黄尾鰤排泄物中蛋白质、总脂肪检测方案(抽提萃取)

智能文字提取功能测试中

Recently, yellowtail kingfish (Seriola lalandi) is being cultured in recirculating aquaculture systems (RAS).  Yellowtail kingfish have a poor faecal integrity, which makes the removal of faeces by traditional RAS technology difficult. Reducing the faecal waste load in RAS can be achieved by reducing the amount of faeces  produced (e.g., increasing digestibility) and/or increasing the removal of faeces. This study assessed the effect of  partial fish meal replacement by plant ingredients and the effect of ingredient grinding size on the amount of  faecal waste produced and faecal characteristics, like faecal removal efficiency and particle size distribution  (PSD), in yellowtail kingfish. This was investigated during two 35-d experiments, where fish were fed restrictively (experiment R) or to apparent satiation (experiment S). For each experiment, individual batches of four  experimental diets were produced according to a 2 × 2 factorial design (protein source × ingredient grinding  size). The formulas used were identical for both experiments. FM100 diets contained only fish meal as protein  source, whilst at FM30-P70 diets approximately 70% of the fish meal were replaced by plant protein ingredients.  The effect of ingredient grinding size was tested by including 40% of either a fine or coarse grinding mixture.  Tanks were stocked with 20 fish and 27 fish for experiment R and experiment S, respectively. For each tank, fish  performance, faecal waste production, faecal removal efficiency and faecal PSD were measured. During both  experiments, ingredient grinding size did not affect the faecal removal efficiency or PSD, whilst fish fed the fine  FM30-P70 diets restrictively showed a lower faecal waste production. The inclusion of plant ingredients resulted  in a lower absolute growth and higher FCR. Furthermore, fish fed the FM30-P70 diets showed a higher faecal  waste production, a smaller PSD and a lower faecal removal efficiency. This ultimately resulted in a higher  amount of non-removed faeces by 58.3% and 37.1% compared to FM100 diets for the experiment R and  experiment S, respectively. In conclusion, the replacement of fish meal with plant ingredients in yellowtail  kingfish diets is challenging due to the adverse effects on fish performance, faecal waste production and faecal  characteristics. However, feeding yellowtail kingfish to apparent satiation partly reduced these adverse effects of  plant ingredient inclusion in terms of faecal waste production and faecal characteristics. Reducing the ingredient  grinding size of yellowtail kingfish diets tended to lower the faecal waste production, whilst not negatively  affecting the fish performance or faecal characteristics.  近年来，黄尾鰤正在循环养殖系统（RAS）中养殖。黄尾鰤的排泄物完整性较差，这使得用传统的RAS技术去除排泄物变得困难。可以通过减少排泄物产生的量（如增加消化率）和/或增加排泄物的清除量来减少RAS中的排泄物废物负荷。本研究评估了植物成分替代部分鱼粉的影响，以及成分研磨大小对黄尾鰤排泄物特性的影响，如排泄物去除效率和粒径分布（PSD）。这在两个35天的实验中进行了研究，在这些实验中，鱼被限制性地喂食（实验R）或表观饱足感（实验S）。对于每个实验，根据2×2因子设计（蛋白质源×成分研磨尺寸）生产4个实验饲粮。两个实验所使用的公式是相同的。FM100饲粮中只含有鱼粉作为蛋白质来源，而在FM30-P70饲粮中，我们大约含有70%的鱼粉被植物蛋白成分所取代。通过加入40%的细研磨或粗研磨混合物来测试成分研磨粒径的影响。实验R和实验S分别饲养了20条鱼和27条鱼。测量了每个鱼缸的鱼类性能、排泄物废物产生、排泄物去除效率和排泄物PSD。在两个实验中，成分研磨尺寸均不影响排泄物去除效率和PSD，而严格饲喂FM30-P70精细饲粮的鱼的排泄物废物产量较低。加入植物成分导致了较低的绝对生长和较高的FCR。此外，饲喂FM30-P70饲粮的鱼的排泄物废物产量较高，PSD较小，排泄物清除效率较低。与实验R和实验S的FM相比，FM100饲粮中，未去除的排泄物量分别增加了58.3%和37.1%。综上所述，黄尾鰤鱼饲粮中用植物成分替代鱼粉，由于对鱼类生产性能、排泄物废物产生和排泄物特性的不利影响，具有挑战性。然而，喂食黄尾鰤明显的饱足在一定程度上减少了植物成分在排泄物废物产生和排泄物特性方面的不利影响。减少黄尾鰤饲粮的成分粉磨尺寸往往会降低排泄物废物的产生，同时不会对鱼的生产性能或排泄物特性产生负面影响。Aquaculture 562 (2023) 738875Contents lists available at ScienceDirectAquaculture Aquaculture 562 (2023) 738875P. Horstmann et al. journal homepage： www.elsevier.com/locate/aquaculture 膳食蛋白源和配料粒度对限制进食且明显饱足的黄尾鰤鱼类性能，排泄废物产量和特性的影响 Effect of dietary protein source and ingredient grinding size on fishperformance， faecal waste production and characteristics of yellowtailkingfish (Seriola lalandi) fed restrictively and to apparent satiation Peter Horstmann°aD，Roel M. Maas ， Xander V. de Boer， Theodorus M.B. de Jong，Thomas W.O. Staessen， Fotini Kokou "， Johan W. Schramab，* Kingfish Zeeland， The Kingfish Company， Kats， the Netherlands. 号称“养鱼特斯拉”的荷兰国王鱼公司bAquaculture and Fisheries Group， Wageningen University and Research， Wageningen， the Netherlands 荷兰瓦格宁根大学与研究中心水产养殖和渔业研究院ARTICLEINFO ABSTRACT Keywords： Recently， yellowtail kingfish (Seriola lalandi) is being cultured in recirculating aquaculture systems (RAS).RAS waste production Yellowtail kingfish have a poor faecal integrity， which makes the removal of faeces by traditional RAS tech-Faecal characteristics nology difficult. Reducing the faecal waste load in RAS can be achieved by reducing the amount of faecesFaecal removal efficiencyproduced (e.g.， increasing digestibility) and/or increasing the removal of faeces. This study assessed the effect ofFaeces recoveryTotal suspended solidspartial fish meal replacement by plant ingredients and the effect of ingredient grinding size on the amount offaecal waste produced and faecal characteristics， like faecal removal efficiency and particle size distribution(PSD)， in yellowtail kingfish. This was investigated during two 35-d experiments， where fish were fed restric-tively (experiment R) or to apparent satiation (experiment S). For each experiment， individual batches of fourexperimental diets were produced according to a 2× 2 factorial design (protein source x ingredient grindingsize). The formulas used were identical for both experiments. FM100 diets contained only fish meal as proteinsource， whilst at FM30-P70 diets approximately 70% of the fish meal were replaced by plant protein ingredients.The effect of ingredient grinding size was tested by including 40% of either a fine or coarse grinding mixture.Tanks were stocked with 20 fish and 27 fish for experiment R and experiment S， respectively. For each tank， fishperformance， faecal waste production， faecal removal efficiency and faecal PSD were measured. During bothexperiments， ingredient grinding size did not affect the faecal removal efficiency or PSD， whilst fish fed the fineFM30-P70 diets restrictively showed a lower faecal waste production. The inclusion of plant ingredients resultedin a lower absolute growth and higher FCR. Furthermore， fish fed the FM30-P70 diets showed a higher faecalwaste production， a smaller PSD and a lower faecal removal efficiency. This ultimately resulted in a higheramount of non-removed faeces by 58.3% and 37.1% compared to FM100 diets for the experiment R andexperiment S， respectively. In conclusion， the replacement of fish meal with plant ingredients in yellowtailkingfish diets is challenging due to the adverse effects on fish performance， faecal waste production and faecalcharacteristics. However， feeding yellowtail kingfish to apparent satiation partly reduced these adverse effects ofplant ingredient inclusion in terms of faecal waste production and faecal characteristics. Reducing the ingredientgrinding size of yellowtail kingfish diets tended to lower the faecal waste production， whilst not negativelyaffecting the fish performance or faecal characteristics. 1. Introduction Yellowtail kingfish (Seriola lalandi) has gained attention due to itshigh market value and rapid growth rate (Miegel et al.， 2010； Sorianoet al.， 2018). Yellowtail kingfish is predominantly cultured in sea cages in Spain， Mexico， Chile， Japan， Australia and New Zealand (Moran et al.，2009； Soriano et al.， 2018). However， increasing efforts are being madeto shift the cultivation of yellowtail kingfish in recirculating aquaculturesystems (RAS) (EUMOFA， 2020). One of the advantages of using RAS isthat waste water can be treated to (partly) recover the solid faecal waste Abbreviations： NSP， non-starch polysaccharides； PSD， particle size distribution； TSS， total suspended solids.. * Corresponding author at： Aquaculture and Fisheries Group， Wageningen University and Research， P.O. Box 338， 6700， AH， Wageningen， the Netherlands. E-mail address： johan.schrama@wur.nl (J.W. Schrama). ( Received 1 July 2022； Received in revised form 20 September 2022； Accepted 25 September 2022 ) ( Available online 29 September 2022 ) ( 0044-8486/C 2022 The Authors. Published by Elsevier B.V. This is an open access article under the CC BY license ( http：//creativecommons . org/ li censes/ b y /4 .0/ ). ) and remove/convert dissolved toxic waste products from the system andeffluent water， allowing the reuse of water (Amirkolaie， 2011). How-ever， the cultivation of yellowtail kingfish in RAS is challenging due totheir poor faecal integrity， which is also described as ‘diarrhoea-like'.The unstable faecal consistency and fine faecal particles make it difficultto remove the faecal material from the water， resulting in high con-centrations of total suspended solids (TSS) in the system and effluentwater (Moran et al.，2009). Due to the potential impact of TSS on animalhealth， system performance， operating costs and environmental eutro-phication， the management of faecal waste is a key factor in the successof a RAS (Amirkolaie， 2011； Brinker et al.