- drafted by C.Y. Cho (Canada)
| 3.1.1 | Since apparent digestibility (AD) can be measured with accuracy, the AD of feed ingredients should be determined as part of the routine quality control process. |
| 3.1.2 | For faecal collection the settling or rotating screen principle as used in the “Guelph” or “St Pee” system may be recommended. However, the hand “Stripping” technique should be used with great care. |
| 3.1.3 | Metabolizability and net energy are recommended for the evaluation of complete diets as a measure of productivity. |
| 3.1.4 | Near “normal” retention of both nitrogen and energy is crucial to determine retained energy (RE), net energy (NE), metabolizable energy (ME) and digestible energy (DE). |
| 3.1.5 | Open-circuit indirect calorimetry is recommended for the measurement of heat production. |
| 3.1.6 | Feeding standards must be based on nutritional energetics. |
- drafted by C.B.Cowey (Canada)
| 3.2.1 | A minimum of six dietary nutrient levels or treatments is recommended for nutrient requirement studies |
| 3.2.2 | Experimental diets should be isoenergetic on a digestible energy or metabolizable energy basis; with DE preferred over ME because values are more reliable and are additive. |
| 3.2.3 | Carcass analysis should be carried out at the beginning and end of the experiment. |
| 3.2.4 | At least triplicate tanks of fish should be used per dietary treatment, as one tank of fish represents a single block observation. |
| 3.2.5 | An approved statistical analysis should be used to define the nutrient requirement level (ie. Broken line, polynomial regression). |
| 3.2.6 | Nutrient requirement studies should be of sufficient duration that there is a 5 to 10 fold increase in fish weight. |
| 3.2.7 | The growth response of experimental fish should be monitored using a model such as specific growth rate or more preferably on the basis of the cube root of fish weight. Furthermore, nutrient requirement should be determined at the maximal possible rates of growth. |
| 3.2.8 | Fish should not be fed on a restricted feeding regime but should be fed to satiation several times per day. |
| 3.2.9 | Amino acid requirements should be measured by using the classic dose-response method. |
| 3.2.10 | Apparent digestibility should be determined in free-swimming fish using an appropriate method by which faeces are naturally voided, and avoiding methods which rely on the handling and removal of fish from the water. |
-drafted by M. Jobling (Norway)
| 3.3.1 | Include “digestibility” estimations as a routine procedure in “growth trials” rather than relying on “previously published or perceived values” from other studies or fish species. |
| 3.3.2 | Since quantitative estimates of feed intake will rarely be available and total collection of faeces will seldom be possible reliance must be placed on the use of an “indirect method” incorporating an “indigestible marker”. |
| 3.3.3 | Select markers with care and be aware of pitfalls in using potentially degradable markers such as “mineral ash”, “cellulose” and “fibre”. |
| 3.3.4 | Include a “marker” routinely in diets such that faecal collection can be carried out at several times during the course of the trial. Do not make “special batches” of “marked feed” for digestibility trials conducted in parallel to the growth trial on other groups of fish, or for estimations of digestibility conducted after the completion or towards the end of the growth trial. |
| 3.3.5 | Search for alternative suitable internal markers such as n-alkanes and 5α-cholestane for measuring lipid digestibility, as acid-insoluble ash (AIA) may be present at too low a concentration within feed ingredients. |
- drafted by J.R. Sargent (UK)
| 3.4.1 | Always perform detailed fatty acid analyses on feed oil samples used as it is important to specify all (n-3) and (n-6) fatty acids even if they are present in trace amounts; these traces can often be very important (ie. presence of 20:4(n-6) in fish oils). |
| 3.4.2. | Assume (with some certainty) that all species require both (n-3) and (n-6) essential fatty acids (EFA). |
| 3.4.3 | Determine whether the species can elongate/desaturate C:18 polyunsaturated fatty acids (PUFA) to C:20 and C:22 PUFA. |
| 3.4.4 | For species that can elongate/desaturate take care with the (n-6)/(n-3) ratio, and avoid excessively high values (ie. (n-6/n-3) ratios of 10/ or 5/1). It would be preferable not to have (n-6/n-3) higher than 1, or preferably less. |
| 3.4.5 | For species that can not elongate/desaturate and therefore require polyunsaturated fatty acids (ie. fish oil), do not neglect (n-6) content (ie. 20:4(n-6) which may be critical, especially in endovascular function. A “guesstimate” could be a (n-3)/(n-6 ratio of > 10:1, although 5:1 would probably be acceptable although difficult to achieve practically. Care should also be taken with fish oils having 20:5 > 22:6 (ie. many sardine, pilchard and anchovie oils and all fish oil concentrates), as too high 20:5(n-3) may interfere (depress) eicosanoid production from 20:4(n-6). |
| 3.4.6 | The concern in paragraphs 4 and 5 is about a “desirable” (n-6)/(n-3) ratio for eicsanoid production, as this is involved in stress reactions and also in immune systems; obvious consequences for overcrowding and disease. |
| 3.4.7 | Remember the important role of 22:6(n-3) in renal, visceral, and perirenal development. In particular, the requirement for 22:6(n-3) during the early larval developmental stages is likely to be critical. |
| 3.4.8 | Considerable further research is needed to determine the optimal (n-6)/(n-3) ratio for fish. |
| 3.4.9 | Remember that EFA can be retained directly by animals, and so at least a complete generation may be required to elicit pathologies or abnormalities in fish. |
- drafted by E.Pfeffer (Germany)
| 3.5.1 | The contribution of carbohydrates to the DE supply of salmonids is to be expected only from α-linked glucose polymers. |
| 3.5.2 | Starch should be gelatinized before use and the degree of gelatinization estimated simply and reliably by amylo-glucosidase degradability. |
| 3.5.3 | Even with highly degradable starch the amount digested is generally negatively correlated with the amount ingested. |
| 3.5.4 | As a consequence, digestibility of any source of starch should be determined at least at two dietary levels, the slope in change of digestibility being an important piece of information for the feed compounder. |
