FEEDS AND FEEDING IN FRY AND FINGERLING CULTURE

J.J. van der Wind

R. & D. Department, Trouw & Co. N.V. International,
Putten, The Netherlands

ABSTRACT

The present problem of dry feed formulation for and feeding of fry and fingerlings are discussed against the background of successful use of artificial diets for salmonids fry and the use of live foods organisms for a number of other species.

Generally, rotifers (Brachionus plicatilis) and brine shrimp (Artemia salina) nauplii are adequate starter feeds. The composition and the size of these organisms is reviewed and compared with the known or assumed requirements of the fry. For most species, especially when they have small fry, no suitable dry feed has been developed yet, even when particle size and composition were technically identical to live food organisms.

Directions for further research are given.

RESUME

L'état actuel en ce qui concerne la composition et la technologie des aliments secs adaptés à l'élevage de poissons au stade larvaire est décrit en perspective de l'utilisation d'organismes vivants et l'application d'aliments secs à l'alevinage des salmonides.

En général les larves de poissons peuvent être élevées en utilisant des organismes vivants comme Brachionus plicatilis ou nauplies d'Artémia salina. La composition et les autres caractères de ces organismes sont discutés. Jusqu'à prêsent les essais à l'aide d'aliments secs n'ont pas connu beaucoup de succès, même si la composition et la granulation en étaient comparables à celles des organismes vivants.

Des suggestions sont faites pour des recherches plus approfondies.

1. INTRODUCTION

The subject of “feeds and feeding in fry and fingerling culture” is a broad one, because it is in fact a variety of problems superimposed on a variety of fish species. It is therefore very difficult to deal with all aspects in one review.

We are dealing in this Workshop only with fry and fingerling of fresh water fishes, so the salt water species can be excluded from the discussion. To a certain extent this facilitates the discussions, since some very intriguing and difficult questions can be omitted, i.e. the questions of the fresh water supply of salt water fry and fingerlings, the osmotic pressure and the related mineral requirements, the different behaviour of proteins in salt water, and so on. But, already the fresh water fry and fingerlings will give us enough problems and difficulties to discuss. In the programme of this Workshop already some species are mentioned and every one of these species has its own feeds and feeding problems. In my opinion it is important to find a best solution per species first and then try to find a greatest common divisor for all species later.

The latter is certainly a nice goal for many of us involved, but for the next year it seems to be unrealistic to expect that big solution.

Despite this general statement, it is not my intention to discuss in particular the 13 species discussed in this Workshop, partly in order not to infringe the experience papers, partly because it would bring us too far away from the general points which are also appropriate.

At the recently held “Symposium on Finfish Nutrition and Feed Technology” in Hamburg several papers were presented concerning the rearing of larvae of several species, including the feeds used and the feeding methods which were applied. Inevitably this paper will repeat to a certain extend some of those papers.

2. NUTRITIONAL REQUIREMENTS OF FRY AND FINGERLING

2.1 General

The available information on the nutritional requirements of fry and fingerling for most species is very limited. A complex of reasons is responsible for this situation, e.g.:

-   economically not important

-   organisms are too small to perform the normal experimental procedures

-   husbandry and raising techniques are not yet available or reliable

Only when the species have enough economic momentum and after the fish have reached a certain size, the nutritional requirements have been studied more intensively. In reality this means that quite some work has been done on Salmonidae, followed by the Cyprinidae.

The initial size of the fry is of primary importance for the size of the feed particle to be supplied. It is also known that the size of the fry after the yolk sac resorption is highly correlated with the size of the fish eggs (Nash, 1977). The following survey gives an indication of the variation in egg sizes (Nash, 1977, Huet, 1970):

The variation is enourmous and consequently there is a big variation in initial fry sizes. It can be noted that salmon and trout eggs are much bigger than e.g. carp eggs. This can be the basis for the fact that it is much easier to start with salmon or trout fry than with for instance carp fry.

It is also reasonalbe to assume that in nature nearly all fry start as plankton eaters, either first phytoplankton followed by zooplankton, or directly zooplankton. Only the very big fry will start immediately with bigger preys, which in turn have taken plankton as their feeding material. This means that in this primary food chain plankton plays an important role.

