  r-  co  co vo  ro  •  •  o o f< LO  T J  cn  vo co CO 0 0 r o CM CO • • n • T j rH r o < vo rH  •  •  o rH  <N  • g m  cn  • rH ro  a H  0  tw  rt  g  &  •  a.  rro  CD rH CO rH r - < r ro ro ro CO v o rH * r vo  rH r o  Q  W  c  <  rr- m rCO rH rH • • •r-i • T j CM O LO  LO ro ro  • 00  <J CM  cn  (0  CV  CO rH VO r -  LO c n ro ro  H h  H  0  co  r - LD r -  CO rH rH r o  CD CM O CO CM  LO  LO CO v o  c n CM  rH  CM  IH  «a* rH 00 cn rH rH  0  c  rrj  •H H  tu Q  r o rH r o  cn  CM  S  (0  rH (ti  +J PH-P 0 CD e a) M -M - H (0 • H SH O W Tj  C PH  •H  d (ti  rH  CO T5  cn •H rH  0  -H  CD  0 s-l 3 O  >  C -H rH  H  (0  QJ  >1 rH  O  Ci (ti • H CO rH -tt  rd  U 4-i 3 cu  <  O  -H  S CD (ti - H  CO  X!  CO T>  C  rH CO c n  rH rH  rd  0  rd  o  ai  >  >1 H  LO  rH  rH  cn  M ni  10  r-  O  co  CO 0 0 v o CM CM r o r~ r o r co CM CO c n v o v o  •H  •H  ^  ro  <«  0  c  LO CM  CO rH v o  ro vo  H  CD •P C  ro  rn  T3 o  CM CM rcn  ro  CO [— VO VO  S ^ m m rr-  o  cri  vo o r-  ^  *i 1  o  VO CM CM  <3<  CM rH CM  CO  o  • • •  OH 4-)  u rH(ti 0 -P uO u w 105  A P P E N D I X 3- F I L L E T F A T T Y A C I D P R O F I L E S F O R E X P E R I M E N T 2  Unpigmented. week 0 17B  25B  1  %  %  35B %  40B %  Corr. GC Corr. GC Fatty Acids RT GC Corr. GC 2.93 2.05 4.40 2.21 14:0 6.29 2.19 4.32 0.88 16:0 10.67 7.70 15.18 4.46 14.87 7.43 15.96 8.34 4.10 2.18 4.68 1.68 16:1n7 12.20 3.07 6.05 1.23 3.35 1.64 3.73 1.56 18:0 17.89 1.90 3.75 1.12 18:1n9 19.37 11.40 22.47 6.17 20.57 10.20 21.92 11.42 0.42 0.90 0.43 18:1n7 20.00 0.61 1.20 4.63 2.26 4.86 2.68 18:2n6 21.80 2.84 5.60 1.39 1.14 0.31 1.03 0.53 1.14 0.58 18:3n3 24.80 0.58 20:1n9 26.48 26.6 4.25 8.38 2.37 7.90 4.24 9.11 4.71 18:4n3 20:4n6 31.9 4.63 9.13 2.94 9.80 5.07 10.89 5.14 22:1n11 32.7 2.80 5.52 1.53 5.10 2.59 5.57 2.57 20:5n3 33.7 1.70 0.09 0.19 0.22 22:4n6 38.94 0.11 0.22 0.51 1.55 1.17 40.13 0.74 1.46 0.77 2.57 0.72 22:5n3 22:6n3 41.50 7.91 15.59 6.32 21.07 7.20 15.47 8.00 100 50.79 50.73 100 30.00 100 46.54 Total % 23.24 21.53 23.72 Tot. Sat. (s) 76.28 78.47 Tot. Unsat. 76.76 (u) u:s 3.30 3.64 3.22 0.95 0.97 1.33 P :s 5.16 % Unident. 