The present invention relates to a fish feed, to a method for its manufacture, to the use of said feed in fish feeding, and to a method for feeding fish. The fish feed according to the invention provides a higher production of roe, better growth of fish and a smaller gutting loss compared to the feeds of the prior art.

Background of the invention In Finland, Sweden and also in some parts of Denmark the production of rainbow trout is economically very dependent on the simultaneous production of roe. The Norwegian salmon industry on the other hand does not focus on roe production as the fish flesh quality is considered to be lower when the fish get sexually mature. Therefore, the development of feeds for salmon and trout has been focused on the growth of fish which do not become sexually mature .

The amount of roe in a fish differs and is dependent e.g. on how close to spawning the fish is slaughtered. Normally the GSI (gonadosomatic index; roe% by weight of a live fish) of rainbow trout is 10-14% in December- January at Finland's conditions.

The nutrients required by fish for growth, reproduction and other physiological functions are similar to those of other animals, i.e. protein, minerals, vitamins, and lipids. Ingredients used in commercial fish feeds can therefore be classified as protein (amino acid) sources, lipid sources, carbohydrates and vitamin and mineral sources.

Fish feeds are available in different compositions and pellet sizes designed for fish of various species at different growth periods. A fish fry needs a diet rich in vitamins and protein for its growth. As fish increases in size the energy level in feed rises in order to enable effective growth. Since high quality protein is more expensive than fat, high energy content is usually provided by increasing the fat content of the feed. The conventional extruded feed used for trout larger than 350-500g in their last growth period contains approximately 36% protein and 35% fat by weight of feed. It has now been found that it is possible to remarkably increase roe production for human consumption and/or the growth of fish by feeding the fish during vitellogenesis, i.e. during the period when the fish is developing and producing roe in its body cavity, with a feed having a higher amount of protein than the conventional feed. Roe production and fish flesh yield are increased also when the fish is fed with the high protein feed only during part of the vitellogenesis, e.g. during early (endogenic) vitellogenesis or, alternatively, only during late (exogenic) vitellogenesis, but then the increases in fish growth and roe production are smaller than those observed when the high protein feed is provided during the whole period of vitellogenesis.

Summary of the invention

An object of the present invention is a method for feeding fish for human consumption wherein the method comprises the steps of providing a feed composition to the fish during vitellogenesis, i.e. during the period when the fish is developing and producing roe in its body, wherein the feed comprises carbohydrates and a high amount, preferably 38-48% protein, and 25-34% fat.

A further object of the invention is the use of a fish feed comprising a high amount of protein for increasing roe production and/or the growth of fish by feeding the fish with said fish feed during vitellogenesis, i.e. during the period when the fish is developing and producing roe in its body. More specifically, the fish feed for the use according to the invention comprises carbohydrates and 38-48% protein and 25-34% fat. The invention also relates to a fish feed comprising carbohydrates and 38-48%, preferably 39-45%, even more preferably 41-42% protein and 25-34%, preferably 28-33%, and more preferably 30-33% fat, together with vitamins and additives used in fish feeds.

Another object of the invention is a process for preparing a fish feed which comprises the steps of mixing the feed raw materials except part of the fat, extruding the obtained mixture and then vacuum coating the obtained extrudate to obtain pellets of suitable size, wherein the feed comprises carbohydrates and 38-48%, preferably 39-45%, protein and 25-34%, preferably 28-33%, fat and wherein most of the fat is added in the step of vacuum coating the pellets. Detailed description of the invention

The feed according to the invention is particularly suitable for fish used for roe production, such as rainbow trout, salmon, brown trout and whitefish, especially rainbow trout and whitefish, during the period when the fish is starting to develop and is producing roe in its body cavity, i.e. during vitellogenesis.

