Effect of fish oil replacement and probiotic addition on growth,body composition and histological parameters of yellowtail (Seriola dumerili)

Fish (175 g of initial weight) were fed in triplicated groups with four diets formulated by 0% (FO 100), 75% (FO 25) and 100% (with and without probiotics, FO 0 and FO 0+) of fish oil replacement consisting of a mixture of linseed, sunflower and palm oils. After 109 days, growth and nutritional parameters were not affected by the treatment; however, fish fed with 0% of fish oil showed the lowest survival rate and without differences between the same diet with probiotics. As for biometric parameters, significant differences in the viscerosomatic index (VSI) were observed between fish fed the FO 0+ diet and the FO 100 and FO 25 diets. Results obtained from histological analysis did not detect inflammation in gut samples, while liver samples showed a remarkable steatosis in all four treatments. Total fish oil replacement produced a significant difference in the width of the lamina propria. The dietary inclusion of probiotics in the FO 0+ diet seems to favour a recovery of intestine histology. In addition, as fish oil substitution increased, the width of the lamina propria also increased. In conclusion, it is possible to affirm that the four diets administrated to Seriola dumerili did not compromise the correct development of the animals.     

Partial or total replacement of fishmeal by solvent-extracted cottonseed meal in diets for juvenile rainbow trout (Oncorhynchus mykiss)

The effect of solvent‐extracted cottonseed meal (SCSM) as a partial or total replacement of fishmeal was studied in juvenile rainbow trout (Oncorhynchus mykiss). Six experimental diets SCSM0, SCSM25, SCSM50, SCSM75, SCSM75A and SCSMT, containing a gradient of SCSM 0, 152, 305, 465, 460 and 610 g kg^?1 to replace 0, 112.5, 225, 337.5, 337.5 and 450 g kg^?1 fishmeal protein were fed to triplicate groups (initial body weight of 39.2 ± 0.1 g) for 8 weeks. The diet SCSM75A was supplemented with lysine and methionine, to be similar to SCSM0 for juvenile rainbow trout. Faeces were colleted after 4 weeks of normal feeding for apparent digestibility coefficients (ADC) of dry matter, crude protein and gross energy determination. Total replacement of fishmeal adversely affected growth performance. Fish fed with diet SCSMT had significantly (P < 0.05) lower weight gain, specific growth ratio, feed conversion efficiency (FCE) and protein efficiency ratio than fish fed with other diets. The FCE of SCSM75 and SCSM75A were significantly lower (P < 0.05) than those of fish fed with SCSM0 diets. The ADC of the dry matter of SCSM75 and SCSMT were significantly lower than the SCSM0 diet, and the ADC of crude protein and the energy of SCSMT were the lowest (P < 0.05). The ADC of threonine, proline, alanine, valine, isoleucine, leucine, lysine and methionine of fish fed with diet SCSMT were lower. Lysine and methionine supplement positively affected the ADC of SCS75A diet. There were no significant differences in the fish body composition. It is shown that SCSM can be utilized in the juvenile rainbow trout diet up to 305 g kg^?1, to replace about 50% of fishmeal protein in this experiment.     

Effects of docosahexaenoic acid on growth,survival and swim bladder inflation of larval amberjack ( Seriola dumerili, Risso)

The effect of docosahexaenoic acid (DHA) on the growth performance, survival and swim bladder inflation of larval Seriola dumerili during the rotifer feeding period was investigated in two feeding experiments. Amberjack larvae at 3 day post hatching were fed rotifers enriched with (1) freshwater C hlorella (Chlo), (2) a mixture (2:1, v/v) of Chlo and DHA‐enriched C hlorella (DHA‐Chlo), (3) DHA‐Chlo and (4) DHA‐Chlo and commercial DHA emulsion, in triplicate for 7 days. The average DHA contents of the rotifers were 0.0, 0.4, 1.0 and 1.9 mg g^?1 DM respectively. The survival rate was improved by the enrichment of rotifers with DHA‐Chlo alone, and DHA‐Chlo and emulsion. Growth and swim bladder inflation of fish fed rotifers enriched with DHA‐Chlo were significantly (P < 0.05) improved, however, with increased levels of DHA further improvement was not found. DHA content in the larval whole body proportionally increased with the DHA level in the rotifers. These results suggest that DHA enrichment of rotifers is effective to improve the growth, survival rate and swim bladder inflation of amberjack larvae. The DHA requirement of amberjack larvae is estimated to be 1.5 mg g^?1 on a dry matter basis of rotifers.     