6.14 3.38 Fish identification Retention time Percent area under gas chromatograph peak Percentage corrected for content of BHT and unidentified peaks p =(20:5n3, 22:5n3, 22) Percentage of unidentified peaks 2  5  6  3  4  Corr. 4.35 16.42 3.31 3.23 22.48 0.85 5.28 1.14 9.27 10.12 5.06 0.43 2.30 15.75 100 24.00 76.00 3.17 0.96 6.53  1  2 3 4  5 6  106  Unpigmented. Week 6. Aerobic packaging 14A  24A  1  %  Fatty Acids 14:0 16:0 16:1n7 18:0 18:1n9 18:1n7 18:2n6 18:3n3 20:1n9 18:4n3 20:4n6 22:1n11 20:5n3 22:4n6 22:5n3 22:6n3 Total % Tot. Sat. (s) Tot. Unsat.  %  3  4  %  %  Corr. GC RT GC Corr. GC 3.81 1.52 6.29 2.37 4.53 1.13 10.67 8.05 15.38 4.39 14.81 5.64 4.86 1.92 12.20 2.70 5.16 1.44 3.24 1.17 17.89 1.67 3.19 0.96 19.37 11.66 22.28 6.55 22.09 8.31 0.81 0.29 20.00 0.08 0.15 0.24 5.30 1.65 21.80 2.46 4.70 1.57 1.15 0.39 24.80 0.56 1.07 0.34 1.34 26.48 2.01 3.84 3.84 1.91 26.6 2.55 4.87 1.14 0.63 2.12 31.9 5.35 10.22 3.63 12.24 3.91 32.7 6.04 1.92 33.7 2.61 4.99 1.79 6.68 1.49 38.94 1.79 3.42 1.98 1.82 0.89 40.13 1.17 2.24 0.54 41.50 7.30 13.95 3.32 11.20 5.08 100 37.43 52.33 100. 29.65 23.10 21.85 76.90 78.15 2  32A  30A  Corr. GC Corr. 4.01 4.06 1.68 15.07 6.24 14.88 4.74 5.13 1.99 3.08 3.13 1.29 22.20 8.79 20.96 0.76 0.77 0.32 4.43 4.41 1.86 1.03 1.04 0.43 3.65 3.58 1.53 4.01 5.10 1.68 10.45 5.03 11.99 6.51 5.13 2.73 3.46 3.98 1.45 2.24 2.38 0.94 13.57 5.98 14.26 100 41.94 100 21.96 22.25 78.04 77.75  (u) 3.58 3.49 3.33 u:s 0.95 0.92 0.87 P :s 5.05 10.77 % Unident. 8.06 Fish identification Retention time Percent area under gas chromatograph peak Percentage corrected for content of BHT and unidentified peaks p =(20:5n3, 22:5n3, 22) Percentage of unidentified peaks 5  6  3.55 1.05 4.53  1  2 3 4  5 6  107  Unpigmented. Week 6. Oxygen barrier packaging 14B  24B  1  %  Fatty Acids 14:0 16:0 16:1n7 18:0 18:1n9 18:1n7 18:2n6 18:3n3 20:1n9 18:4n3 20:4n6 22:1n11 20:5n3 22:4n6 22:5n3 22:6n3 Total % Tot. Sat. (s) Tot. Unsat. itu  %  30B %  Corr. GC RT GC Corr. GC 6.29 2.47 4.55 0.78 3.91 2.14 10.67 8.03 14.79 3.08 15.44 7.78 5.21 2.64 12.20 2.76 5.08 1.04 3.21 1.63 17.89 1.68 3.10 0.64 19.37 12.02 22.14 4.62 23.16 11.11 0.15 0.41 20.00 0.08 0.15 0.03 5.36 2.35 21.80 2.57 4.73 1.07 0.59 1.09 0.22 1.10 0.54 24.80 26.48 2.18 4.02 0.89 4.46 2.47 4.55 0.72 3.61 4.50 26.6 31.9 32.7 5.76 10.61 2.28 11.43 5.74 4.61 3.06 33.7 3.40 6.26 0.92 1.35 1.42 38.94 1.83 3.37 0.27 2.31 1.21 40.13 1.21 2.23 0.46 41.50 7.23 13.32 2.93 14.69 7.79 100 19.95 100 52.32 54.28 22.56 22.44 77.44 77.56 2  3  4  32B %  Corr. GC Corr. 4.09 1.41 