The feed according to the invention comprises carbohydrates and 38-48%, preferably 39- 45%, more preferably 40-44 %, and even more preferably 41-42% of protein, while the content of fat is 25-34%, preferably 28-33%, and more preferably 30-33%. The protein content can be tailored depending on the particular fish species to be fed. For rainbow trout the most preferred amount of protein is 41-42%. For whitefish, a preferred amount of protein for the use according to the invention is 44-45%, while the amount of fat is 25- 26%.

An important protein source in the feed according to the invention is fish meal. However, the amount of fish meal does not need to be as high as conventionally used in a feed for brood fish which means cost savings to the fish farmer. Other protein sources include for example soy, soy protein, wheat, wheat gluten, corn, corn gluten (i.e. corn protein), faba bean, lupine, pea protein, rapeseed/canola, potato protein, cottonseed meal, animal protein sources, such as blood meal, hemoglobin meal and other animal proteins, such as poultry and insect proteins. Sources of fat include for example vegetable oils, such as rapeseed oil, palm oil, soybean oil, olive oil, linseed oil and cottonseed oil, fish oil and animal fats. Preferred sources of fat are rapeseed oil and fish oil. By adjusting the proportions of rapeseed oil and fish oil in the fish feeds during the growth of the fish it is possible to obtain optimal levels of the healthy fatty acids eicosapentanoic acid (EPA) and docosahexaenoic acid (DHA) in the fish flesh.

The feed according to the invention may also comprise added nucleotides and/or nucleosides, especially in free form, to further improve the yield of roe. Nucleotides and nucleosides which are building blocks of DNA and RNA and of ATP accelerate cell division. Even a minor amount of added nucleotides/nucleosides has an effect but preferably the amount of added nucleotides/nucleosides in the feed should be at least 0.01%, more preferably at least 0.05%, even more preferably at least 0.1% and further preferably at least 0.5% or more by weight of the feed. The upper limit of added nucleotides is not critical since nucleotides are not harmful to the fish even in high amounts. Especially when the feed contains under 30% fish meal it is beneficial to add nucleotides/nucleosides to the feed, and particularly under 20% of fish meal, nucleotides/ nucleosides in some form should be added to the feed to further guarantee an increased roe production. It is known that the diet of fish needs to contain the essential amino acids arginine, histidine, isoleucine, leucine, lysine, methionine, phenylalanine, threonine, tryptophan and valine since fish cannot synthesize these amino acids from other molecules. Some amino acids, such as cysteine and tyrosine, are non-essential amino acids since cysteine can be synthesized from methionine and tyrosine from phenylalanine. In order to guarantee a sufficient level of the essential amino acids in the fish diet, for example methionine and lysine have been added to fish feeds to provide an adequate intake of these amino acids.

It has now been found that the roe production of fish is further improved by including in the diet higher amounts of added amino acids. More specifically, especially amino acids tryptophan, methionine and cysteine/cystine have been found to have a positive effect on roe production. Therefore, the feed according to the invention preferably contains at least one of added amino acids from the group of methionine, tryptophan, cysteine, cystine and lysine in higher amounts than recommended by NCR (National Research Council) in 2011 or in higher amounts as compared to the other amino acids.

Preferably in the feed according to the invention at least one of the amino acids tryptophan, methionine and lysine is added in such amounts that the feed contains said amino acid(s) in higher amounts than recommended by NRC in 2011. A preferred added amino acid is methionine. More preferably in the feed according to the invention at least two of the amino acids tryptophan, methionine and lysine are added in such amounts that the feed contains said amino acids in higher amounts than recommended by NRC in 2011. Even more preferably in the feed according to the invention the amino acids tryptophan, methionine and lysine are added in such amounts that the feed contains said amino acids in higher amounts than recommended by NRC in 2011. Other amino acid(s) in addition to amino acids tryptophan, methionine and/or lysine may also be added to the feed according to the invention in amounts higher than recommended by NRC.