， 2005； Brinker and Rosch，2005； Chen et al.， 1993； Fernandes and Tanner， 2008； Moran et al.，2009； Schumann et al.， 2016； Unger and Brinker， 2013). In practice，waste management issues might be controlled by either lowering theamount of faecal waste excreted (by improved nutrient digestibility) orimproving the faecal removal efficiency (by improved faeces integrity)(Amirkolaie， 2011； Bureau and Hua， 2010； Cho and Bureau， 1997；Kokou and Fountoulaki， 2018； Tran-Tu et al.， 2018). Seriola spp. feeds currently rely on the inclusion of fish meal (Can-debat et al.， 2020； Dam et al.， 2019； Liu et al.， 2019). Due to theincreasing demand and stagnating supply of marine ingredients， a shiftfrom marine-based to more plant-based diets took place over the lastdecades (Kissinger et al.， 2016； Staessen et al.， 2020a). For yellowtailkingfish， information on fish meal replacement by alternative in-gredients is scarce. Studies with European seabass (Dicentrarchus labrax)and rainbow trout (Oncorhynchus mykiss) reported an increased faecalwaste production and reduced faecal integrity when substituting fishmeal with plant protein sources (Brinker and Friedrich， 2012； Foun-toulaki et al.， 2022； Staessen et al.， 2020a). These negative effects ofplant ingredient inclusion on nutrient digestibility and thus faecal wasteproduction were promoted when rainbow trout were fed to apparentsatiation (Staessen et al.， 2020a). Moreover， it was observed that inparticular fat digestion was negatively affected under conditions ofsatiation feeding. Alterations in the digestibility of nutrients will resultin changes in faecal composition， which could influence the faecalintegrity (Moccia et al.， 2007； Patterson and Watts， 2003； Refstie et al.，2005； Reid et al.， 2009). This shows the importance of feeding level onfaecal waste production and characteristics. Another way to affect the faecal waste production and characteristicsmight be by altering the dietary physico-chemical properties such asingredient grinding size (Callan et al.， 2007； Kahlon et al.， 2006；Staessen et al.， 2020a；Tran-Tu et al.， 2018). The amount of faecal wastecan be altered by the ingredient grinding size as this affects nutrientdigestibility (Tran-Tu et al.， 2018). However， ingredient grinding sizecan also affect chyme viscosity (Tran-Tu et al.， 2018)， which may lead toaltered faecal characteristics. Currently， information on the impact ofingredient grinding size， applied to produce RAS feeds， on the faecalwaste production and characteristics is lacking. This study investigated in yellowtail kingfish whether the proteinsource or ingredient grinding size affect the faecal waste production andcharacteristics， determined as faecal removal efficiency and particle sizedistribution (PSD). The effect of fish meal replacement was tested byformulating diets differing in their main protein source (PS)： FM100 /FM30-P70. The effect of ingredient grinding size (GS) was investigatedby diets altering partly in their ingredient grinding size： fine / coarse.This was obtained by using different screen sizes during milling (fine-1mm / coarse-1.5 mm). To investigate the effect of feeding level， this wasstudied during two independent experiments： restricted / satiationfeeding. 2. Materials and methods 2.1. Diets The effect of protein source and ingredient grinding size on faecalwaste production and characteristics of yellowtail kingfish was investigated during two experiments， where fish were fed restrictively(experiment R) or to apparent satiation (experiment S). Four experi-mental diets were formulated according to a 2×2 factorial design， withprotein source and ingredient grinding size as factors. For each experi-ment， separate batches of these experimental diets were produced. Theeffect of the first factor， protein source， was tested by formulating dietswhich differed in their main protein source by replacing approximately70% of the fishmeal by plant protein ingredients. FM100 diets containedonly fish meal as protein source (68.43%)， whilst at FM30-P70 dietsapproximately 70% of the fish meal was replaced by equal amounts ofwheat gluten， pea protein concentrate and soy protein concentrate(Table 1).FM30-P70 diets were supplemented with DL-methionine andtaurine to fulfil the nutrient requirements regarding amino acids.Furthermore， monocalcium phosphate was added to the FM30-P70 dietsto ensure that phosphorus was not a limiting factor for growth. In alldiets， a minimum of 9.5% fish oil was present to fulfil the requirementsfor essential fatty acids. Additional 1.78% fish oil was added to theFM30-P70 diets in order to achieve a fat content equal to the FM100diets. The diet composition among experiments was similar (Table 1).The effect of the second factor， ingredient grinding size， was tested byincluding 40% of either a fine or coarse grinding mixture. This grindingmixture consisted of 50% fish meal and 50% wheat which were groundusing either a 1 mm (fine) or a 1.5 mm screen (coarse). This procedurewas applied because fine grinding of fish meal is challenging due to itsfat content. The rest of each diet was ground using a 1.5 mm screen.Grinding was done by a hammermill (LHM20/16， 1.5 kW； Condux In-ternational， Mankato， United States of America). The analysed nutrientcomposition is given in (Table 2) and particle size distribution of dietmixture prior to extrusion and physical pellet characteristics in Table 3.The contrast in particle size distribution of the diet mixtures between the Table 1 Diet composition of FM100 and FM30-P70 diets fed during the restricted andsatiation experiment. Protein source FM100 FM30-P70 Ingredients (g/kg) Fish meal LT 484.3 Wheat gluten 150.0 Pea protein concentrate 150.0 Soy protein concentrate" 150.0 Fish oil 95.0 112.8 Monocalcium phosphate 10.0 DL-methionine 4.0 Taurine 5.5 8.0 Premix' 15.0 15.0 Yttrium oxide 0.2 0.2 Grinding mixture Wheat 200.0 200.0 Fishmeal 200.0 200.0 Faroese Fish meal， minimally 71% CP LT (Koster Marine Proteins GmbH，Hamburg， Germany). ( DAmygluten (Tereos Starch & Sweeteners， Aalst， Belgium).Pisane FO (Cosucra， Warcoing， Belgium). ) d SoycomilR(ADMSpeciality Ingredients B.V.， Amsterdam，TTheNetherlands). e Fish oil (BioCeval GmbH & Co. KG， Cuxhaven， Germany). Premix composition. Vitamins (IU or mg/kg complete diet)： Vitamin B1-15mg； Vitamin B2-15 mg； Vitamin B6-15 mg； Vitamin B5-50 mg； Vitamin B3-150mg； Biotine -0.7 mg； B-12-0.05 mg； Folic acid -3 mg； Vitamin C -500 mg(given as ascorbic acid C， phosphate)； Vitamin E-100 IU； A-vitamin A palmitate-10，000 IU； D-Rovimix D3-500-2500 IU； Ks K-menadione sodium bisulphite(51%)-15 mg； Inositol-450 mg； Betaine-500 mg； Choline (given as cholinechloride)-1000 mg； Anti-oxidant BHT (E300-321)-100 mg； Calcium propi-onate -1000 mg. Minerals (mg/kg complete diet)； Ferric sulphate-50 mg； Zincsulphate -80 mg； Cobalt sulphate -0.2 mg； Copper sulphate -8 mg； Sodiumselenite -0.2 mg； Manganese sulphate -30 mg； Magnesium sulphate-750 mg；Chromic chloride-1 mg； Calcium iodate -2 mg. 8 Grinding mixture was grinded at 1 mm for fine diets and at 1.5 mm for coarsediets. Table 2Analysed nutrient content of the experimental diets. Feeding level Restricted feeding Satiation feeding Protein source FM100 FM30-P70 FM100 FM30-P70 Ingredient grinding size Fine Coarse Fine Coarse Fine Coarse Fine Coarse Analysed nutrient content (g/kg DM) Dry matter (DM， g/kg) 940.8 952.2 953.8 958.6 964.0 968.2 975.1 954.9 Crude protein 551.2 549.3 541.6 538.2 550.5 553.2 546.2 545.6 Crude fat 169.3 166.9 168.0 170.1 163.4 160.6 158.1 163.9 Total carbohydrates 155.8 159.1 220.2 223.4 159.6 159.6 226.0 220.2 Starch and sugars 138.7 140.7 156.2 161.7 143.7 145.6 159.0 156.0 NSPb 17.0 18.4 64.0 61.7 16.0 14.0 67.1 64.2 Gross energy (kJ/g DM) 22.7 22.2 23.1 23.2 22.2 22.2 22.9 23.0 Ash 123.7 124.7 70.2 68.3 126.5 126.7 69.6 70.3 Phosphorus 16.8 16.8 11.0 10.7 17.7 18.4 11.2 11.6 Calcium 25.9 25.7 10.7 10.4 26.5 26.9 10.8 10.6 dTotal carbohydrates content (on DM basis) was calculated as： 1000 -(crude protein + crude fat +ash). Non-starch polysaccharides (NSP) content (on DM basis) was calculated as： Total carbohydrates -(starch + sugars). Table 3Physical pellet characteristics and dietary particle size distribution (PSD， %； prior to extrusion) of the experimental diets. Feeding level Restricted feeding Satiation feeding Protein source FM100 FM30-P70 Fine FM100 FM30-P70 Ingredient grinding size Fine Coarse Coarse Fine Coarse Fine Coarse Physical pellet characteristics Hardness (kg) 5.4 5.1 6.1 5.6 5.4 6.5 6.0 6.2 Durability (%) 0.1 0.1 na 0.1 0.1 0.1 0.1 0.1 Bulk density (g/L) 571 561 na 571 613 622 569 636 Gelatinization degree (%) 87.9 82.5 88.6 81.6 81.3 81.3 75.1 73.7 Dietary particle size distribution <40 um 0.4 0.2 5.4 5.2 0.4 0.4 4.1 4.1 40-80 um 8.9 5.1 38.8 34.9 4.9 1.1 37.7 33.7 80-150 um 36.4 35.3 29.6 28.4 25.8 24.1 34.1 32.3 150-250 um 27.3 30.3 9.2 12.2 49.8 50.7 10.7 11.6 250-315 um 5.7 5.7 3.4 3.4 5.6 5.6 3.1 3.1 314-425 um 7.2 6.9 4.6 4.2 6.5 5.9 3.9 3.2 425-630 um 10.4 10.2 8.0 7.8 6.6 6.3 6.0 5.6 >630 um 3.7 6.3 1.1 3.9 0.3 6.0 0.4 6.3 aDurability expressed as feed fines (%). Calculated as： I%oFM ×P%x pum+I%wheat gluten x P%x um +I%pea protein ×P%x um +I%osoy protein × Pox um +I%ogrinding mixture fine × P9x pm+ I%gringind mixture coarse × P%ox um， where I% is the inclusion of each ingredient within the diet and P% the fraction (in %) of particles within each of the fractions (e.g. x-40 um). fine and coarse diets were larger in the experiment S compared to theexperiment R (Table 3). The diets were produced for each experiment individually (differentbatches)by Research Diet Services (Wijk bijDuurstede， TheNetherlands) by extrusion using a Clextral BC45 laboratory scale twin-screw extruder (Clextral， Firminy， France) with a 2 mm die， resultingin 3 mm sinking pellets. After extrusion， the pellets were dried for 3 h(70°C) and afterwards cooled to room temperature. After cooling， partof the oil (80 g/kg) in the formula was added to the experimental dietsby vacuum coating (Vacuum core coater， PegasusQ-10VC， %4 H/VVnozzle nr. 6502) at the Animal Science Group (Wageningen Universityand Research， Wageningen， The Netherlands). Diets were producedapproximately one week prior to the start of the experiments. 