However, both the production and the composition of plankton in nature vary considerably and it is very difficult to use this as a reference for the nutritional requirements of fry as a control in feeding experiments. A good example of the differences in composition of some natural fish foods was given by Yurkowski and Tabachek (1979). Luckily it has been shown, that 2 organisms can be used as natural live food for a variety of species, in any case for most of the species of interest for us. It concerns the rotifer Brachionus plicatilis and the nauplii of the brine shrimp Artemia salina. Depending on the size of the fry they start with Brachionus followed by Artemia nauplii and later on by bigger preys or food particles or they start directly with Artemia nauplii. The rotifer measures around 150–200 micron and the Artemia nauplius around 200–500 micron. The importance of these sizes, relative to dry food, will be discussed later.

The fact that these two live foods are used successfully for a variety of fry species shows that they supply all the required nutrients in a sufficient quantity. Therefore we will use their compositions as guidelines for the discussions of the specific requirements. In several cases we will compare these figures with what is known from the Salmonidae.

2.2 Protein

The protein level of Brachionus plicatilis depends on their diet and age. Reported values on dry matter basis (d.m.) vary from 50–60%, with an average value of about 54% (Scott and Baynes, 1978, Yoshida and Hoshii, 1978). For the nauplii of Artemia salina a variation in the protein content is also reported, due to origin, feeding, age, etc. The reported values on d.m. vary even from 41.6% till 59.7% (Benijts, et.al., 1975). Taking into account the protein level on d.m. of the brine shrimp eggs (Utah and San Fransisco origin) and of the adult brine shrimps, which are reported to be about 50% (unpublished data) and 58% (Gallagher and Brown, 1975) respectively, we may assume that also nauplii of the Artemia salina will have on average a protein content of around 50–55% on dry matter. In fact these protein levels of both natural foods, being about 50–55%, are in very close agreement with the experience with trout and salmon fry, which show the best performance with artificial feeds containing 55–80% protein on d.m. (Cowey, 1979).

However, since protein is only N × 6.25 it is important to have a look at the amino acid compositions. Table I gives a survey of data of Brachionus, Artemia and Artemia-eggs, trout-eggs, egg albumin and of a fry starter. The composition of the rainbow trout eggs is given as this is in any case a major part of the first feeding of rainbow trout fry in their yolk sac stage. The amino acid composition of the nauplii of Artemia can be expected to be in between those of the eggs and the shrimp itself. One could argue from the figures in Table I that the Rotifera might be at least marginal in their supply of Arginine, Methionine, Lysine, Histidine and Tyrosine, although it is also possible that the supply in the other natural foods is excessive. On the other hand Scott and Baynes (1978 a) report about a nutritional deficiency of Rotifera for turbot fry. It would be wise to gather more information about the amino acid compositions of natural foods or food combination which have given a good fry growth, in order to be able to quantify the practically needed amino acid compositions.

Although the exact requirements for certain amino acids of fry of trout and salmon species are not well known, it seems that the existing trout or salmon starters which are on the market fulfil the requirements satisfactorily. However, a point which certainly needs more examination is the role of chitin in the feeding of fry and fingerlings. A part of the total nitrogen in some natural food is found in the chitin. An indication of the possible importance can already be taken from some publications of research with eels (Deelder, 1978).

Of course the presence of chitinases in the digestive tract is a necessity in that case. For fry and fingerlings our knowledge on the presence of proteinases and chitinases (and other enzymes as well) is far too limited.

2.3 Fat

Also the fat levels in the Rotifera vary according to their diet and age: from 5.4 to 17.1% on d.m. with an average of about 12% at the age of 2 days. Especially age seems to influence the fat level, in fact a decreasing effect. This might be the reason that Yoshida and Hoshii (1978) report only a value of about 7%. The lipid level in the nauplii of Artemia varies also. Reported levels are between 7.0% (6 day old nauplii) and 27%. It seems that the fat level in the nauplii is on average about 20% on d.m., whereas it is in the brine shrimp itself lowered till 6 or 7% on d.m. (aether extraction!). Anyway it seems that most of the natural food for fish, including fry, contains between 10 and 20% lipids on d.m. (see also Yurkowski and Tabachek, 1979).