Effect of different rearing conditions on body lipid composition of greater amberjack broodstock (Seriola dumerili)

The aim of this study was to assess the effect of two rearing conditions: outdoor environment with great volume tanks (500 m³) and low stocking density (~0.4 kg m^?3); and indoor environment with smaller volume tanks (10 m³) and higher stocking density (~5 kg m^?3), on muscle, liver and ovary lipid composition of Seriola dumerili broodstock born in captivity. The rearing conditions tested seem to affect the pattern of lipid body deposition in broodstock fish of S. dumerili, increasing the muscle and liver triacylglycerides (TG) accumulation of fish kept under indoor conditions, probably due to reduced energy expenditure in swimming, with some variations in the fatty acid profile that may respond to the differences in the culture conditions tested. No significant differences were found for Gonadosomatic Index (GSI) or ovary lipid deposition between groups in this study, which may suggest that the conditions tested do not have a major effect on ovary development. However, one season later the females kept under outdoor conditions released eggs spontaneously, whereas those kept under indoor conditions did not spawn, suggesting that the conditions tested actually have an effect on the broodstock's reproductive fitness. More studies are needed in order to evaluate whether the conditions tested have or not any influence on ovary development.     

Partial replacement of fishmeal by fermented soybean meal in diets for juvenile white shrimp (Litopenaeus vannamei)

As an important feed ingredient, fermented soybean meal (FSM) has been widely used in aquatic animals due to its stable sources and reasonable price. Here, we evaluated the potential for replacement of fishmeal (FM) with FSM in diet of Litopenaeus vannamei. Five isonitrogenous (410 g/kg) and isolipidic (80 g/kg) diets were formulated: a control diet containing 320 g/kg FM and four experimental diets in which FM in control diet was replaced by FSM at 10 (FSM10%), 20 (FSM20%), 30 (FSM30%) and 40% (FSM40%). An 8‐week feeding trial was conducted in fifteen fibreglass tanks with 50 shrimps per tank. After 8 weeks trial, FSM20% had significantly enhanced growth performance (p < 0.05). No significant difference was found in body composition and digestive enzyme activities of all groups (p > 0.05). Through real‐time quantitative PCR analysis, tor and s6k expression levels of FSM20% were significantly up‐regulated (p < 0.05). Results of western blot showed that the phosphorylation of S6K was not significantly affected by different dietary treatments (p > 0.05), which suggested mTOR signalling pathway was not affected by FSM diets. Based on the above data, 20% replacement of FM with FSM was reasonable and advantageous for L. vannamei diet.     

Assessment of soybean meal in diets for discus (Symphysodon aequifasciata HECKEL) farming through a fishmeal replacement study

A 12‐week feeding trial was carried out with discus (Symphysodon aequifasciata HECKEL), a valuable ornamental species produced in South East Asia, to determine the suitability of using soybean meal as a partial fishmeal substitution in the diet. Juvenile discus (4.3–4.8 g) was fed eight experimental diets with graded levels of soybean meal replacing fish meal (0%, 10%, 20%, 25%, 30%, 35%, 40%, 50%). Growth performance (relative growth rate, specific growth rate) decreased and feed utilization (feed conversion ratio, protein efficiency ratio) was worsened at 30% replacement and higher. Amino‐acid analysis indicated methionine and lysine deficiencies at dietary high soybean replacement levels, although the essential amino‐acid requirements for this species are unknown. The level of soybean in diets negatively affected both dry matter and protein digestibility. This present trial showed the need for further studies involving amino‐acid supplementation, different soy products and palatability enhancement to improve utilization of diets containing soybean meal to allow higher than the maximum 30% replacement level suggested here.     

High‐quality spontaneous spawning in greater amberjack (Seriola dumerili,Risso 1810) and its comparison with GnRHa implants or injections

Production of sufficient high‐quality eggs of greater amberjack (Seriola dumerili) still constitutes the main bottleneck for commercial production of this species. The main objective of this study was to compare the quality of spontaneous spawn of greater amberjack with those obtained by either GnRHa injection or GnRHa implant protocols. Captive amberjack broodstock were distributed in three circular tanks of 40 m³. Broodstock from Tank 1 were not hormonally induced and spawned spontaneously, whereas those of Tank 2 were intramuscularly injected with GnRHa (20 µg/kg body weight) and those of Tank 3 were given EV‐500 µg GnRHa implants. The number of eggs per spawn obtained in the broodstock without hormonal treatment was larger than in those obtained with injections or implants. Egg quality was best in broodstock with spontaneous spawn, followed by GnRHa‐injected fish and then GnRHa implants. Besides, size of larvae from control and injected broodstock was similar between them and significantly higher (p < 0.01) than those from GnRHa implant spawn. Overall, this study showed that it is possible to obtain very high‐quality spontaneous spawn in greater amberjack, providing the adequate conditions. Furthermore, GnRHa weekly injections lead to similar egg viability and hatching rates than spontaneous spawn and higher fertilization rates than GnRHa hormonal implants, which is better than in previous studies.     