2.2. Fish， rearing conditions and housing facilities The experiments were carried out in accordance with the Dutch andEuropean law on the use of experimental animals. The Animal WelfareBody of Wageningen University and Research (The Netherlands) classifiedthese experiments as non-invasive. Fish were kept and handled in agree-ment with EU-legislation. Yellowtail kingfish (Seriola lalandi) of mixed sexwere obtained from a commercial fish farm (Kingfish Zeeland B.V.， TheKingfish Company， Kats， The Netherlands). At the beginning and the endof the experiments， fish were batch weighted (Mettler-Toledo ICS429) to determine initial and final weight and growth. One day prior weighting，fish were starved. Per tank， 20 fish of 105 g and 27 fish of 39 g with werestocked for experiment R and experiment S， respectively. Tanks wereconnected to the same RAS (filled with artificial seawater)， the latterconsisting of a sump， settling tank， drum filter， protein skimmer， andtrickling filter. The system’s refreshment rate was adjusted to keep theNO3-N concentration below 100 mg/L. The water flow over each tank wascontrolled (Magnetic-inductive flow sensor， SM 6000； ifm electroic， Essen，Germany) and kept constant at 7.0±0.05 L/min. The outlet of each tankwas connected to an individual swirl separator (column height 44 cm；diameter 24.5 cm； Aqua Optima AS， Pulford， United Kingdom) to quantifyfeed spillage after feeding and to collect faeces. Water quality parameters were measured daily from the commonoutflow to ensure that the pre-set water quality parameters remainedwithin optimum conditions for yellowtail kingfish. The pH (WTW Multi3630 IDS - SenTix 940) was maintained within the range of 7.0 to 7.6，water temperatures were kept at 23.5±0.3°C and 23.5±0.1°C(WTWMulti 3630 IDS -FDO 925) and salinity at 35.2±0.7 and 34.0±1.3ppm(WTW Multi 3630 IDS - TetraCon 925) during experiment R and S，respectively. During both experiments， the dissolved oxygen concentrationin the outlet water was maintained at a level above 5.0 mg/L (WTW Multi3630 IDS - FDO 925). Maximum allowable values for TAN (total ammo-nium nitrogen， NH4-N and NH3-N combined； Merck Aquamerck Colori-metric Ammonium test)， NO2-N (Merck Aquamerck Colorimetric Nitrite test)， NO3-N concentrations (Merck MQuant Nitrate test strips) were <2mg/L， < 0.06 mg/L， < 1 mg/L， and < 100 mg/L， respectively. Thephotoperiod was set at 20 L：4D for the entire duration of both experiments.Light went on at 7：30 am and switched off at 3：30 am. 2.3. Experimental procedures and sampling During both experiments， treatments were tested for 5-weeks (35 d)and randomly distributed over a total of 12 (experiment R - triplicate)and 16 tanks (experiment S - quadruplicate). During experiment R， fishwere restrictively fed to maintain an equal amount of feed given per tankper day on dry matter (DM) basis for all treatments. The feeding levelwas set at 20 g/kgBW/d which is approximately 80% of the predi-1cated satiation level. Throughout the experiment， the daily amount offeed was gradually increased based on the average initial fish weight andthe predicted daily growth assuming a FCR of 1 for all treatments. Thedaily amount of feed was divided into two equal portions， which werehand fed at 9：00 and 15：00 h. During experiment S， fish were fed twice aday at 9：00 and 15：00 h. Each feeding moment lasted maximally 1 h orwas terminated earlier if fish stopped eating. During the first 3 days ofeach experiment， the feeding level gradually increased until the desiredfeeding level was reached. This allowed the fish to adapt to the diet.Fifteen minutes after feeding， the glass bottles attached to the swirlseparators were checked for feed pellets to determine feed spillage.Mortality was checked twice a day before feeding. Faeces for digestibility analysis were collected overnight for 5 daysduring week 5 by settling (Amirkolaie et al.， 2005). Bottles， which wereconnected beneath the swirl separators， were submerged in ice water tominimize bacterial degradation of the sample. Faecal samples werepooled per tank and stored at -20 °C until further analysis. Faecescollection for determination of faeces removal efficiency was done at theend of the fourth week (experiment R) and end of the fifth week(experiment S). The collecting method was the same as for the faecalsamples collected for digestibility purposes， expect that faecal materialwas collected continuous for 48 h (excluding feeding moments). Faecescollection for determination of faecal PSD was done twice weekly duringthe last two weeks of the experiment (3 h collection during the day aftermorning feeding). One sample per week was used for PSD analysis with aparticle size analyser and one sample per week for PSD analysis bysieving. After collection， faeces were stored on ice until further analysis.Feed samples were taken by pooling 100 g per experimental diet perweek. 2.4. Analysis Faeces collected for digestibility and faeces removal efficiency weredried at 70 °C until constant weight (Staessen et al.，2020a). Thereafter，faeces were pooled per tank and ground (mixer mill， IKA A11 basic).Feed and faeces were analysed as described by Staessen et al. (2020a).For dry matter determination， faeces and feed were analysed gravi-metrically by drying for 4 h at 103 °C until constant weight (ISO 6496，1999). Ash was determined gravimetrically by combustion for 4 h at550°C in a muffle furnace (ISO 5984， 2002) until constant weight. Theash fraction was dissolved in concentrated sulphuric acid by autoclaving(121°C， 20 min) to determine yttrium by ICP-AES (NEN 15510， 2007).Total nitrogen was determined according to Kjeldahl’s method (ISO5983-2，2009)； crude protein was calculated with a protein conversionfactor of 6.25. Crude fat was determined gravimetrically using acidhydrolysis (HydrothermQ， C. Gerhardt GmbH & Co. KG， Konigswinter，Germany) followed by petroleum-ether extraction (Soxhlet method； ISO6492， 1999). Total starch and gelatinized starch were analysed todetermine the gelatinization degree of starch in the experimental diets(Nutrilab， Giessen， The Netherlands). Total starch was analysed enzy-matically using amyloglucosidase after washing with 40% ethanol.Gelatinized starch was analysed according to the modified glucoamylasemethod described by Zhu et al. (2016). For digestibility calculations， starch content (including sugars) of pelleted diets and faeces was ana-lysed as described above for total starch analysis， leaving out the ethanolwashing step. Gross energy was measured using bomb calorimetry(C7000， IKA werke， IKA analysentechnik， Staufen， Germany). PSD of the ingredient mixtures of both diets (prior to extrusion) wasinvestigated by sieving a 50 g sub-sample through a stack of sieves(mesh sizes： 630 um， 425 pm， 315 um， 250 um， 150 um， 80 um and 40um； 10 min sieving time， interval of 6 s， amplitude of 2 mm/'g； Retsch，AS 200 control， Haan， Germany). Pellet hardness was tested using ahardness tester (KAHL Pellet Hardness Tester； AMANDUS KAHL GmbH& Co. KG， Hamburg， Germany). Durability (% feed fines) was deter-mined by sieving a 200 g sub-sample through a sieve (1 mm mesh size； 2min sieving time， interval of 6 s， amplitude of 2 mm/’g；Retsch， AS 200control， Haan， Germany). Bulk density was determined with a 1 L cyl-inder with slide， fall weight and filling cylinder (Biotechnion， Wage-ningen， The Netherlands). Faecal PSD was analysed as a measure to determine faecal charac-teristics. Faecal PSD was determined by using a laser particle analyser(240 s time interval and 90% confidence interval； DIPA 2000， DonnerTechnologies， Or Akiva， Isreal). The particle size analyser was connectedto a liquid flow controller (LFC) in combination with a mechanic stirrer(LFC-101； 150 ml/min flow speed； 20% stirrer speed， around 55 rpm).Prior to the application of the faecal material to the LFC， faeces weresieved using a screen size of 850 um and the upper size was discarded.Tocorrect for the upper size range of the particle size analyser (850 um)，the particle fraction above and below 850 um was determined during thelast two weeks by sieving. Therefore， collected faeces were shortly ho-mogenized (200 rpm， 15 s， MR3000， Heidolph Instruments， Schwabach，Germany) and a sub-sample was applied to an 850 um sieve. Both thefiltrate (< 850 um) and residue (> 850 um) were individually collectedwith pre-weight 1.5 um glass fibre filter (90 mm diameter， grade 696，VWR， Radnor， USA) using a vacuum pump. Filters were stored at -20°Cuntil further analysis. To determine the collected organic matter (OM)mass of the fractions < 850 um and > 850 um， filters were dried andcombusted as described above. 