The composition of the fat of the rotifera is not known. There is some information about the fatty acid composition of the fat of nauplii of Artemia and of the brine shrimp itself. These data are summarized in Table II together with information on the fatty acid composition of rainbow trout eggs, yolk sac fry of rainbow trout, brook trout eggs and fry and the average of some adult fresh water fishes.

As most authors report their fatty acid analysis in a different way, it is impossible to make a direct comparison. It is a pitty that so little is known about the fatty acid composition of Rotifera and nauplii. As a result it is not possible to judge their adequacy for developing fry. Typical is the absence of the important 22:6 (ω3) fatty acid in the nauplii and Artemia salina itself, but this might be a matter of analytic techniques. In the review of Castell (1979) it is shown that salt water species have an other ω6/ω3 ratio in their body than fresh water species have. This is undoubtedly influenced by the fatty acid composition of their natural food. Since Brachionus plicatilis and Artemia salina are salt water living organisms, so belong to the food chain of salt water species, it can be argued that their fat composition might not be optimal for fry of fresh water species. This needs however much more investigations. The fatty acid composition of the yolk fat, which is deposited by the parent fish, can also be taken into consideration.

2.4 Minerals

The ash content of the Brachionus is reported to be about 18.5 on d.m.. This level is certainly influenced by the salt content of the water, which adheres to the organisms (Yoshida and Hoshii, 1978). For nauplii the reported values vary from 6.5 to 14.7%, with an average of about 10%. Again the age seems to play a role (higher levels for older nauplii, Benijts, et.al., 1975). For the brine shrimp itself a value is reported of about 22% on d.m. (Gallagher and Brown, 1975). This again may be influenced by the salt water as the reported Sodium level was even 5.1%.

The composition of the minerals (on d.m.) is as follows:

The low Ca level and the high P level are interesting items to note and need further consideration, especially when they are compared with the Ca and P levels in for instance trout eggs and trout fry:

¹⁾ unpublished data
²⁾ Ogino and Kamizono (1975)

A considerable increase of Ca in the whole body of the developing fry is noticeable, not only in absolute terms, but also relative to phosphorus and magnesium. It is nog unrealistic to assume that rotifers and brine shrimp nauplii fed over a prolonged period will give deficiency symptoms for minerals, as in fact they are described by Ogino and Kamizono (1975) for developing rainbow trout fry and carp.

2.5 Carbohydrates

This includes both the so-called crude fibre and the nitrogen free extracts. Reported values for Brachionus plicatilis are around 13.8% on d.m. (Scott and Baynes, 1978, Yoshida and Hoshii, 1978). This applies for 2 days old Rotifera. With increasing age this level decreases. For the nauplii of Artemia salina two different values are reported of 22.7% and of 6.0% on d.m. respectively (Benijts, et.al. 1975). The reliability of these figures seems doubtful. For brine shrimp itself one value 3.6% on d.m. of crude fibre is reported. We do not know anything about the further composition of these fractions, and neither do we know anything about a possible nutritional role. It can be assumed that eventually only the simple sugars can be utilized as energy sources, whereas the polysaccharides and crude fibre are of no value.

2.6 Vitamines

We are not aware of publications concerning the vitamin levels in Rotifera or in nauplii of Artemia salina. There is one report of vitamin levels in brine shrimp itself (Gallagher and Brown, 1975), but to use these figures as a reference for their nauplii is not justifiable. We must assume that for the first development of the yolk sac fry and fry of fish species the vitamin levels in the eggs are of primary importance. In this connection the carry-over of the vitamins from the parent fish via the eggs plays a major role - and should not be forgotten in the lay-out of the formulations of feeds for the brood fish! An example of a specific effect of a vitamin-like substance on reproduction is given by Deufel (1965).

However, as soon as the fry or fingerling stage is reached, there certainly is a need for a supply of vitamins via the feed. The nutritional requirements for trout and salmon fry and fingerlings are quite well established (Halver, 1972). For the other fresh water species this certainly is not the case, but if we take for the time being the rainbow trout figures, including a possible Vitamin C requirement, it can be assumed to be sufficient.

2.7 Water

This is in a way maybe the most important nutrient. Fresh Rotifera contain about 90% water and for nauplii of Artemia values between 85 and 95% are reported. It is obvious that as soon as dry food is given to fry and fingerling the required water intake has to come from the environment.