| 3.5.5 | As the digestibility of starch decreases with increasing dietary concentration, the contribution of carbohydrates to total DE should always be less than 20% within intensive salmonid production diets. |
| 3.5.6 | Indigestible starch appears to stimulate feed intake more than cellulose does. |
| 3.5.7 | Three quarters of the faecal organic matter loaded into water from fish is composed of carbohydrates. |
| 3.5.8 | Species differences need further investigation. |
| 3.5.9 | As there is some evidence to suggest that glucose utilization may be affected by dietary chromium oxide inclusion level, alternative markers may have to be used for digestibility studies. |
- drafted by F.J. Schwarz (Germany)
| 3.6.1 | Gross mineral requirements can be determined by the dose-response technique using a wide range of dietary mineral inclusion levels (ie. ranging from suboptimal to possibly subtoxic levels), and calculating mineral requirement by regression analysis. |
| 3.6.2 | In addition to the use of tissue mineral retention or storage as a criterion for determining mineral requirement, there is a need to develop other criteria for estimating requirement such as specific enzyme indicators. |
| 3.6.3 | Studies on comparative mineral availability should be carried out under identical conditions with fish on a suboptimal mineral supply status. |
| 3.6.4 | Considerably further research and understanding is required concerning mineral homeostasis in fish. |
| 3.6.5 | Gross requirement for minerals in fish is dependent upon numerous factors, including fish performance (requirement for a high growth rate), diet (semi-synthetic or practical diets are necessary for requirement studies), chemical binding form, interactions with other dietary elements, interactions with other water borne elements, and interactions with environmental conditions. |
| 3.6.6 | The dietary mineral requirements of fish are species specific, and so mineral requirements must be determined for individual fish species. |
| 3.6.7 | Further research is required concerning elements of special interest, such as phosphorus with respect to water pollution. |
- drafted by J.E.Halver (USA)
| 3.7.1 | Recommended guidelines for determining dietary vitamin requirements or allowances include: |
| |
| 3.7.2 | Requirement studies should ensure that the pH of the diet is recorded together with the methodology (including calibration) employed for dietary and tissue vitamin analysis, and that the experimental test diets be formulated so as to elicit an optimal or maximum growth response. |
- drafted by T. Storebakken (Norway)
| 3.8.1 | Experiments should be carefully designed in order to identify the effect of the additive and identify eventual side-effects with sufficient accuracy. |
| 3.8.2 | Based on European legislation (75/24/EEC; 83/228/EEC; 87/153/EEC), the following testing procedure is suggested: |
| |
| 3.8.3 | Dietary studies with carotenoids should ensure the use of appropriate analytical techniques which take configural isomers into account, the stabilization of feed and faecal samples with liquid nitrogen so as to avoid carotenoid breakdown, and the use of long-term growth studies which monitor the effect of fish size and within-treatment variability. |
- drafted by A.G.J. Tacon (FAO/EIFAC) and M. Jobling (Norway)
| 3.9.1 | During growth trials fish should be fed ad libitum or “to satiation”, and should not receive a restricted ration as this introduces competitive interactions and increases disparity in feed acquisition. Furthermore, concentrated feeding over a limited time period may improve feed distribution in some species (ie. salmonids) since food supply under these conditions is not a “defensible resource”. |
| 3.9.2 | Some attempt should be made to assess feed intake, at the very least on a “tank basis”. Although this should be sufficient for most studies, in some studies it may be advantageous to collect data on individual feed intake using labelled feeds methods such as x-radiography. |
| 3.9.3 | Nutrient requirement studies should be conducted under conditions mimicking as far as possible those of the intended farm production unit and environment, including experimental rearing facility (indoor or outdoor tank, cage or pond), feed preparation technique, feeding method, water quality, photoperiod, and fish stocking density. |
| 3.9.4 | Full feed ingredient descriptions, including International Feed Number (IFN), chemical composition and particle size should be provided when reporting dietary formulations and the results of nutritional feeding trials. |
| 3.9.5 | The growth performance of experimental fish should be at least equal to or greater than that of the target fish species under practical farming conditions so that dietary nutrient requirements can be ascertained under conditions of maximum attainable growth. |
| 3.9.6 | Nutrient digestibility studies should ensure that a range of different feed ingredient inclusion levels and ingredient particle sizes be tested, that digestibility measurements be separated on the basis of feed preparation method employed (ie. cold pelleted feeds, conventional steam pelleted feeds, and extrusion pelleted feeds), and that fish be fed to satiation several times per day under similar on-farm ambient conditions. In addition, it is also recommended that research efforts be focused on the further development of simple in-vitro digestion techniques for the rapid estimation of nutrient digestibility. |
| 3.9.7 | In the unique case of semi-intensive pond farming systems, it is also recommended (in addition to 3.9.3) that a combination of food-web isotope tracer studies and mathematical modelling studies be undertaken so as to study the nutritional contribution of natural food organisms in the overall nutritional budget of pond reared fish. |
| 3.9.8 | So as to ensure the applicability and rapid transfer of research data to farmers it is recommended that where ever possible nutrition experiments be conducted in-situ on representative fish farms and that the data generated from these on-farm research studies also be evaluated from an economic viewpoint. It is also recommended that an analysis of fish quality and tissue histology be undertaken on a routine basis during these extended nutritional feeding trials whenever possible. |