2.5. Calculations and data analysis Absolute growth (g) was calculated as the difference between theaverage individual initial (W；) and final (W) body weight (BW； g). Theabsolute feed intake (FIabs； g/d) was calculated as FItot /t， where FItot is thetotal feed intake (g DM) and t is the number of days during the experi-mental period. Feed intake per metabolic body weight (FImbw； g/kg/d)was calculated as FI / MBW， where MBW is the metabolic body weight(kgo.8) which was calculated as (WG / 1000)0.8. The geometric mean BW(Wc； g) was calculated as e(nWt+lnW0)/2). Feed conversion ratio (FCR) wascalculated on dry matter basis (g/g) as (FI x dmF / 1000)/ (Wf-Wi)，where dmF is the dry matter content of the feed (g/kg). Survival (%) wascalculated as (Nf-N；) × 100， where Ni is the number of fish at thebeginning and Nr the final number of at the end of the experiment. Apparent digestibility coefficient (ADC， %) of organic matter， crudeprotein， crude fat， carbohydrate， starch and gross energy were calcu-lated according to Cheng and Hardy (2002) using yttrium as inertmarker： ADC (%)=100×(1-((Ydiet /Yfaeces)x (Nfaeces /Ndiet)))， whereY is the inert marker percentage of the diet or faeces and N is the nutrientpercentage (or kJ/g gross energy) of the diet or faeces. Organic matter(g/kg DM) and total carbohydrates in feed and faeces were calculated as1000- ash and as 1000 - (crude protein + crude fat +ash)， respectively.Faecal waste production， faecal removal efficiency and non-removedfaeces per feed intake were calculated according to Fountoulaki et al.(2022). Faecal waste production (g OM/kg FI) was determined onorganic matter basis as the amount of non-digested feed per kg feedintake as (100%-ADCoM)×1000， where ADCom is the organic matterdigestibility during week 5. Faeces removal efficiency (FR48h， %) wascalculated as the percentage of collected faeces by settling throughout48 h continuous faeces collection in relation to the total amount of faecal waste production. In detail， this was calculated as the amount of yttriumcollected by settling (Yrecovered， g) in relation to the total amount ofyttrium given via the fed (Ydiet g) as Yrecovered / Ydiet ×100%. The non-removed faeces per feed intake (g OM/kg FI) was calculated as thedifference between the total amount of faecal waste produced and theamount of faeces removed as ((100%-FR48h)×(100%-ADCoM))×1000， where FR48h and ADCom is the faeces removal efficiency duringthe 48 h continuous faeces collection and ADCoM the organic matterdigestibility during week 5. PSD data from the particle size analyser was obtained on volumetricbasis in size classes of 1 um (upper size class 850 um). Data was con-verted into cumulative volume percentage. The upper size range wascorrected by the percentage of particles greater than 850 um. Thefraction of particles > 850 um was determined by sieving as describedabove according to Brinker et al. (2005). 2.6. Statistical analysis Tanks were used as the experimental unit in all statistical analysis (n=12 in experiment R； n=16 in experiment S). Statistical analyses weredone separately per experiment. A two-way ANOVA using a generallinear model was used to investigate the effect of protein source andingredient grinding size and their interaction. In the case of a significantinteraction effect (p <0.05)， a Tukey HSD test (honest significant dif-ference； 95% significance) was performed to compare treatment means.Statistical analyses were performed by using the statistical program IBMSPSS Statistics 27 (IBM， New York， United States of America). 3. Results 3.1. Experiment- Restricted feeding Fish performance (Tables 4 and 5)： Mean survival was high (97.9%)and did not differ between dietary treatments (p > 0.05). Initial bodyweight (105 g) was similar between treatments (p >0.05). As intendedduring the restricted feeding experiment， the absolute feed intake (3.95g DM/d) was equal among treatments. Differences were observed forfinal body weight， growth and FCR between the diets with differentprotein sources (p<0.05).FM100 diets resulted in a higher final BW andgrowth， and a lower FCR compared to FM30-P70 diets. Nutrient digestibility (Tables 6 and 7)： Apparent digestibility co-efficients (ADC， %) of organic matter， crude protein， crude fat， total Table 4 Main effect of protein source on growth performance of yellowtail kingfish fedthe experimental diets restrictively (3 replicates) and to apparent satiation (4replicates) for 35 days. FM100 FM30-P70 SEM PS Restricted feeding Survival (%) 99 97 1.9 ns Initial body weight (g) 105 105 0.7 Final body weight (g) 276 258 0.2 ★ FIabs (g DM/fish/d) 4.0 4.0 0.001 FImbw (g DM/kg9.8/d) 16.3 16.7 0.08 Growth (g/d) 4.9 4.4 0.07 *★ FCR 0.81 0.91 0.013 Satiation feeding Survival (%) 100 100 0.4 ns Initial body weight (g) 40 39 1.1 Final body weight (g) 192 161 4.4 FIabs (g DM/fish/d) 3.4 3.1 0.07 ★ FImbw (g DM/kg/d) 24.0 23.5 0.20 Growth (g/d) 4.4 3.5 0.10 ★★* FCR 0.79 0.89 0.005 ★★★ FIabs - feed intake absolute； FImbw - feed intake metabolic body weight； FCR-feed conversion ratio (on DM basis)； PS - protein source； GS - ingredientgrinding size. Values are means and the standard error of the means (SEM)； ns-not significant p >0.05；*-p<0.05； **-p<0.01； ***-p<0.001. Table 5 Main effect of ingredient grinding size on growth performance of yellowtailkingfish fed the experimental diets restrictively (3 replicates) and to apparentsatiation (4 replicates) for 35 days. Fine Coarse SEM GS Restricted feeding Survival (%) 97 99 1.9 ns Initial body weight (g) 105 105 0.7 ns Final body weight (g) 269 266 0.2 ns FIabs (g DM/fish/d) 4.0 4.0 0.001 一 FImbw (g DM/kg/d) 16.4 16.6 0.08 一： Growth (g/d) 4.7 4.6 0.07 ns FCR 0.85 0.86 0.013 ns Satiation feeding Survival (%) 100 100 0.4 ns Initial body weight (g) 39 40 1.1 ns Final body weight (g) 175 178 4.4 ns FIabs (g DM/fish/d) 3.2 3.3 0.07 ns FImbw (g DM/kg/d) 23.8 23.7 0.20 ns Growth (g/d) 3.9 4.0 0.10 ns FCR 0.84 0.83 0.005 # FIabs - feed intake absolute； FImbw - feed intake metabolic body weight； FCR -feed conversion ratio (on DM basis)； PS - protein source； GS - ingredientgrinding size. Values are means and the standard error of the means (SEM)； ns-not significant p > 0.1； # -tendency p <0.1. Table 6 Main effect of protein source on apparent digestibility coefficient (ADC， %) ofyellowtail kingfish fed the experimental diets restrictively (3 replicates) and toapparent satiation (4 replicates) for 35 days. FM100 FM30-P70 SEM PS Restricted feeding Organic matter 84.7 73.8 0.97 Crude protein 93.2 90.7 0.54 ★★ Crude fat 90.3 74.4 1.27 ★★★ Total carbohydrates 48.9 32.4 2.60 Starch and sugars 80.1 72.3 1.37 ★★★ Energy 87.5 77.4 0.93 *** Satiation feeding Organic matter 82.1 75.9 0.68 ★★★ Crude protein 91.6 92.0 0.32 ns Crude fat 89.4 83.1 0.99 ★★★ Total carbohydrates 41.5 31.1 1.92 *** Starch and sugars 79.5 68.6 1.14 ★★★ Energy 84.8 79.6 0.55 PS - protein source； GS - ingredient grinding size. Values are means and thestandard error of the means (SEM)； ns - not significant p > 0.05；**-p<0.01；***-p<0.001. Table 7 Main effect of ingredient grinding size on apparent digestibility coefficient(ADC， %) of yellowtail kingfish fed the experimental diets restrictively (3 rep-licates) and to apparent satiation (4 replicates) for 35 days. Fine Coarse SEM GS Restricted feeding Organic matter 80.3 78.2 0.97 # Crude protein 92.1 91.7 0.54 ns Crude fat 83.1 81.6 1.27 DS Total carbohydrates 43.6 37.8 2.60 # Starch and sugars 77.7 74.6 1.37 Energy 83.2 81.7 0.93 ns Satiation feeding Organic matter 79.7 78.3 0.68 # Crude protein 91.8 91.9 0.32 ns Crude fat 86.6 85.9 0.99 ns Total carbohydrates 40.0 32.6 1.92 ★★ Starch and sugars 76.3 71.8 1.14 Energy 82.8 81.6 0.55 PS - protein source； GS - ingredient grinding size. Values are means and thestandard error of the means (SEM)； ns - not significant p > 0.1；# -tendency p<0.1；*-p<0.05；**-p<0.01. carbohydrates， starch and energy were affected by the protein source (p< 0.01)， being higher in fish fed the FM100 diets than the FM30-P70diets. A tendency for a higher organic matter (p = 0.064) and totalcarbohydrate ADC (p=0.054) was observed for fine diets in comparisonto the coarse diets. In restrictively fed fish， ADC of all nutrients wasunaffected by the interaction effect of protein source and ingredientgrinding size (supplementary data). Faecal waste production and characteristics： The total amount of faecalwaste production， faecal removal efficiency and the amount of non-removed faeces were affected by dietary protein source p <0.05；Figs.1，2 and 3). The amount of faecal waste production of fish receivingthe FM100 diets (153.4 g OM/kg FI) was 41.4% lower compared to fishreceiving the FM30-P70 diets (261.9 g OM/kg FI； p <0.001). Fineingredient grinding showed a tendency (p=0.064) for a reduced faecalwaste production compared to course ingredient grinding. Fish fed theFM100 diets had higher faeces removal efficiency of 45.4% compared to23.4% for fish fed the FM30-P70 diets (p <0.001)， which is a 94.5%higher removal efficiency at the FM100 diets. No ingredient grindingeffect， nor an interaction effect was observed on faeces removal effi-ciency (p>0.05). Consequently， FM100 diets showed a lower amount ofnon-removed faeces (83.8 g OM/kg FI) by 58.3% compared to the FM30-P70 diets (201.2 g OM/kg FI). Faecal PSD was significantly (p<0.001)affected by the protein source (Table 8， p<0.001). Fish fed FM30-P70diets excreted larger amounts of small faecal particles compared to fishfed FM100 diets (p <0.001). 