Based on the recommendation of NRC (2011), rainbow trout requires 0.3% tryptophan, 0.7% methionine, 1.1% of methionine + cysteine, and 2.4% lysine (% by weight of the feed calculated as digestible amino acids). In the feed and method according to the invention, methionine is added in such amounts that the feed contains digestible methionine over 0.7%, preferably over 1.0%, more preferably over 1.4%, and even more preferably over 2.1% by weight of the feed. Tryptophan is added in such amounts that the feed contains over 0.3%, preferably over 0.4%, more preferably over 0.6% and even more preferably over 1.0% digestible tryptophan. The amount of digestible lysine in the feed according to the invention is over 2.4%, preferably over 2.8%, more preferably over 3.2%, and even more preferably over 7% by weight of feed.

The fish feed according to the invention comprises also vitamins needed for the growth and roe production of the fish. The level of vitamins in the feed is adjusted by adding to the feed a vitamin premixture which comprises all the vitamins the fish needs. The vitamins may include any vitamins such as vitamin C, vitamin E (tocopherol), vitamin B, vitamin A, vitamin D and vitamin K.

Vitamins C and E provide together a strong anti- oxidative effect. Vitamin E includes various tocopherols (α, γ and δ tocopherol) which protect polyunsaturated fatty acids, such as eicosapentanoic acid (EPA) and docosahexaenoic acid (DHA), from oxidation. In addition to its own antioxidative effect vitamin C also re-activates the function of vitamin E.

Vitamin A refers to a group of compounds which have the same biological activity. Major forms of vitamin A include for example retinol, retinoic acid and carotenoids. Carotenoids are pro-vitamin A forms, among which beta-carotene is most important. Astaxanthin which is a common additive in fish feeds also belongs to carotenoids. Astaxanthin provides the red color of the salmon flesh and has also an antioxidative effect. The amounts of vitamins and astaxanthin in a fish feed are known to a person skilled in the art and can be adjusted accordingly. For example, in the feed according to the invention the amount of vitamins may be lower than that in a feed designed for brood fish or fry. The feed according to the invention may also comprise beta-glucan for enhancing phagocytic activity and thus for activating the immune system. If beta-glucan is used, its amount is within the limits known to a person skilled in the art of fish feeds.

Other possible additives in the feed according to the invention are phytase enzyme and other beneficial enzymes. As phosphorus in plant materials is stored in the form of phytic acid, phytase is needed in the gastrointestinal tract to degrade phytic acid and release phosphorus. The natural phytase activity in the gastrointestinal tract of fish is almost negligible and if vegetable based sources of phosphorus are used in fish feeds, added phytase enables to release phosphorus from phytic acid. The amount of added phytase may be for example approximately 750 FYT/kg feed.

The feed according to the invention preferably also comprises lecithin. In fish feeds, lecithin is used principally to emulsify fat, thus improving digestion and feed efficiency, but lecithin has also a nutritional role as a source of phospholipids, notably such as phosphatidylcholine, phosphatidylinositol and phosphatidylethanolamine. Lecithin functions also as an antioxidant to prolong the shelf life of the feed.

Fish feeds are available in different pellet sizes and densities to accommodate the needs of species in various growth stages. The pellet size should in general be approximately 20- 30% of the size of the fish species mouth gape. In practice, for rainbow trout the pellet size varies from 1.7 mm pellets for small fry to 9 or even 12 mm pellets for the last growth period. For the use according to the invention the preferred pellet size is 7 to 12 mm, preferably 7 to 9 mm in case of rainbow trout. A fish can eat a smaller pellet size than recommended for a certain growth stage and therefore the actual pellet size is not decisive.

By the method according to the invention it is possible to get a much higher flesh yield especially in fish that produce roe. When the test feed according to the invention and a control feed (conventional feed with 36% protein and 35% fat) were fed for about 6 months to fish that should get sexually mature that year, it was found that the test feed had a better feed conversion rate (FCR) than the control feed. The test feed also provided a higher flesh yield and a much higher roe which means a remarkable increase in the absolute amount of roe. A reduced gutting loss was also observed, also when the test feed was fed only during the last 4.5 months of vitellogenesis.