It is clear that for bigger fry of all species this is not a problem as such, but with the exeption of the species with bigger fry (trout and salmon) it may pose a problem in the initial feeding phase of the species with small fry.

This will be discussed in detail in the following section.

3. HOW TO MEET THE REQUIREMENTS

We have already seen, that Rotifera and nauplii of brine shrimp are good starter feeds for most of the fry of fresh water species and that their compositions can be used as references, because they supply all the necessary required nutrients sufficiently (although maybe not always in optimal levels!). The question now arises if and how these live foods can be replaced by artificial feeds. Theoretically it is not too difficult to imitate those compositions on dry matter. It would mean to make a formula with:

55 – 60% protein

12 – 16% fat

12 – 15% ash

10 – 15% carbohydrates

Also it would not be too difficult to meet the specific compositions of the protein, fat and minerals as outlined in the first part of this review. However, the success of numerous attempts to replace the live foods is very limited. In fact only the attempts with Salmonidae have been successful up till now. Work with carp, grass-carp, grayling, bream, pike and several salt water species have shown that a replacement of Rotifera and/or nauplii of Artemia salina very often resulted in poor performances (Huisman, 1979; von Luckowicz, 1976; Horvath, 1979; unpublished own data; Dabrowski et al., 1978; Gatesoupe, et al., 1977).

The reason for these failures up till now, are not absolutely clear as it remains very difficult to prove the nutritional adequacy of a feed if the feeding technique and husbandry is not optimal. In the initial phase a lot of attention was given to the problem of the water stability of starter feeds and to the particle sizes. It was thought that the particle sizes in use for Salmonidae were too big for fry of other species. The smallest size in use for trout and salmon are around 0.3–0.4 mm. The problems of the manufacturing of such small particles are already numerous. The number of dry food particles taken per day by fry of rainbow trout may be as low as 5 per day. This makes it imperative that all individual particles must be of the same, complete composition. In fact each particle should contain the required levels and compositions of protein, fat, minerals, vitamins. For the production of these particles it is necessary that certain ingredients are ground very fine and that special grades of ingredients are used, for instance from minerals, trace elements, vitamins, etc. A survey of the problems encountered was given for instance by Van Limborgh (1979).

For fry smaller than those of trout and salmon the problems to get the right particle size and composition are even bigger. In accordance with the sizes of Rotifera and nauplii it was thought that particle size have to be as small as 0.1–0.3 mm (100–300 u).

With special equipment it is nowadays indeed possible to produce such small particle sizes, even meeting the requirement of equal composition per particle. Theoretically however, it is possible that some of the nutrients (amino acids) become unavailable through the process (heat treatments in extrusion, flaking etc.) or are destroyed (vitamins for instance). The number of particles per gram of those feeds depends on the composition, but an example is the following:

Compared with rotifers and nauplii the particle sizes achieved are not too different, and also the weight per rotifer or nauplii is not too different from the mentioned particle weights. The wet weight of Brachionus plicatilis is said to be around 0.0025 mg and of nauplii of Artemia salina between 0.013 (initial) and 0.045 mg (after 2 days). Regarding the water stability it can be said that the newer production techniques enable us to produce good water-stable particles. Details and surveys can be found in the publications of Meyers (1979), Luquet and Rumsey (1979) and of Van Limborgh (1979). An indication of the water stability is for instance the percentage of Nitrogen, dissolved in water after a certain time. From 2 commercial available starters this was measured:

It can be concluded that there are differences between products, but it is possible to bring the losses of the nitrogen into the water down to 5–10%. So, it can be concluded that particle sizes, particle weight and water stability of artificial starter feeds are comparable with rotifers and nauplii. To answer the question why the results are often so dissappointing, it is therefore needed to realize that there are other differences between live food and artificial feeds:

The motion of live food remains an important point, especially for those species which react only to moving preys (pike f.i.). The recognition of the food as such by fry is and will be very difficult to build-in in artificial feeds. The possibility to deform is a big advantage of natural food during the actual uptake by the fry. It enables the fry to ingest bigger particles than they could swallow with fixed particle forms. For the latter the size and form of the mouth is decisive (Thorpe and Wankowski, 1979). The water content of the feed particles is a factor which needs much more attention. It is said repeatedly that one of the main problems in the use of artificial starter feeds is the water pollution. In all the systems one tries to supply the same density of food particles per liter as with natural food. As sizes and weights are about comparable, it is well possible to achieve the required density. However, the total dry matter supply at the same time is at least 10 times higher. Or with other words: for the same dry matter uptake the larvae need only to swallow 10% of the particles taken normally with live food. This means, even if we assume the same attractability (which also remains to be seen!), that one supplies with artificial feeds at least 10 times more dry matter. This must inevitably lead to water pollution problems. A lot of research, also in the techniques is needed, to solve these problems.

It is of importance to note two other points. Firstly the fact that in certain publications the use of lyophilized natural food has not given the required performances (von Luckowcz, 1976). This might stress the importance of the water level per se in the natural food as a required nutrient, but it could also be a proof for the idea that the enzymes present in live food are a necessity for the digestion by the small fry (and they are destroyed in the drying process!). On the other hand this has not been shown to be important in the Salmonidae.

Secondly there is the work of Appelbaum (1979) reporting very positive results with Corigonidae fed with alkan-grown-yeasts. It might be of interest to follow this work, as maybe yeast cells, brought into water, exert an other behaviour as water stable prepared starter feeds.

4. FEEDING METHODS

In the previous section the feeding methods were already indirectly mentioned as being very important. It is not well possible to give general recommendations. The feeding method will have to be adapted to the feeding behaviour of the fry of the specific species, since it is very difficult to adapt the feeding behaviour of the fry to the feeding method. This would certainly give a very low survival rate, which is not the intention in a mass rearing system.

If the feed particles meet the nutritional and physical requirements posed by the fry it will be technically no too difficult to bring them on the spot. The feeding technique however is also closely related to the total husbandry and only in a combined effort with all disciplines involved, will it be possible to find the most optimal solution.

5. REFERENCES

Appelbaum, S., 1979 The suitability of alkan-yeast (hydrocarbon grown yeast) as a first nutrient for Coregonus albula (L) fry. In: Halver, J.E. and K. Tiews (Ed.): Finfish Nutrition and Fishfeed Technology 1: 515 – 524. Heenemann GmbH & Co, Berlin.

Atchison, G.J., 1975 Fatty acids levels in developing brook trout (Salvelinus fontinalis) eggs and fry. J. Fish. Res. Board Can. 32: 2513 – 2515

Barahona-Fernandes, M.H. and M. Girin, 1977 Effect of different food levels on the growth and survival of laboratory-reared seabass larvae (Dicentrarchus Labrax (L)). Actes de Colloques du C.N.E.X.O., 4: 69 –84.

Benijts, F., E. van Voorden and P. Sorgeloos, 1975 Changes in the biochemical composition of the early larval stages of the brine shrimp, Artemia salina L. 10th European Symposium on Marine Biology, Vol. 1: 1–9.

Castell, J.D., 1979 Review of lipid requirements of fishes. In: Halver, J.E. and K. Tiews (Ed.): Finfish Nutrition and Fishfeed Technology 1: 59 – 84. Heenemann GmbH & Co, Berlin.

Cowey, C.B., 1979 Protein and animo acid requirements of finfish. In: Halver, J.E. and K. Tiews (Ed.): Finfish Nutrition and Fishfeed Technology 1: 3 – 16. Heenemann GmbH & Co, Berlin.

Dabrowski, K., H. Dabrowski and C. Grudniewski, 1978 A study of the feeding of common carp larvae with artificial food. Aquaculture, 13: 257 – 264.

Deelder, C.L., 1978 A short note on the intensive culture of eel (Anguilla anguilla L.). Aquaculture, 13: 289 – 290.

Deufel, J., 1965 Pigmentierungsversuche mit Cantaxanthin bei Regenbogenforellen. Arch. Fischereiwiss., 16: 125 – 132.

Gallagher, M., and W.D. Brown, 1975 Composition of San Fransisco Bay brine shrimp (Artemia salina). J. Agric. Food Chem., 23 (4): 630 – 632.