3.2. Experiment- Satiation feeding Fish performance (Tables 4 and 5)： Mean survival was high (99.8%) and did not differ between dietary treatments (p > 0.05). Initial bodyweight (39 g) was similar between treatments (p >0.05). The feedintake was higher for fish fed the FM100 diets compared to fish fed theFM30-P70 diets (p <0.01). Differences were observed for final bodyweight， growth and FCR between the diets with different protein sources(p<0.05). FM100 diets resulted in a higher final BW and growth， and alower FCR compared to the FM30-P70 diets. None of the performanceparameters was affected by ingredient grinding size or interaction effect(p>0.05). Nutrient digestibility (Tables 6 and 7)： Apparent digestibility co-efficients (ADC， %) of organic matter， crude fat， total carbohydrates，starch and energy were affected by the dietary protein source (p <0.001)， being higher in fish fed the FM100 diets than the FM30-P70diets. Fine ingredient grinding had a positive effect on the starch，energy，total carbohydrate ADC (p <0.05) and tended to improve organicmatter ADC (p=0.063) compared to the coarse ingredient grinding.Only starch ADC was affected by the interaction effect (p < 0.01， sup-plementary data). Fine ingredient grinding resulted in an increasedstarch ADC for FM100 diets， whilst ingredient grinding size did notaffect the starch ADC of FM30-P70 diets (supplementary data). Faecal waste production and characteristics： The total amount of faecalwaste production， faecal removal efficiency and the amount of non-removed faeces were affected by dietary protein source (p < 0.05；Figs. 1，2 and 3). The amount of faecal waste production of fish receivingthe FM100 diets (179.3 g OM/kg FI) was 25.7%lower compared to fishreceiving the FM30-P70 diets (241.3g OM/kg FI； p < 0.001). Fineingredient grinding showed a tendency (p=0.063) of a reduced faecalwaste production compared to coarse ingredient grinding (p <0.1). Fishfed the FM100 diets had higher faeces removal efficiency of 46.9% Fig. 1. Main effects of protein source and ingredient griding size on faecal waste per feed intake (g OM/kg feed intake) of yellowtail kingfish during (a) restrictedfeeding and (b) satiation feeding； OM-organic matter； FI- feed intake； error bars indicate standard error of means； #-tendency p <0.1； ***-p<0.001. Fig. 2. Main effects of protein source and ingredient grinding size on faeces removal efficiency (%) of yellowtail kingfish during (a) restricted feeding and (b)satiation feeding； error bars indicate standard error of means； ns - not significant p> 0.05； ***-p<0.001. Main effect of protein source on faecal particle size distribution (%， PSD) ofyellowtail kingfish fed the experimental diets restrictively (3 replicates) and toapparent satiation (4 replicates) for 35 days. FM100 FM30-P70 SEM PS Restricted feeding <40 um 0.5 0.08 40-100 um 5.2 0.25 ★ 100-250 um 14.1 25.6 1.55 ★ 250-850 um 19.0 30.0 0.93 ★ >850 um 64.0 38.7 2.42 Satiation feeding <40 um 0.5 0.04 40-100 um 6.2 0.41 100-250 um 25.9 29.1 1.52 250-850 pm 32.4 34.9 1.58 >850 um 36.0 29.3 3.30 Values are means and the standard error of the means (SEM)； ns - not significantp> 0.1； #-tendency p < 0.1； **-p<0.01；***-p<0.001. compared to 37.5% for fish fed the FM30-P70 diets (p<0.001)， which isa 25.0% higher removal efficiency at the FM100 diets. Ingredientgrinding size nor the interaction effect affected faeces removal efficiency(p >0.05). Consequently， FM100 diets (95.3 g OM/kg FI) resulted in areduced amount of non-removed faeces by 37.1% compared to theFM30-P70 diets (151.4g OM/kg FI). Faecal PSD at the fraction < 40 umwas significantly (Table 8， p <0.01) affected by the protein source.Compared to the other diets， a larger number of particles were observedin the fraction 250-850 um for fish fed the fine FM30-P70 diet (Table 9，PS × GS； p<0.05). 4. Discussion 4.1. Fish performance and nutrient digestibility In the current study， the effect of dietary protein source， ingredientgrinding size and feeding level on fish performance， faecal waste pro-duction and characteristics of yellowtail kingfish were investigated.Diets were comparable in crude protein and fat content. Carbohydratecontent， in particular non-starch polysaccharides (NSP)， was higher inFM30-P70 diets. Plant ingredient inclusion resulted in reduced feed intake when fishwere fed to apparent satiation. Similar results were observed in a studywith rainbow trout by Staessen et al. (2020a). Literature suggests thatfeed intake is regulated by digestible energy intake (Houlihan et al.，2001； Jobling， 1983). This was not the case in the current study， as fishin the FM30-P70 treatments had a lower digestible energy intake. It islikely that the lower feed intake of yellowtail kingfish， fed the FM30-P70diets， is related to the lower palatability of plant ingredients (Houlihan Table 9 Main effect of ingredient grinding size on faecal particle size distribution (%，PSD) of yellowtail kingfish fed the experimental diets restrictively (3 replicates)and to apparent satiation (4 replicates) for 35 days. Fine Coarse SEM GS Restricted feeding <40 um 0.3 88 0.08 ns 40-100 um 3.9 0.25 ns 100-250 um 20.0 19.7 1.55 ns 250-850 um 24.8 24.2 0.93 DS >850 um 50.9 51.9 2.42 ns Satiation feeding <40 um 0.5 0.5 0.04 ns 40-100 um 5.9 0.41 ns 100-250 um 28.5 26.5 1.52 DS 250-850 um 35.2 32.1 1.58 # >850 um 30.0 35.3 3.30 ns Values are means and the standard error of the means (SEM)； ns- not significantp > 0.1； # -tendency p <0.1. et al.， 2001； Kokou and Fountoulaki， 2018； Sinha et al.， 2011). Plantprotein sources are generally inferior to fish meal in terms of palatability(Houlihan et al.， 2001； Sinha et al.， 2011) which can be related to thepresence of antinutritional factors (ANF). ANF， such as saponins， tan-nins， protease inhibitors， lectins， phytates and NSP， can negatively affectfeed intake (Galkanda-Arachchige et al.， 2019； Houlihan et al.， 2001；Kokou and Fountoulaki， 2018；Krogdahl et al.， 2010； Sinha et al.，2011).For example， saponins which are present in soy and pea products， areknown to have a bitter taste (Francis et al.， 2001；Houlihan et al.， 2001；Kokou and Fountoulaki， 2018； Krogdahl et al.， 2010). During both experiments， fat digestibilities of the FM100 diets werecomparable to those reported by Candebat et al. (2020) and Liu et al.(2019)， whereas higher crude protein digestibilities were observed inthe current study. These results contrast with the lower crude proteinand fat digestibility observed by Dam et al. (2019)， Pirozzi et al.(2019)，and Booth and Pirozzi (2021). Factors that may allude to the differencesin nutrient digestibility between studies are for instance； differences infaecal collection method， water temperature， and other environmentalconditions (Amirkolaie et al.， 2006； Dam et al.， 2019； Pirozzi et al.，2019). In the current study， the inclusion of plant ingredients wereassociated with higher carbohydrate inclusion. This negatively affectednutrient digestibility， in particular fat digestibility， which is consistentwith results in rainbow trout (Staessen et al.， 2020a). However， thenegative effect was greater in the current study with yellowtail kingfishthan in the study with rainbow trout. This suggests that yellowtailkingfish are more sensitive to carbohydrates compared to other fishspecies (Booth et al.， 2013；Maas et al.， 2019； Staessen et al.，2020a). Theoverall negative effects of FM30-P70 diets on nutrient digestibility areexpected due to the higher intake of NSP. NSP are considered as indigestible carbohydrates that can act as ANF (Maas et al.， 2020a； Sinhaet al.， 2011). Literature shows that increasing levels of NSP negativelyaffect the digesta viscosity， resulting in a reduced interaction of enzymeswith the substrate and adversely affect gut morphology and physiology(Maas et al.