In the method according to the invention the time of starting the feeding with the new feed depends on the age and species of the fish involved as well as on the temperature of farming water, the amount of light etc., i.e. on factors which have an effect on the timing when the fish is producing roe in its body. The reproductive cycle of a rainbow trout spawning for the first time starts approximately 12 months before the actual spawning time. The oocyte maturation and yolk incorporation (vitellogenesis) is a hormonally controlled and regulated process wherein vitellogenin produced by the liver is stored in the maturing egg cells. Vitellogenesis comprises various phases starting from previtello- genesis, followed by endogenic and exogenic vitellogenesis. A rainbow trout enters previtellogenesis at Finland's conditions during spring (April-May), followed by endogenic vitellogenesis, and finally exogenic vitellogenesis until spawning.

In the method according to the invention a feed composition is provided to the fish during vitellogenesis. The period of vitellogenesis in different fish species varies and is dependent also on light and temperature conditions. A preferred time period of feeding rainbow trout according to the method of the invention is at least about 3 months, more preferably 3-4 months, even more preferably about 5 months, 5-6 months, 6-7 months, 7-8 months, 8-9 months, 9-10 months or over 10 months. A more preferred time period for feeding rainbow trout according to the invention is 6-8 months at Finland's conditions. Even shorter periods of time (1-2 months, typically at least 1 month, more typically at least 2 months, most typically at least 3 months) have an effect on roe production and/or flesh yield, especially on roe production, of salmonids and are within the scope of the invention.

In one embodiment of the invention the fish is fed according to the invention for 3-4 months during early vitellogenesis. In a more preferred embodiment of the invention the fish is fed according to the invention for 3-4 months during late vitellogenesis. In still another embodiment of the invention the fish is fed according to the invention during part of vitellogenesis or during any combination of time periods during vitellogenesis. In a preferred embodiment the fish is fed according to the invention during whole

vitellogenesis.

The invention is also directed to the use of a fish feed comprising a high amount of protein for increasing roe production for human consumption and/or the growth of fish by feeding the fish with said fish feed during vitellogenesis, i.e. during the period when the fish is developing and producing roe in its body. More specifically, the fish feed for the use according to the invention comprises carbohydrates and 38-48% protein and 25-34% fat. Preferred feed compositions are as disclosed in this specification. Both the produced roe and the fish flesh are aimed for human consumption.

In this disclosure "%" means percent by weight if not otherwise specified.

The following experimental part illustrates the invention. Example 1

The tested fish was rainbow trout that should get sexually mature soon and had an averag starting weight of 0.88 kg. The average end weight when slaughtered was 2.85 kg. The feeds comprised conventional fish feed ingredients with the following differences in the compositions of the feeds:

The control feed (%): Protein 36.0

Fat 35.0

The test feed (% ) : Protein 41.0

Fat 31.0

Both feeds had the same amount of digestible energy. The fish (1328 pes) were divided into 8 cages (166 fish/cage). Four cages were fed with the control feed (Hercules LP 7 mm) and four cages with the test feed, starting late May until early September. Due to high water temperature some mortality occurred and fish had to be starved for a prolonged period during summer. In September each cage was divided into two cages, in order to be able to change the feeding of the subgroups. The groups were marked as follows:

AA - received control feed only

AB - received control feed at the beginning of the test (approximately 3 months) and the test feed during the end of the test period (during part of the exogenic vitellogenesis, i.e. approximately 3 months)

BB - received test feed during the whole test period (approximately 6 months)

BA - received test feed at the beginning of the test (approximately during endogenic vitellogenesis, 3 months) and the control feed during the end of the test period

(approximately 3 months)

Feeding was otherwise identical for each cage. The fish were fed according to the recommendations of the feed producer for the commercial feed (Raisioagro Oy' s feeding table for rainbow trout, 2014), based on the temperature of the farming water and the average weight in each cage. Samples were taken and the fish were weighed at the beginning of the test, in early September, and in early December.