Gatesoupe, F.J., M. Girin and P. Luquet, 1977 Recherche d'une alimentation artificielle adaptée à l'élevage des stades larvaires des poissons. II - Application à l'élevage larvaire du bar et de la sole. Actes de colloques du C.N.E.X.O., 4: 59 – 66.

Halver, J.E., 1972 Fish Nutrition, New York, Academic Press. 713 pp.

Horvath, L., 1979 The rearing of warmwater fish larvae. In: Halver, J.E. and K. Tiews (Ed.): Finfish Nutrition and Fishfeed Technology I: 349 – 357. Heenemann GmbH & Co, Berlin.

Huet, M., 1970 Traité de pisciculture. Bruxelles, Ch. de Wijngaert. 718 pp.

Huisman, E.A., 1979 The culture of grass carp (Ctenopharyngodon idella Val.) under artificial conditions. In: Halver, J.E. and K. Tiews (Ed.): Finfish Nutrition and Fishfeed Technology I: 491 – 500. Heenemann GmbH & Co, Berlin.

Limborgh, C.L. van, 1979 Industrial production of ready to use feeds for mass rearing of fish larvae. In: Halver, J.E. and K. Tiews (Ed.): Finfish Nutrition and Fishfeed Technology II: 3 – 11. Heenemann GmbH & Co, Berlin.

Luckowicz, M. von, 1976 Anfütterung von Karpfenbrut. Fischer und Teichwirt, 27: 68 – 69.

Luquet, P. and G.L. Rumsley, 1979 Formulation et technologie des aliments secs pour poissons. In: Halver, J.E. and K. Tiews (Ed.): Finfish Nutrition and Fishfeed Technology II: 21 –35. Heenemann GmbH & Co, Berlin.

Meyers, S.P., 1979 Formulation of water-stable diets for larval fishes. In: Halver, J.E. and K. Tiews (Ed.): Finfish Nutrition and Fishfeed Technology II: 13 – 20. Heenemann GmbH & Co, Berlin.

Nash, C.E., 1977 The breeding and cultivation of marine fish species for mariculture. Actes de Colloques du C.N.E.X.O., 4: 1 – 10.

Ogino, C. and M. Kamizono, 1975 Mineral requirements in fish. 1. Effects of dietary salt-mixture levels on growth, mortality and body composition in rainbow trout and carp. Bull. Jap. Soc. Sci. Fish., 41: 429 – 434.

Scott, A.P. and S.M. Baynes, 1978 Effect of algal diet and temperature on the biochemical composition of the rotifer, Brachionus plicatilis. Aquaculture, 14: 247 – 260.

Scott, A.P. and S.M. Baynes, 1979 The effect of unicellular algae on survival and growth of turbot larvae (Scophtalmus maximus L.). In: Halver, J.E. and K. Tiews (Ed.): Finfish Nutrition and Fishfeed Technology I: 423 – 433. Heenemann GmbH & Co, Berlin.

Thorpe, J.E. and J.W.J. Wankowski, 1979 Feed presentation and food particle size for juvenile atlantic salmon, Salmo salar L. In: Halver, J.E. and K. Tiews (Ed.): Finfish Nutrition and Fishfeed Technology I: 501 – 513. Heenemann GmbH & Co, Berlin.

Yoshida, M. and H. Hoshii, 1978 Nutritive value of rotifera, a zoo-plankton, for poultry feed. Japan. Poultry Sci., 15 (2): 64 – 68.

Yurkowski, M. and J. L. Tabachek, 1979 Proximate and amino acid composition of some natural fisch foods. In: Halver, J.E. and K. Tiews (Ed.): Finfish Nutrition and Fishfeed Technology I: 435 – 448. Heenemann GmbH & Co, Berlin.

6. TABLES

Table I: Amino-acid compositions from Rotifera, Artemia, trout eggs and eff albumin and a fry starter (expressed as % of protein)

1) Yoshida and Hoshii (1978)

2) unpublished own data

3) Gallagher and Brown (1975)

Table II: Summary of fatty acid compositions.

1) Benijts et al. (1975);

2) Gallagher and Brown (1975);

3) unpublished own data (RBT = rainbow trout);

4) Atchison (1975); eggs 10 days after fertilization, fry 70 days after fertilization;

5) Castell (1979).