， 2020b； Refstie et al.， 1999； Sinha et al.， 2011). Moreover，NSP are known to have the potential to bind bile acids in the gastroin-testinal tract. Because NSP are largely indigestible by fish， this can resultin an increasing faecal bile acid loss， as shown in rainbow trout (Staessenet al.， 2020b). Fat digestion is largely dependent on bile acids， the loss ofwhich is expected to result in the lower fat digestion for the FM30-P70diets (Kortner et al.， 2013； Sinha et al.， 2011； Staessen et al.， 2020a，2020b). Furthermore， FM100 diets are expected to have containedgreater amounts of bile acids and their precursors cholesterol andtaurine (all present in fish meal)， whereas they were absent in the plantingredients (Kortner et al.， 2013； Staessen et al.， 2020a，2020b). Another factor that could have resulted in lower nutrient digestibilityof fish receiving FM30-P70 diets is the amount of undigested starch inthe gastrointestinal tract. Fish fed FM30-P70 diets had a greater starchintake and lower starch digestibility. According to the literature， starchcan induce osmotic imbalances and fermentation processes in thegastrointestinal tract， which can affect nutrient digestion (Amirkolaieet al.， 2006； Booth et al.， 2013； Hung et al.， 1990； Kokou and Foun-toulaki， 2018； Refstie et al.， 2005； Sinha et al.， 2011； van Barneveld，1999). Comparing the nutrient digestibilities of FM100 diets amongexperiments， restrictive feeding resulted in higher nutrient digestibility.An increasing feeding level shortens the gut transit time (Bromley，1994)， which ultimately leads to lower nutrient digestibility (Hunget al.， 1990； Miegel et al.， 2010； Staessen et al.， 2020a). However，contradictory results were observed for FM30-P70 diets among experi-ments， where restrictive feeding resulted in a lower nutrient di-gestibility， especially for fat. This can be explained by the findings ofHung et al.(1990)， who showed that disaccharides (breakdown prod-ucts of starch) have a greater negative effect on osmolality and waterretention in the distal intestine of white sturgeon (Acipenser trans-montanus) compared to starch. Accordingly， it is expected that fish fedthe FM30-P70 diets restrictively had a slower gut transit rate， which incombination with the higher gelatinisation degree resulted in greaterbreakdown of starch into mono-and disaccharides， whereas this was notreflected in the starch digestibility data (starch ADC was analysed asstarch， including sugars) (Hung et al.， 1990； Miegel et al.， 2010). In thiscase， the hypothesised greater breakdown of starch would be expectedto increase the negative effects on nutrient digestibility of fish fed theFM30-P70 diets restrictively， whereas this was not the case for FM100diets (Hung et al.， 1990). This suggests that yellowtail kingfish are lesstolerant of mono-and disaccharides compared to starch. However， sinceno clear evidence is presented， this remains only a hypothesis.Overall， itis hypothesised that the higher dietary carbohydrate intake， especiallyNSP， is responsible for the poorer nutrient digestibility of yellowtailkingfish when fed FM30-P70 diets. A positive trend for fine ingredient grinding on organic matter di-gestibility was observed during both experiments. Literature on the ef-fect of ingredient grinding size on nutrient digestibility is conflicting(Callan et al.， 2007； Moreira et al.， 2009； Sveier et al.， 1999； Tran-Tuet al.， 2018； Zhu et al.， 2001). Studies on pigs have shown a positivetrend of fine ingredient grinding on nutrient digestibility (Callan et al.，2007； Moreira et al.， 2009)， whilst in studies on various fish species thistrend was absent or reversed (Sveier et al.， 1999； Tran-Tu et al.， 2018；Zhu et al.，2001). The low contrast in grinding size (only 40% of the totaldiet was ground differently) could explain the absence of a significanteffect for the factor ingredient grinding size. The positive trend of fineingredient grinding on organic matter digestibility could be explainedby a decrease in dietary viscosity and an increased particle surface area，which ultimately improves the mixing of the chyme and the effective-ness of endogenoOuuss enzymes (Callan et al.， 2007； Sinha et al.， 2011；Tran-Tu et al.， 2018). Besides the desired ingredient grinding size contrast among fine and coarse diets， an unintended ingredient grinding size contrast occurredwithin the factor protein source， as FM30-P70 diets had a smalleringredient grinding size compared to FM100 diets. It could be that thisunintended ingredient grinding size contrast positively affected thenutrient digestibility of the FM30-P70 diets compared to the FM100diets. Since higher nutrient digestibilities were observed for the FM100diets， and the effect on nutrient digestibility was greater for the factorprotein source than for the factor ingredient grinding size， this unin-tended ingredient grinding size contrast is unlikely to have had a deci-sive influence on the results. 4.2. Faecal waste production and characteristics One of the challenges of farming yellowtail kingfish in RAS is con-trolling the amount of TSS in the system and discharge water. In prac-tice， both the amount of faecal waste produced and faecal removalefficiency affect the amount of TSS (Amirkolaie， 2011； Bureau and Hua，2010； Cho and Bureau， 1997； Kokou and Fountoulaki， 2018； Tran-Tuet al.， 2018). Both FM100 diets and fine ingredient grinding (tended to)had a positive effect on organic matter digestibility， although the effectbeing greater for the former. As faecal waste production follows theamount of non-digested feed (Kokou and Fountoulaki， 2018)， the FM100and fine (Tendency) diets showed resulted in a lower amount of faecalwaste production. Another important step in waste management is the efficient removalof faecal waste. Faecal characteristics of yellowtail kingfish wereassessed as faecal removal efficiency and PSD， both measurements beingin line with each other. Regarding the effect of ingredient grinding size，it was hypothesised that coarse ingredient grinding would result inhigher chyme viscosity， thus improving faecal integrity (Tran-Tu et al.，2018). In the current study， no effect of ingredient grinding size onfaecal integrity was observed. In terms of protein source， average faecesremoval efficiency over both experiments were 46.2% and 30.4% forFM100 and FM30-P70 diets， respectively. Thus， the inclusion of plantingredients resulted in lower faecal integrity compared to FM100 diets.This is in line with findings of Brinker and Friedrich (2012)， whoobserved lower faeces integrity of rainbow trout when fish meal wasentirely replaced by plant protein ingredients. However， in a study withEuropean seabass， reduced faeces removal efficiency was only observedwhen fish meal was replaced with field peas or feather meal， whilst noeffects were observed when fish meal was replaced with sunflower cake，wheat distillers grain， soy protein concentrate or corn gluten meal(Fountoulaki et al.， 2022). The lower faecal removal efficiency of FM30-P70 diets can beexplained by a treatment effect regarding the faecal waste composition(Patterson and Watts， 2003； Reid et al.， 2009). Fish receiving the FM30-P70 diets had higher carbohydrate intake and lower digestibility，resulting in a higher proportion of undigested carbohydrates in thegastrointestinal tract and ultimately in the faeces. Moccia et al. (2007)and Reid et al. (2009) concluded that faecal composition does notsignificantly influence faecal characteristics. Different findings wereobtained by Patterson and Watts (2003)， who found that the main sourceof non-removed faecal particles are derived from the indigestible cel-lulose fraction and gelatinized starches. Another explanation could be the aforementioned effect of greatercarbohydrate inclusion on osmolality and bacterial fermentation pro-cesses (Amirkolaie et al.， 2006； Booth et al.， 2013； Furuichi and Yone，1981； Hung et al.， 1990； Kokou and Fountoulaki， 2018； Refstie et al.，1999； Shimeno et al.， 1977； Sinha et al.， 2011； van Barneveld，1999).Fermentation processes can lead to gas production， the entrapment ofwhich in faeces can lead to poor faecal integrity (diarrhoea) (Hung et al.，1990； Kokou and Fountoulaki， 2018； Refstie et al.， 1999； Sinha et al.，2011； van Barneveld， 1999). When comparing the experiments amongeach other， similar faecal removal efficiencies were observed for FM100diets. However， lower faeces removal efficiencies were observed in fishfed the FM30-P70 diets restrictively than in fish fed the FM30-P70 diets to apparent satiation. This can be due to the aforementioned hypoth-esised greater presence of mono- and disaccharides for fish fed theFM30-P70 diet restrictively (Hung et al.， 1990； Miegel et al.， 2010)，which in turn has greater implications for faecal integrity (Hung et al.