4.08 5.33 15.41 14.87 5.05 1.76 5.09 3.27 3.12 1.13 21.23 7.71 22.29 0.43 0.78 0.15 4.49 1.65 4.77 1.03 0.38 1.10 8.60  3.06  8.85  10.97 5.02 14.51 5.85 2.87 8.30 2.71 2.98 2.31 1.00 9.02 14.89 3.12 100 34.59 100 22.75 22.08 77.25 77.92  i \  ( ) 3.53 3.46 3.43 u:s 1.04 0.96 0.97 P :s 5.39 5.98 10.74 % Unident. Fish identification Retention time Percent area under gas chromatograph peak Percentage corrected for content of BHT and unidentified peaks p =(20:5n3, 22:5n3, 22) Percentage of unidentified peaks 5  6  3.40 0.89 13.78  1  2 3 4  5 6  108  Pigmented. Week 0 124D  124C  1  %  %  124B %  124A %  Corr. GC Corr. GC Fatty Acids RT GC Corr. GC 3.64 0.56 6.29 4.21 1.42 14:0 2.59 4.57 1.64 16:0 10.67 9.10 16.07 6.25 16.05 6.47 16.57 2.96 4.21 1.64 4.20 0.77 12.20 2.92 5.16 1.64 16:1n7 17.89 3.75 1.54 3.94 0.89 18:0 1.86 3.29 1.46 18:1n9 19.37 13.12 23.17 8.50 21.83 7.94 20.34 3.61 0.06 0.15 0.08 20.00 0.11 0.19 18:1n7 4.83 1.78 4.56 0.84 18:2n6 21.80 2.94 5.19 1.88 1.00 1.10 0.39 18:3n3 24.80 0.65 1.15 0.43 1.81 4.65 1.75 4.48 0.71 20:1n9 26.48 26.6 5.20 9.18 1.50 3.85 1.23 3.15 0.60 18:4n3 20:4n6 31.9 8.91 3.10 7.94 1.41 22:1n11 32.7 5.78 10.21 3.47 1.80 4.62 1.82 4.66 0.98 20:5n3 33.7 3.03 5.35 38.94 0.23 0.41 0.13 0.33 0.15 0.38 22:4n6 2.00 0.83 2.13 0.31 22:5n3 40.13 1.06 1.87 0.78 22:6n3 41.50 8.03 14.18 7.64 19.62 8.92 22.85 5.10 100 18.82 100 38.93 100 39.04 Total % 56.62 24.02 24.15 Tot. Sat. (s) 23.93 75.85 76.07 75.98 Tot. Unsat. (u) u:s 3.18 3.16 3.14 1.23 0.89 1.09 P :s 6.51 6.98 5.97 % Unident. Fish identification Retention time Percent area under gas chromatograph peak Percentage corrected for content of BHT and unidentified peaks p =(20:5n3, 22:5n3, 22) Percentage of unidentified peaks 2  5  6  3  4  Corr. 1.54 14.36 2.39 4.10 13.42 2.91 0.60 1.79 1.71 2.91 5.73 1.97 1.20 45.38 100 23.44 76.56 3.27 1.45 10.31  1  2  3 4  5  6  109  Pigmented. Week 6. Aerobic packaging 2A  4A  1  %  %  12A %  Corr. GC Fatty Acids RT GC Corr. GC 4.45 1.22 2.29 4.37 1.74 14:0 6.29 10.67 8.40 16.02 5.92 15.14 5.01 16:0 5.22 1.69 16:1n7 12.20 2.28 4.35 2.04 18:0 17.89 1.89 3.60 1.24 3.17 1.24 18:1n9 19.37 12.08 23.04 8.61 22.03 7.89 20.00 0.06 0.15 18:1n7 21.80 2.85 5.44 1.77 4.53 1.67 18:2n6 0.95 0.03 18:3n3 24.80 0.48 0.92 0.37 4.14 20:1n9 26.48 2.22 4.23 1.62 4.37 2.82 18:4n3 26.6 2.15 4.10 1.71 0.71 1.82 0.04 20:4n6 31.9 22:1n11 32.7 5.63 10.74 4.21 10.77 3.13 5.32 1.40 20:5n3 33.7 2.86 5.45 2.08 3.15 0.09 22:4n6 38.94 1.40 2.67 1.23 1.92 1.77 40.13 1.01 1.93 0.75 22:5n3 41.50 6.89 13.14 5.03 12.87 5.96 22:6n3 52.43 100 39.09 100 33.96 Total % 23.99 22.77 Tot. Sat. (s) 77.23 76.01 Tot. Unsat. 