Feeding Strategy Study For Atlantic Salmon (Salmo Salar L.) In Recirculating Aquaculture Systems

Atlantic salmon have been introduced to many countries for cultivating becauseof the advantages of large individual, good flesh and high contents of unsaturatedfatty acids. However, the traditional cage aquaculture is gradually replaced by therecirculating aquaculture systems (RAS), which is due to its vulnerability to naturalenvironmental impacts, high dependence on natural resources (land and water) andlarge impact on the environment. In aquaculture, feed is the major factor indetermining the efficiency and feeding strategy plays a determinant role in regulatingenvironmental impacts of farming. By now, the impacts of feeding strategies onAtlantic salmon and RAS systems have not been reported. This study was conductedto investigate the effects of feed and feeding strategies on the growth of Atlanticsalmon and nitrogen and phosphorus emission of the system. We also constructedempirical weight-based feeding-growth model and nitrogen and phosphorus emissionmodels. Meanwhile, a workshop experiment was used to verify these models. Themain conclusions were as follows:(1) The feed type and daily satiation degree significantly affected salmon growthand feed utilization. Compared with restricted feeding groups (80%satiation),satiation groups had better growth, lower feed conversion ratio and improvedgastrointestinal digestive enzyme activity, while having higher mortality rate(+53.74%~+115.26%). The two factors had significant impacts on the flesh qualityof Atlantic salmon. Increased feeding could improve vitamin E content, water holdingcapacity and pH of muscle, but also reduce muscle hydroxyproline (HYP) content.Although no significant effect s on rheological properties of feces were found, thefecal viscosity in satiation groups was slightly improved. This helped to reduce theconcentrations of small suspended solids in water and improved solids removal rates.(2) Feeding rate, feeding frequency and stocking density had different effectssequence on the growth, digestion of salmon and water quality. For growth performance, the impact order was feeding rate> feeding frequency> breeding densityand the optimal combination was1.2%feeding rate,4times/d frequency and10kg/m3stocking density. For digestion, the impact order was stocking density> feeding rate>feeding frequency and the optimal combination was1.0%feeding rate,2times/dfeeding frequency and15kg/m3stocking density. For nitrogen and phosphorusemissions, the impact order was stocking density> feeding rate> feeding frequency.The optimal combination for reducing total ammonia nitrogen and total nitrogenexcretions was1.0%feeding rate,3times/d feeding frequency and10kg/m3stockingdensity, while the optimal combination for reducing phosphorus excretion rate was0.8%feeding,3times/d feeding frequency and10kg/m3stocking density.(3) The combinations of feeding rate and feeding frequency combination havesignificant effects on growth, feed utilization, body composition and nitrogen andphosphorus emissions of Atlantic salmon. The1.6%feeding rate and4times/dfeeding frequency combination can significantly improve the salmon growth andreduce feed conversion ratio, while improving utilization of nitrogen and phosphorusand the system emissions for N and P. Therefore, the use of80%-90%satiationfeeding and high feeding frequency combinations can get better growth efficiency andenvironmental impact in feeding management of juvenile salmon.(4) The feed loading was adopted to describe the relationship between feeding andwater quality in RAS. Feed loading had significant effects on growth, feed utilizationand mortality of salmon. When fed a load of2.4g/L, Atlantic salmon had optimumgrowth, feed utilization and survival rate. Meanwhile, the deposition rate of nitrogenand phosphorus were the biggest (48.21%and33.74%). Under these conditions, thesystem output the smallest proportion of nitrogen (36.07%), while the proportion ofphosphorus output was the largest (73.93%).(5) Based on the above results, the growth model and excretion model for N and Pwere as follows:Feeding-growth model: G=-0.023F×lnW+0.224F-0.016lnW+0.682N excretion model: No=2.10×10-4F+4.94×10-4W1.0117P excretion model: Po=3.69×10-4F+2.61×10-4W0.7605The above model accuracy was tested through a designed workshop experiment.The deviation of feeding growth model was12.68%, while nitrogen and phosphorusemission model deviation were17.93%and23.65%. The results showed that themodel had good predictive ability.