，1990). However， as literature is scarce on the effect of different carbo-hydrate types on faecal characteristics， in particular osmolality， waterreabsorption and bacterial fermentation， this remains only as anobservation. In summary， the inclusion of plant ingredients reducedfaeces removal efficiency. This could be due to both differences in faecalcomposition as well as the negative effects of carbohydrates (especiallymono- and disaccharides) of the plant ingredients. It might be worth-while to investigate the effect of dietary carbohydrate inclusion andtype， on faecal characteristics. Both the amount of faecal waste production and faecal removal ef-ficiency are important factors influencing waste management in RAS asthey determine the amount of non-removed faeces (Amirkolaie， 2011；Bureau and Hua， 2010； Kokou and Fountoulaki， 2018； Tran-Tu et al.，2018). Ingredient grinding size did not affect the amount of non-removed faeces. However， the inclusion of plant protein ingredientsnegatively affected the amount of non-removed faeces. In particular， itbecame clear that the inclusion of plant ingredients increased theamount of non-removed faeces， due to both a lower nutrient digestibilityand faecal removal efficiency. In practice， a higher amount of non-removed faeces would result in higher concentrations of TSS in thesystem water， potentially impairing animal health and system perfor-mance， whilst increasing the operation costs and ultimately contributeto environmental eutrophication when discharged into natural waters(Amirkolaie， 2011； Brinker et al.， 2005； Brinker and Rosch， 2005； Chenet al.， 1993； Fernandes and Tanner， 2008； Schumann et al.， 2016； Ungerand Brinker， 2013). 5. Conclusion Replacing fish meal with plant ingredients in yellowtail kingfishdiets remains a challenge due to the negative effects on fish perfor-mance， faecal waste production and faecal characteristics (faecal parti-cle size distribution and faecal removal efficiency). However， whenfeeding yellowtail kingfish to apparent satiation， these negative effectsof plant protein ingredients on faecal waste production and faecalcharacteristics were partially reduced. Reducing the ingredient grindingsize of yellowtail kingfish diets showed a tendency towards reducedfaecal waste production without affecting fish performance and faecalremoval efficiency. CRediT authorship contribution statement Peter Horstmann： Conceptualization， Methodology， Formal anal-ysis， Investigation， Data curation， Writing - original draft. Roel M.Maas： Conceptualization， Methodology，Data curation， Writing-review& editing. Xander V. de Boer： Investigation， Data curation， Writing-review & editing. Theodorus M.B. de Jong： Investigation， Data cura-tion， Writing-review & editing. Thomas W.O. Staessen： Conceptual-ization， Writing - review & editing， Funding acquisition， Projectadministration. Fotini Kokou： Conceptualization， Writing - review &editing. Johan W. Schrama： Conceptualization， Methodology， Writing- review & editing， Supervision， Project administration. Declaration of Competing Interest The authors declare the following financial interests/personal re- lationships which may be considered as potential competing interests：This work is part of the Healthy Happy Kingfish project applied forby Kingfish Zeeland B.V. under the subsidy scheme Innovation ProjectsAquaculture 2019 and， granted by the RVO (Netherlands EnterpriseAgency) under the application number 19111000012. This project ispartly funded by The European Union with support of the European Maritime and Fisheries Fund (EMFF). Data availability Data will be made available on request. Acknowledgement We would like to thank the staff of the aquaculture research facility(in particular Wian Nusselder and Menno ter Veld) for their technicalsupport in conducting the experiment. Furthermore， we would like toacknowledge Ronald Booms， Samara Hutting and Tino Leffering fortheir support during the lab analysis. This work is part of the HealthyHappy Kingfish project applied for by Kingfish Zeeland B.V. under thesubsidy scheme Innovation Projects Aquaculture 2019 and， granted bythe RVO (Netherlands Enterprise Agency) under the application number19111000012. This project is partly funded by The European Union withsupport of the European Maritime and Fisheries Fund Appendix A. Supplementary data Supplementary data to this article can be found online at https：//doi.org/10.1016/j.aquaculture.2022.738875. References Amirkolaie， A.K.， 2011. Reduction in the environmental impact of waste discharged byfish farms through feed and feeding. Rev. Aquac. 3， 19-26. https：//doi.org/10.1111/i.1753-5131.2010.01040.x. Amirkolaie， A.K.， Leenhouwers， J.I.， Verreth， J.A.J.， Schrama， J.W.， 2005. Type ofdietary fibre (soluble versus insoluble) influences digestion， faeces characteristicsand faecal waste production in Nile tilapia (Oreochromis niloticus L.). Aquac. Res.36， 1157-1166. https：//doi.org/10.1111/j.1365-2109.2005.01330.x. Amirkolaie， A.K.， Verreth， J.A.J.， Schrama， J.W.， 2006. Effect of gelatinization degreeand inclusion level of dietary starch on the characteristics of digesta and faeces inNile tilapia (Oreochromis niloticus (L.)). Aquaculture 260， 194-205. https：//doi.org/10.1016/j.aquaculture.2006.06.039. Booth， M.A.， Pirozzi， I.， 2021. The interaction of feeding regime and dietary specificationon growth and nutrient utilisation in Yellowtail Kingfish Seriola lalandi. Aquaculture544，737094. https：//doi.org/10.1016/j.aquaculture.2021.737094. Booth， M.A.， Moses， M.D.， Allan， G.L.， 2013. Utilisation of carbohydrate by yellowtailkingfish Seriola lalandi. Aquaculture 376-379， 151-161. https：//doi.org/10.1016/j.aquaculture.2012.11.024. Brinker， A.， Friedrich， C.， 2012. Fish meal replacement by plant protein substitution and1Sguar gum addition in trout feed. Part II： effects on faeces stability and rheology.Biorheology 49， 27-48. https：//doi.org/10.3233/BIR-2012-0605. Brinker， A.， Rosch， R.， 2005. Factors determining the size of suspended solids in a flow-through fish farm. Aquac. Eng. 33，1-19. https：//doi.org/10.1016/j.aquaeng.2004.10.003. Brinker， A.， Koppe， W.， Rosch， R.， 2005. Optimizing trout farm effluent treatment bystabilizing trout feces： a field trial. N. Am. J. Aquac. 67， 244-258. https：//doi.org/10.1577/A04-078.1. Bromley， P.J.， 1994. The role of gastric evacuation experiments in quantifying thefeeding rates of predatory fish. Rev. Fish Biol. Fish. 36-66. Bureau， D.P.， Hua， K.， 2010. Towards effective nutritional management of waste outputsin aquaculture， with particular reference to salmonid aquaculture operations. Aquac.Res. 41， 777-792. https：//doi.org/10.1111/j.1365-2109.2009.02431.x. Callan， J.J.， Garry， B.P.， O’Doherty， J.V.， 2007. The effect of expander processing andscreen size on nutrient digestibility， growth performance， selected faecal microbialpopulations and faecal volatile fatty acid concentrations in grower-finisher pigs.Anim. Feed Sci. Technol. 134， 223-234. https：//doi.org/10.1016/j.anifeedsci.2006.09.018. Candebat， C.L.， Booth， M.， Codabaccus， M.B.， Pirozzi， I.， 2020. Dietary methionine sparesthe requirement for taurine in juvenile yellowtail kingfish (Seriola lalandi).Aquaculture 522，735090. https：//doi.org/10.1016/j.aquaculture.2020.735090. Chen， S.， Timmons， M.B.， Aneshansley， D.J.， Bisogni， J.J.， 1993. Suspended solidscharacteristics from recirculating aquacultural systems and design implications.Aquaculture 112， 143-155. https：//doi.org/10.1016/0044-8486(93)90440-A. Cheng， Z.J.， Hardy， R.W.， 2002. Apparent digestibility coefficients and nutritional valueof cottonseed meal for rainbow trout (Oncorhynchus mykiss). Aquaculture 212，361-372. https：//doi.org/10.1016/S0044-8486(02)00260-0. Cho， C.Y.， Bureau， D.P.， 1997. Reduction of waste output from salmonid aquaculturethrough feeds and feeding. Progress. Fish-Culturist. 59， 155-160. https：//doi.org/10.1577/1548-8640(1997)059<0155：ROWOFS>2.3.CO；2. Dam， C.T.M.， Elizur， A.， Ventura， T.， Salini， M.， Smullen， R.， Pirozzi， I.， Booth， M.， 2019.Apparent digestibility of raw materials by yellowtail kingfish (Seriola lalandi).Aquaculture 511，734233. https：//doi.org/10.1016/j.aquaculture.2019.734233. EUMOFA， 2020. Recirculating Aquaculture Systems. https：//doi.org/10.2771/66025. Fernandes， M.， Tanner， J.， 2008. Modelling of nitrogen loads from the farming ofyellowtail kingfish Seriola lalandi (Valenciennes， 1833). Aquac. Res. 39， 1328-1338.https：//doi.org/10.1111/j.1365-2109.2008.02001.x. Fountoulaki， E.， Vasilaki， A.， Nikolopoulou， D.， Schrama， J.， Kaushik， S.J.， Antony JesuPrabhu， P.， 2022. Faecal waste production， characteristics and recovery in Europeanseabass (Dicentrarchus labrax) is affected by dietary ingredient composition.Aquaculture 548，737582. https：//doi.org/10.1016/j.aquaculture.2021.737582. Francis， G.， Makkar， H.P.S.， Becker， K.， 2001. Antinutritional factors present in plant-derived alternate fish feed ingredients and their effects in fish. Aquaculture 199，197-227. https：//doi.org/10.1016/S0044-8486(01)00526-9. Furuichi， M.， Yone， Y.， 1981. Change of blood sugar and plasma insulin levels of fishes inglucose tolerance test. Bull. Japanese Soc. Sci. Fish. 47， 761-764. https：//doi.org/10.2331/suisan.47.761. Galkanda-Arachchige， H.S.C.， Wilson， A.E.， Davis， D.A.