2  ii (u)  i\  3  4  16A %  Corr. GC Corr. 3.59 1.19 3.33 14.75 5.77 16.16 4.45 4.98 1.59 3.78 3.65 1.35 23.23 7.48 20.95 0.14 0.05 4.70 4.92 1.68 1.09 0.36 1.01 7.73 8.30 2.76 0.12 8.12 9.22 2.90 4.73 4.12 1.69 0.27 0.11 0.31 1.79 5.21 0.64 17.55 8.14 22.79 100 100 35.71 23.27 22.00 76.73 78.00  3.39 3.55 3.17 u:s 0.86 0.88 1.22 P :s 8.28 4.87 5.55 % Unident. Fish identification Retention time Percent area under gas chromatograph peak Percentage corrected for content of BHT and unidentified peaks p=(20:5n3, 22:5n3, 22) Percentage of unidentified peaks 5  6  3.30 1.26 5.90  1  2 3 4  5 6  no  Pigmented. Week 6. Oxygen barrier packaging 2B  4B  1  %  %  Fatty Acids 14:0 16:0 16:1n7 18:0 18:1n9 18:1n7 18:2n6 18:3n3 20:1n9 18:4n3 20:4n6 22:1n11 20:5n3 22:4n6 22:5n3 22:6n3 Total % Tot. Sat. (s) Tot. Unsat.  RT 6.29 10.67 12.20 17.89 19.37 20.00 21.80 24.80 26.48 26.6 31.9 32.7 33.7 38.94 40.13 41.50 2  GC 1.12 4.28 1.04 1.04 5.42 0.31 1.33 0.24 1.06 1.06 0.32 2.67 1.33 1.70 0.47 3.49 26.88 3  %  %  Corr. GC Corr. GC 4.17 0.84 3.50 1.19 15.92 3.27 13.64 4.44 4.55 1.44 3.87 1.09 3.87 0.86 3.59 1.04 20.16 4.62 19.27 6.64 1.15 0.10 0.42 0.04 4.95 1.02 4.26 1.39 0.89 0.23 0.96 0.29 3.94 0.97 4.05 1.14 2.29 1.27 3.94 0.55 1.19 9.93 2.41 10.05 3.08 5.17 1.36 4.95 1.24 9.09 2.26 6.32 2.18 1.75 0.63 2.63 0.52 12.98 3.96 16.52 4.70 100 30.80 100 .23.97 20.73 23.96 79.27 76.04 4  16B  12B  Corr. GC Corr. 4.54 3.86 2.43 14.42 8.22 15.35 5.53 4.68 2.96 3.57 3.38 1.91 21.56 13.04 24.35 0.13 5.30 4.51 2.84 1.14 0.94 0.61 3.70 4.12 4.88 9.11 10.00 5.68 10.61 4.74 4.42 2.54 2.58 7.34 1.38 1.55 1.69 0.83 15.26 6.23 11.63 100 100 53.55 21.66 23.45 76.55 78.34  ii (u) A 3.62 3.17 3.82 u:s 1.17 0.99 0.82 P s 15.55 4.85 5.56 % Unident. Fish identification Retention time Percent area under gas chromatograph peak Percentage corrected for content of BHT and unidentified peaks p=(20:5n3, 22:5n3, 22) Percentage of unidentified peaks 6  3.26 0.76 5.87  1  2  3 4  5 6  ill  APPENDIX 4- RAW DATA FOR FILLETS FROM EXPERIMENT 2 diet week 