Fish larval nutrition and feed formulation – knowledge gaps and bottlenecks for advances in larval rearing (a larvanet review).

The requirements of most of the nutrients in fish larvae are unknown, due to difficulties in making the graded, stable and easily bio-available diets needed in dose response experiments. Instead, researchers have tried to find indirect methods to measure the nutrient requirements. One method is to analyze the nutrient profile of the main natural diet, copepods, assuming that these organisms cover the requirements. Another is to extrapolate the requirements in fish to the larval stage. Other methods based on nutrient composition and utilization of the yolk and the uptake and metabolism of radiolabelled nutrients have also been applied. An important aspect is the definition of the requirement, e.g. the requirement for pigmentation and immune function may be different from that for optimal growth and survival. Furthermore, requirements may differ between species and with environmental conditions. In the present review, we look at some requirement estimates according to the present knowledge and compare them to the nutrient composition of live feeds. The protein requirement, extrapolated from juveniles in Atlantic cod and Atlantic halibut, appears to be higher than the protein content of their feed organisms, rotifers and Artemia, respectively. There are two experiments which measure quantitatively the requirements for phospholipids (PL) in fish larvae (ex. Cahu et al., 2003) and which indicate requirements of more that 45% of the lipid. This is higher than the PL content of rotifers and Artemia. The enrichment of rotifers with protein and PL is limited by the volume of their digestive tract, since these nutrients cannot be stored in excess in the rotifer tissues. Rotifers, but not always Artemia, can be enriched to contain enough n-3 fatty acids to cover the requirements in cold water species. Rotifers can occasionally drop below fish requirement levels in certain micronutrients, but enrichment can easily be obtained according to linear dose response models for most of these nutrients. There are thus many gaps in the knowledge before we can claim to feed fish larvae according to their nutrient requirements. First of all, the requirements must be determined. In order to do the proper dose response experiments one will need an experimental diet. The technology for both formulated feeds and for enrichment of live feeds is steadily improving, and some of the nutrients can be studied with the presently available tools. However, further improvement of formulated diets for fish larvae is necessary, both from a scientific and producers point of view. Acknowledgement: This work was financed by Cost Action FA0801-Larvanet