， 2019. Success of fishmealreplacement through poultry by-product meal in aquaculture feed formulations： ameta-analysis. Rev. Aquac. 12， 1624-1636. https：//doi.org/10.1111/raq.12401. Houlihan， D.， Boujard， T.， Jobling， M.， 2001. Food Intake in Fish. https：//doi.org/10.1002/9780470999516. Hung， S.S.O.， Groff， J.M.， Lutes， P.B.， Fynn-Aikins， F.K.， 1990. Hepatic and intestinalhistology of juvenile white sturgeon fed different carbohydrates. Aquaculture 87，349-360. https：//doi.org/10.1016/0044-8486(90)90072-U. Jobling， M.， 1983. A short review and critique of methodology used in fish growth andnutrition studies. J. Fish Biol. 23， 685-703. https：//doi.org/10.1111/j.1095-8649.1983.tb02946.x. Kahlon， T.S.， Berrios， J.D.J.， Smith， G.E.， Pan， J.L.， 2006. In vitro bile acid bindingcapacity of milled wheat bran and milled extruded wheat bran at five specificmechanical energy levels. Cereal Chem. 83， 434-438. https：//doi.org/10.1094/CC-83-0157. Kissinger， K.R.， Garcia-Ortega， A.， Trushenski， J.T.， 2016. Partial fish meal replacementby soy protein concentrate， squid and algal meals in ：low fish-oil diets containingSchizochytrium limacinum for longfin yellowtail Seriola rivoliana. Aquaculture 452，37-44. https：//doi.org/10.1016/j.aquaculture.2015.10.022. ( Kokou， F.， Fountoulaki， E.， 2018. Aquaculture waste production associated withantinutrient presence in common fish feed plant ingredients. Aquaculture 495， 295 - 310. htt p s ： // d o i . or g /10 . 1 0 16 /j. aq u aculture . 20 1 8 . 06.00 3 . ) ( Kortner， T.M.， Gu， J.， Krogdahl， A. ， Bakke， A.M.， 2013. Transcriptional regulation ofcholesterol and bile acid metabolism after dietary soyabean meal treatment in Atlantic salmon (Salmo salar L.). Br. J. Nutr. 109， 593 - 604. htt p s ： / / doi . org / 1 0 . 10 1 7 / S0007 1 14 5 120020 2 4 . ) ( Krogdahl ， A.， Penn， M.， Thorsen， J.， Refstie， S.， Bakke， A.M.， 2010. Importantantinutrients in plant feedstuffs for aquaculture： an update on recent f indingsregarding responses in salmonids. Aquac. Res. 41 ， 333-344. h t t p s ：// doi . o rg / 10.1111 /j. 1 3 65-2109 . 2009.0242 6 .x . ) ( Liu， A. ， Pirozzi ， I. ， Codabaccus， B.， Hines， B. ， Simon， C.， Sammut， J. ， Booth， M.， 2019.Digestible choline requirement of juvenile yellowtai l kingfish (Seriola lalandi).Aquaculture 509， 209-220. https ： / / doi .o r g/ 10.1016 / j .aq u acultu r e.2019.05.02 0 . ) ( Maas， R.M.， Verdegem， M.C.J.， Schrama， J.W.， 2019. Effect of non-starch polysaccharidecomposition and enzyme supplementation on growth performance and nutrientdigest i bility i n Nile t i l apia (Oreochromis niloticus). Aquac. Nutr. 25， 622-632. h ttps： // doi . or g/ 1 0.1 1 1 1 /a nu . 128 8 4 . ) ( Maas， R.M.， Verdegem， M.C.J.， Stevens， T.L.， Schrama， J.W.， 2020a. Effect of exogenous enzymes (phytase and xylanase) supplementation on nutrient digestibility and growth performance of Nile tilapia (Oreochromis niloticus) fed different qualitydiets. Aquaculture 529， 735723. h t t p s： / / doi . org / 10. 1 016 / j. a q uac ul ture . 2020.73572 3 . ) ( Maas， R.M . ， Verdegem， M.C.J.， Wiegertjes， G.F.， Schrama， J.W.， 2020b. Carbohydrateuti l isation by tilapia： a meta-analytical approach. Rev. Aquac. 12， 1851 - 1866. h t t ps ：// d o i . o r g / 1 0. 1 1 1 1 / r a q . 1241 3 . ) ( Miegel， R.P.， Pain， S.J.， van Wettere， W.H.E.J. ， Howarth ， G.S. ， Stone， D.A.J.， 2010. Effectof water temperature on gut transit time， digestive enzyme activity and nutrient ) digestibility in yellowtail kingfish (Seriola lalandi). Aquaculture 308，145-151.https：//doi.org/10.1016/j.aquaculture.2010.07.036. Moccia， R.， Bevan， D.， Reid， G.， 2007. Composition of feed and Fecal waste fromcommercial trout farms in Ontario： physical characterization and relationship todispersion and depositional modelling. Technical Research Report for EnvironmentCanada， Ontario Sustainable Aquaculture Working Group. University of Guelph， p.21. Moran， D.， Pether， S.J.， Lee， P.S.， 2009. Growth， feed conversion and faecal discharge ofyellowtail kingfish (seriola lalandi) fed three commercial diets. New Zeal. J. Mar.Freshw. Res. 43， 917-927. https：//doi.org/10.1080/00288330909510050. Moreira， I.， Kutschenko， M.， Paiano， D.， Scapinelo， C.， Murakami， A.E.， de Qu1addros， A.R.B.， 2009. Effects of different grinding levels (particle size) of soybean hull on startingpigs performance and digestibility. Braz. Arch. Biol. Technol. 52，1243-1252.https：//doi.org/10.1590/S1516-89132009000500023. Patterson， R.N.， Watts， K.C.， 2003. Micro-particles in recirculating aquaculture systems：microscopic examination of particles. Aquac. Eng. 28， 115-130. https：//doi.org/10.1016/S0144-8609(03)00028-1. Pirozzi， I.， Benito， M.R.， Booth， M.， 2019. Protein， amino acid and energy utilisation ofjuvenile yellowtail kingfish (Seriola lalandi)： quantifying abiotic influences.Aquaculture 513，734439. https：//doi.org/10.1016/j.aquaculture.2019.734439. Refstie， S.， Svihus， B.， Shearer， K.D.， Storebakken， T.， 1999. Nutrient digestibility inAtlantic salmon and broiler chickens related to viscosity and non-starchpolysaccharide content in different soyabean products. Anim. Feed Sci. Technol. 79，331-345. https：//doi.org/10.1016/S0377-8401(99)00026-7. Refstie， S.， Sahlstrom， S.， Brathen， E.， Baeverfjord， G.， Krogedal， P.， 2005. Lactic acidfermentation eliminates indigestible carbohydrates and antinutritional factors insoybean meal for Atlantic salmon (Salmo salar). Aquaculture 246， 331-345. https：//doi.org/10.1016/j.aquaculture.2005.01.001. Reid， G.K.， Liutkus， M.， Robinson， S.M.C.， Chopin， T.R.， Blair， T.， Lander， T.， Mullen， J.，Page， F.， Moccia， R.D.， 2009. A review of the biophysical properties of salmonidfaeces： implications for aquaculture waste dispersal models and integrated multi-trophic aquaculture. Aquac. Res. 40， 257-273. https：//doi.org/10.1111/j.1365-2109.2008.02065.x. Schumann， M.， Unger， J.， Brinker， A.， 2016. Floating faeces： effects on solid removal andparticle size distribution in RAS. Aquac. Eng. 78， 75-84. https：//doi.org/10.1016/j.aquaeng.2016.10.007. Shimeno， S.， Hosokawa， H.， Hirata， H.， Takeda， M.， 1977. Comparative studies oncarbohydrate metabolism of yellowtail and carp. Bull. Japanese Soc. Sci. Fish. 43，213-217. https：//doi.org/10.2331/suisan.43.213. Sinha， A.K.， Kumar， V.， Makkar， H.P.S.， De Boeck， G.， Becker， K.， 2011. Non-starchpolysaccharides and their role in fish nutrition- a review. Food Chem. 127，1409-1426. https：//doi.org/10.1016/j.foodchem.2011.02.042. Soriano， E.L.， Ramfrez， D.T.， Araujo， D.R.， Gomez-Gil， B.， Castro， L.I.， Sánchez， C.G.，2018. Effect of temperature and dietary lipid proportion on gut microbiota inyellowtail kingfish Seriola lalandi juveniles. Aquaculture 497，269-277. https：//doi.org/10.1016/j.aquaculture.2018.07.065. Staessen， T.W.O.， Verdegem， M.C.J.， Koletsi， P.， Schrama， J.W.， 2020a. The effect ofdietary protein source (fishmeal vs. plant protein) and non-starch polysaccharidelevel on fat digestibility and faecal bile acid loss in rainbow trout (Oncorhynchusmykiss). Aquac. Res. 51， 1170-1181. https：//doi.org/10.1111/are.14467. Staessen， T.W.O.， Verdegem， M.C.J.， Weththasinghe， P.， Schrama， J.W.， 2020b. Theeffect of dietary non-starch polysaccharide level and bile acid supplementation on fatdigestibility and the bile acid balance in rainbow trout (Oncorhynchus mykiss).Aquaculture 523，735174. https：//doi.org/10.1016/j.aquaculture.2020.735174. ( Sveier， H.， Wathne， E.， Lied， E.， 1999. Growth， feed and nutrient uti l isation and gastrointestinal evacuation time in Atlantic salmon (Salmo salar L.)： the effect of dietary fish meal particle size and protein concentration. Aquaculture 180， 265-282. https ： // do i . o rg / 10.1016 / S0 0 4 4 -8486 ( 99)00196 - 9 . ) ( Tran-Tu， L.C.， Hien， T.T.T.， Bosma， R.H. ， Heinsbroek， L.T.N.， Verreth， J.A.J.， Schrama， J.W.， 2018. Effect of ingredient particle sizes and dietary viscosity on digestion andfaecal waste of striped cat fi sh (Pangasianodon hypophthalmus). Aquac. Nutr. 24， 961-969. h ttps： / / doi.or g/ 1 0.11 1 1 / anu.12632 . ) ( Unger， J .， Brinker， A.， 2013. Floating feces： a new approach for efficient removal of solids i n aquacultural management . Aquaculture 404 - 405， 85 - 94. h t t ps ：/ / d o i .o rg / 1 0.101 6/j . a quacul tu re . 2 0 1 3 .04.01 5 . ) ( van Barneveld， R.J.， 1999. Understanding the nutritional chemistry of lupin (Lupinusspp. ) seed to improve livestock production efficiency. Nutr. Res. Rev. 12， 203 - 230. http s ： // do i . o rg / 10.1079/09 5 44229 9 1 0 87289 3 8 . ) ( Zhu， S.， Chen ， S.， Hardy， R.W. ， Barrows， F.T.， 2001 . Digestibi l ity， growth and excretion response of rainbow trout (Oncorhynchus mykiss Walbaum) to feeds of different ingredient particle sizes. Aquac. Res. 32， 885-893. h t t ps ： / /d o i .o r g/1 0.1046 /j . 13 65 - 2109.2001.00624.x . ) ( Zhu， L. ， Jones， C.， Guo， Q.， Lewis， L. ， Stark， C.R.， Alavi， S.， 2016. An evaluation of totalstarch and starch gelat i nization methodologies in pelleted animal feed. J. Anim. Sci. 94， 1501-1507. h t tps： // doi . or g/ 10.2527 / j a s . 20 15 -9822 . ) https：//doi.org/j.aquaculture.