1  2  bag  3  u:s  4  p:s  5  a  L  b  %CL  6  75 0 3.27 1.45 20.35 31.68 13.13 2.66 n/a 75 0 3.14 1.23 17.65 32.80 12.88 3.55 n/a 3.16 1.09 15.00 37.91 12.35 5.71 75 0 n/a 0 3.18 0.89 15.78 34.92 12.06 2.36 75 n/a 3.17 0.89 9.14 48.27 14.14 4.07 75 6 A 3.39 0.88 11.57 44.11 14.84 5.36 75 6 A 75 6 A 3.55 1.22 16.47 39.99 13.69 4.64 3.30 1.26 19.18 38.39 15.31 5.22 75 6 A B 3.17 0.82 10.09 44.56 12.43 2.35 75 6 6 B 3.82 1.17 16.40 44.28 15.04 5.37 75 3.62 0.99 19.91 41.24 15.23 5.14 75 6 B 75 6 B 3.26 0.76 21.28 38.09 14.74 5.95 3.87 3.30 0.97 0.88 42.09 9.85 0 0 n/a 3.28 3.64 1.33 2.55 38.02 9.47 0 0 n/a 40.59 11.28 5.01 0 0 3.22 0.95 3.16 n/a 40.03 11.04 4.81 0 0 3.17 0.96 4.43 n/a 0 6 A 3.33 0.92 1.47 51.05 12.78 7.08 50.25 12.27 5.03 0 6 A 3.58 0.87 4.83 0 6 A 3.49 0.95 7.91 52.72 12.65 5.16 53.22 15.43 6.67 0 6 A 3.55 1.05 5.39 0 B 3.46 0.97 2.16 54.48 11.86 8.96 6 3.43 0.96 3.43 45.85 12.50 5.68 0 6 B 3.53 1.04 6.34 48.91 13.72 6.04 0 6 B 48.36 12.61 7.09 0 6 B 3.40 0.89 4.66 ppm astaxanthin (added) indiet (75- pigmented, 0- unpigmented) Storage time at -3°C A- oxygen permeable, B- oxygen barrier Ratio of unsaturated to saturated fatty acids Ratio of polyunsaturated (20:5n3, 22:5n3,22:6n3) to saturated fatty acids Percent crude lipid 1  L  3 4  5  6  112  APPENDIX 5- G C PROFILES FOR EXPERIMENT 2 Pigmented fillet at 0 weeks - upper Pigmented fillet after 6 weeks storage at -3°C, aerobic packaging - lower Peak numbers indicate retention time.  113  P-i c n rH r o M< CM r o VO o rH rH c n rH 00 CM c n 00 CM o o fa  o  o  O  o  o  c o VO v o r o o vo o CM c n o  vo o  rH CM o rH  o  rH o  CC CC Tj -H o co c o o CD • • • cc CM CM CM 4-1 CO  CO CO CO CM r - CO •1—1 CO r-~ Tj O ro  <  o  •  o  LO vo rH o o  CO o CM o [— o o O o  LO vo rH o o  ro o cn CM o  o  O  o  o  o  rH (ti d Tj -H CO CD  CC  CO CO CO CO rH CM r - co •r-i r o r x) o ro  <  o  •  o  LO vo rH O o  00 o CM o r- o o o o  LO vo rH o o  00 00 LO CM r-  o  o  o  o  o  fa Q  00 o CM o r- o o o o  LO vo rH o o  00 00 LO CM r-  CM 00 cn CO rH  o  o  o  o  o  rH  rH rH rH rH rH rH LO rH CM c o  CD  u  +  4-) U 4-1 CD d cu Cn CU O "rH - H (ti • H CO T j 4-1 Xi T )  e  Dl (ti (ti  Xi Xi  -tt 4-1 CD •H TJ  o 4H CO d o •H 4-1 (ti  >  cy CO X!  o  0)  e  o  o  •  •  •  o  •  o  •  Q o 4-1 CO  o  o  -M c n • H rH rvo  rH oo CN ro ro  cn rH CM