Aspergillus niger fermented plant protein mix as a potential substitute for fishmeal in the diet of Penaeus vannamei (Boone, 1931)

A 45‐day feeding trial was conducted to evaluate the utilization of fermented plant protein mix (FPPM) as a fishmeal substitute in the diet of Penaeus vannamei. Six isonitrogenous and isolipidic diets were formulated by substituting fishmeal at 0% (control), 50%, 60%, 70%, 80% and 90% (FR‐0, FR‐50, FR‐60, FR‐70, FR‐80 and FR‐90, respectively) using FPPM. Shrimps fed with diets FR‐50, FR‐60 and FR‐70 had no significant difference in weight gain (173.22%–178.07%) with the control group (179.50%). However, the broken‐line regression indicated that 70.8% fishmeal could be substituted by FPPM. The trend of digestibility was reflected in growth performance. Survival had no significant difference among the treatments. Fishmeal substitution had significant (p < .05) variations in protease and amylase activity, but not in lipase. Except lipid, other body composition was not affected due to dietary change. A decrease (p < .05) in haemolymph total protein, cholesterol and triglycerides was observed to increase in fishmeal substitution. A higher fishmeal substitution (FR‐90) had significantly (p < .05) lower total haemocyte count and the reverse was true for phenoloxidase. Results conclude that a combination of fermented ingredients would be more effective rather using alone in shrimp feed, with a great potential in reducing the pressure on fishmeal.     

Partial and total replacement of fishmeal with poultry by-product meal in diets for gibel carp, Carassius auratus gibelio Bloch

Triplicate groups of gibel carp Carassius auratus gibelio Bloch (initial body weight: 4.89 g) were fed for 8 weeks at 24.8–30.8°C with nine isonitrogenous and isoenergetic diets. The control diet (F1) used white fishmeal (FM) as the sole protein source. In the other eight diets (F2–F9), 40.5–100% of FM protein was substituted by poultry by‐product meal (PBM) at 8.5% increments. The specific growth rate (SGR), feed efficiency ratio, protein efficiency ratio, protein retention efficiency and energy retention rate for fish fed PBM diets (F2–F9) were all higher, but not always significantly, than those for fish fed F1. All apparent digestibility coefficients for fish fed PBM diets were lower than those for fish fed F1. Fish fed F1 had a significantly higher hepatosomatic index value than fish fed PBM diets (P<0.05). No significant (P>0.05) effect of diet was found in whole‐body moisture and fat content. Whole‐body protein and energy content for fish fed PBM diets were slightly higher than that for fish fed F1. The optimal replacement level of FM by PBM was estimated by second‐order polynomial regression to be 66.5% in protein.     

Positive phototaxis as the cause of jaw malformations in larval greater amberjack,Seriola dumerili (Risso, 1810): Mitigation by rearing in tanks with low‐brightness walls

Jaw malformations are a serious problem in fingerlings of the greater amberjack, Seriola dumerili. To establish a countermeasure against these malformations, we conducted a larval behavioural test to identify their root cause, and we investigated the ability of low‐brightness rearing tank walls to control their occurrence. Larval distribution was examined in a horizontally illuminated rectangular transparent aquarium (20 cm × 100 cm × 25 cm), and their distribution was biased towards the light source after this lighting was provided from 3 to 13 days post hatching (dph), indicating positive phototaxis; this response disappeared on 22 dph (there are no data between 13 and 22 dph). Larvae were observed to swim phototactically towards the light source and repeatedly collided with the tank wall (walling behaviour). To investigate the ability of wall brightness to control the incidence of jaw malformations, we compared their incidence in a rearing trial with different Munsell colour values for the tank walls: 2.11 (black), 5.85 (grey) and 9.52 (white). The incidence rate of jaw malformations at 25 dph was significantly lower in the tanks with lower‐brightness walls, that is, 5.8%, 22.5% and 26.8% for black, grey and white respectively. Larval walling behaviour was most intense for white, followed in order by grey and black. These results indicate that positive phototaxis elicits larval walling behaviour, causing jaw malformations, and that low‐brightness tank walls control this process.     