In December 30 fish from each cage were weighed, stunned, bled and their gutted yield, and maturity were recorded. The gutting loss was determined as the percentage of gut weight with respect to live and gutted weight of a fish. The gonads were weighed in December. GSI values were determined as the percentage of gonad weight with respect to live as well as gutted weight of a fish. Results

The average FCR (feed conversion rate) and SGR (specific growth rate) values for each feeding group are presented in Table 1. SGR values illustrate the growth % of fish per day, based on live weight. FCR values show the amount of feed needed to produce 1 kg of fish. The smaller the FCR value, the more efficiently the feed is converted into the growth of fish. Relatively high FCR upon the first test period results plausibly from suboptimal farming conditions during hot summer. From the results it can be seen that the group receiving only the test feed remarkably increased its growth when compared to the group receiving only the control feed throughout the experiment. Also the groups receiving the test feed only for part of the test period showed better growth and growth rate than the control group.

Table 1. The average FCR and SGR for each treatment group

TreatFCR 1 FCR II FCR l-ll SGR I SGR II SGR l-ll ment May-Sept Sept-Dec May-Dec May-Sept Sept-Dec May-Dec AA 1.35 0.98 1.15 1.02 0.55 0.78

AB 1.35 0.96 1.15 1.02 0.56 0.78

BA 1.31 0.95 1.13 1.02 0.58 0.80

BB 1.31 0.93 1.11 1.02 0.61 0.82

The gutted fish weight, the gut weight, the gonad weight, and GSI are presented in Table 2.

As can be seen from Table 2, the gutted weight of the test group (BB) was remarkably higher than in the control group (AA) and also higher than in groups AB and BA.

Table 2. Live weight, gutted weight, gonad weight, gut weight, GSI and gut in mature fish in December

Treatment Live g Gutted g Gonads g Gut g GSI1 % GSI2 % Gut %

AA 2754 2128 310.4 224.6 11.3 14.8 10.6

AB 2787 2169 315.6 214.5 11.3 14.6 10.0

BA 2862 2228 313.6 230.5 11.0 14.3 10.4

BB 2992 2313 353.7 230.1 11.8 15.4 9.9

The absolute amount of gonads showed a surprisingly high increase in all three test groups and especially for group BB.

The results also show a surprisingly reduced gutting loss (i.e. weight of gut) for the group fed the test feed.

Table 2 further shows the gonadosomatic index (GSI) ( ) based on live fish(l) and gutted fish(2) weight. The results show an increase in the gonadosomatic index in fish fed with the test feed as compared to the fish fed only the control feed. As already mentioned, fish fed the test feed also had a higher weight than those fed the control feed. Therefore, even though the fish were bigger the GSI values were unexpectedly even higher for group BB than for the control group AA.

Example 2

The tested fish were rainbow trout which were grown at a commercial fish farm. One cage was fed with the test feed and another with the control feed. The test started in May. The compositions of the feeds were as in Example 1. Daily feeding rates used in the test followed the recommendations given in feeding tables (Raisioagro, 2014). The fish were slaughtered in December, i.e. during the time of roe production. About 100 fish from each cage were taken as samples.

FCR (feed conversion rate) values were calculated for the whole group of fish. FCR of the control group was 1.37 while FCR of the test group was 1.28. The FCR of the test group was thus remarkably better than that of the control group.

Other results are shown in Table 3.

Table 3. Results at the end of the test, calculated from those fish that were sexually mature

CONTROL TEST FEED

Average live weight, g 3063 3160

Average gutted weight, g 2377 2479

Gut% (gut/live weight) 7.09 6.10

GSI1% (gonads/live weight) 11.43 12.20

GSI2% (gonads/gutted weight) 14.78 15.63

In addition to the remarkably better FCR values, it can be seen from the results that also the roe production (GSI) of the test group was surprisingly higher in the test group compared to the control group. The results also show a surprisingly reduced gutting loss for the test group.