CO CO  ro  00  o o o  o o o  d  LO vo rH o o  -H 4->  o  o > • (D  00 CM VO r00  o  rH 00 ro ro  fa  •>  o  cn rH ro 00  - P cn cn cn • H vo vo vo cn cn cn  CO  co r o CO rH r~ tTro cu o  rH CO CM LD ro  -tt u rH (L> o (ti M 4-> -H M o EH 4-1  6  w  fa  •  ro  10 o •• o o CM 00  •  rH (ti CO d d CD  ro  CO  o  •  ^  vo vo  •  ro  •  ro  00 CM rH rH CM  CO +J CO CD tH  O  fa o T  O  O  O  n  CD H  ro cri n (M  CD H  o  o  o  i—I ^  O  o  o O CD CO CO CD o r - CM rH O CM CD o CTi CD rH  CU  O i CO CO CM CTi LO CM ^ ^ C\ CD d l <Tl CM O CD LO LD CD  o  O  fa CTi o cn o LO o  H O <H W CO TJ CD 4J 0)  o  O  O  O CM r O  O  • rH rd  O TJ CC • H LO r o CO o • CD CCCM CM 1 1 -P CO  CM ^ r - CM  CM o  .  o  rH  co  cnincomcoLOiocri S r - r - C M o o ^ c M C O LO vH LO O CTi LD CM -r—i <=J< O rH O O CO T J O rH o o o o o o  o  o  o  o  o  2  c o rH CD r o CM CD LO rH c n ^ r H CO o  • n CM o TJ CTi < CM rH  o  rH CO CM rd r- CM cn 00 TJ r o CO • • -H CO LO CD CD 1 1  CM CO CD CM CO CM  •r-i  c CO  3  CC  CO CO rH CD r o CM *D r o rH c n ^ r H r o CO "3" o  c o L n o o L O c o L n L D c o c o r - r - C M c o ^ c M t H LO i—i LD o c n LO r~ -|—|'C <3<OrHOOrH X ! O I - H O O O O C O <J o o o o o o o  - n CM o T j CTi f< CM rH  ,  - P CD CD - H rH rH cn cn ro ro • • CD rH rH  CM co CD CM rH CM CM CM  fa  >  Q  10  ix M O  in CD O C rd •H H  c o L n c o L O o o L O L O c o L O c o r - r - C M o o ^ c M r H L O LO i—I LO O CTl LO I ' LO D ^ ^ O r H O O r H O O C D O r H O O O O C O O CO O O O O O O O r H  fa  LO r-H CM r o  Q cw O (0 •H CO >1  r-i rd  M 0  m CD  o c  rd •H M rd >  CO 0 0 rH CD CO CM CD r o rH r o rH c n CO <3* o D CM o CD CTi CO CM rH 1  fa  Q  •H  0  u  >H 4-1 3 cu O -rH COTj  m  4-J Xi Xi -tt * -tt M i—I 4-i -p at o m D l QJ CD g H 4-1 rd - H - H - H S-l O 4J QTj'Tj4->faEH r  (0 •H (0 >i rH rd  i  H  CM r o CD CM rH CO CM CM CM LO rH  CD O  U 4-J  3 CD O -M CO TJ  CD g -H 4-J  O  LM  co  c o  CD g -H 4-> -tt -M Cn CD rd •rH x5 TJ  •rH 4-J rd  > S-l CD CO  fa  rd o o  o o cn rH o • • cn o rH  o rH  Cn D i rd rd  -t< -tt 4-J CD CD g • H -r-i T j 4->  U O U  rH rd 4-J  w  EH  M  U  rd TJ  a  rd -M  co  CD Di H rd rH rd  xi •H  4-J CO CM •rH CM co r LO • • CD rH rH  Xi Xi Xi  TJ CD N -H TJ  00  rH rH rH rH rH r H LO rH CM ro  0 CD g D i tr>  rd  4-1 CO  TJ -rH CO CD SH  o  rd  CO  C  CO c n CO rH Xi  o  o  -H 4-> rd  > U CD CO  Xi o  a rd CO CD -M O  c  CD TJ  CC  curr -- ll o O  fe  rH O CNJ VD o o ro ro ro O VD CN] L O VD  O  O  o  LO  o  T3  rH o  o  rH  PH  PH  - H CM CNJ rH to i H r o O CD • • • pr. CNJ CM CN]  CTi r o CNJ  CO VO CO CNJ LO  r-  o  •  r H rd  o  1  4-1 CO  TS  -H to CD SH  Ti CD N  CO CTi VD CNJ -n X) r H fiC r H  a  r - rH r- rH rH r o o L O rH CD L O r r o CO rH L O r o ro o CNJ  CN] o  o  rH CTl CTi rd VD rH <=r rr- 00 - H CN] 10 • • CD CM CM  rH  -a  ed  CO CT) r- rH r— rH rH r o L O rH C O L O rH CO VD o CNJ r o r o r-1 L O r o 00 •r~i rH r o o CNJ o o VD rH CNJ ro  4-J •H  <  hH  KH  dP  o\°  U  U  0 lH  <D 0 C  id •H H rd  >  CJ  CO CTi r - rH r~ rH rH r o 00 L O rH oo L O rH C O CO CD o CNJ r o r o rH L O r o 00 r -  cr CU i H  ro o CO rH CNJ  (0  >i H id  c  <  o  rH rH rH rH rH rH Q  0 CO •H  CNJ o  VD * r r o r-  r H CNJ r o LO  CD  Cn Cn rd rd  XiXi  4-> -X -K -fc M 4-) 4-J CD o w 4-1 CD CD M Cn CD CD •H •H rd -rH -r4 - H 4 > CO T5 4-> TS TI  o  o  fi  e  Xi  o  LH  CO  G o  CD CO  u  w  rH rd 4-1 O EH  G  D  •  CM  1  G  rd 4-J CO  CD Cn LO  LO  LO  LO  rd r H  > LO  LO  LO  rd  r^ r- r— VD VO vo  M  • • • CD Q O o o 4-J CO  4-1 -H  4-> <H CTi rH • H CO rH C O L O r-- rro ro o  •H 4-1 rd  fe  CO  •  •  ro  t-H o  u  dP  CO  •  VO  o o o  O o O o o <H LO VO r o CTl  •  rH rd  to  <H  LO  •r-l 4-> rd  > H  LO  LO  Xi O  CD  e •H  H  CTl rH CM  -H Ti *H rd Ti  LO  •  CM  •  CO  Xi  G O > CD to  Xi O rd to CD 4-1 o  G  CO r -  X! O  r o CTi rH CM  CD TJ PH  

Cite

Citation Scheme:

        

Citations by CSL (citeproc-js)

Usage Statistics

Share

Embed

Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                        
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            src="{[{embed.src}]}"
                            data-item="{[{embed.item}]}"
                            data-collection="{[{embed.collection}]}"
                            data-metadata="{[{embed.showMetadata}]}"
                            data-width="{[{embed.width}]}"
                            async >
                            </script>
                            </div>
                        
                    
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:
http://iiif.library.ubc.ca/presentation/dsp.831.1-0089140/manifest

Comment

Related Items