Apparent digestibility coefficient of poultry by‐product meal (PBM) in diets of Penaeus monodon (Fabricius) and Litopenaeus vannamei (Boone), and replacement of fishmeal with PBM in diets of P. monodon

Nutrient apparent digestibility coefficients (ADCs) of pet food grade poultry by‐product meal (PBM) were determined for black tiger prawn, Penaeus monodon and Pacific white shrimp, Litopenaeus vannamei by the indirect method (reference diet and test diet at 7:3 ratio). Subsequently, an 8‐week growth trial was conducted to evaluate the effects of substitution of fishmeal (FM) with PBM in diets of P. monodon (initial weight = 0.21 ± 0.01 g). In the growth trial, six isonitrogenous and isoenergetic diets PBM0, PBM25, PBM50, PBM75, PBM100 and PBMA100, containing a gradient of PBM 0, 88.7, 177.4, 266, 354.7 and 354 g kg^?1 to replace 0, 92.5, 185, 277.5, 370 and 370 g kg^?1 FM were fed to four replicate groups respectively. The diet PBMA100 was supplemented with DL‐Met to be similar to PBM0. The results showed that both P. monodon and L. vannamei had relatively high ADC of crude protein (77.6% and 84.2% respectively) and gross energy (72.8% and 84.0% respectively) for PBM. Litopenaeus vannamei showed significantly higher digestion ability for PBM than P. monodon (P < 0.05). In growth trial, no significant difference in growth performance was observed among shrimp fed the experimental diets. DL‐Met supplementation did not improve the growth of P. monodon. PBM is a suitable protein ingredient for P. monodon feeds and can be used up to 354.7 g kg^?1 to totally replaced FM.     

Evaluation of a fish meal analogue as a replacement for fish meal in practical diets for Pacific white shrimp Litopenaeus vannamei

Three six‐week growth trials and a digestibility trial were conducted to evaluate a fish meal analogue (FMA) as a replacement for fish meal (FM) in shrimp feeds. Trials 1 and 2 evaluated and confirmed the potential of FMA supplementation (0, 48.5, 97, 145.5 and 194 g/kg) as a replacement for FM up to 200 g/kg without balancing for phosphorus (P) in practical diets for juvenile Pacific white shrimp L. vannamei. At the end of trial 1, shrimp offered diets containing 48.5 g/kg FMA exhibited significantly higher weight gain (WG) than those fed with the diet containing 145.5 g/kg FMA. At the end of trial 2, dietary FMA inclusion at 48.5 and 97 g/kg significantly improved WG and protein retention (PR), while reducing FCR and protein content of shrimp body compared to the diet containing 194 g/kg FMA. To determine whether P deficiency is the cause of reduced growth, the third trial was conducted utilizing equivalent diet but balanced for P. At the end of trial 3, shrimp fed diet containing 48.5 g/kg FMA+P showed significantly higher WG and PR than those fed diet containing 145.5 g/kg FMA+P. No decreasing trend of growth was detected in the diets containing FMA compared to the FM‐based diet. Apparent digestibility coefficients of dry matter, energy, protein and amino acids of FMA were determined using chromic oxide as an inert maker and the 70:30 replacement technique. The energy, protein and individual amino acid digestibility of FMA were significantly lower than those of soybean meal and FM which were run at the same time. Results of this work indicate that FMA can replace up to 200 g/kg FM in shrimp diets with supplemental inorganic P. Given the good growth across the range of inclusion without any indication of a growth depression, the low nutrient digestibility of FMA may be due to an atypical response or the product simply does not work with the testing technique.     

Growth performance of juvenile kuruma shrimp,Marsupenaeus japonicus (Bate) fed diets replacing fishmeal with soybean meal

A feeding trail was conducted to evaluate the effect of lowering dietary fishmeal (FM) levels while increasing levels of dehulled soybean meal (SBM) on growth, nutrient utilization and body composition of juvenile kuruma shrimp, Marsupenaeus japonicas. Five experimental diets were formulated to be isoenergetic, isolipidic and isonitrogenous with decreasing FM levels from 40 to 16% while increasing SBM from 0 to 33% respectively. Quadruplicate groups of shrimp (initial wt = 1.5 g) were fed the test diets for 56 days under the flow‐through system. There were no significant differences in final weight (g) and specific growth rate (SGR, % day) among shrimp fed FM40, FM34, FM28 and FM22 diets respectively. Growth parameters significantly decreased in shrimp when fed FM16 diet, which was the lowest level of FM. Feed intake was positively correlated with the SGR of shrimp, and the lowest one was found in shrimp fed FM16 diet. Protein gain and retention, whole body lipid, arginine and methionine significantly decreased in FM16 fed group. Thus, it is concluded that dietary FM could be reduced down to 22% with SBM without compromising growth, nutrient utilization and retention, and whole body composition of kuruma shrimp.     

Apparent digestibility of animal,plant and microbial ingredients for Pacific white shrimp Litopenaeus vannamei

A digestibility trial was conducted to determine apparent digestibility coefficients of dry matter, protein, energy and amino acids of animal, plant and microbial ingredients for Pacific white shrimp, Litopenaeus vannamei. The tested ingredients included traditional soybean meal (TSBM), PepsoyGen soybean meal (PSBM), NutriVance soybean meal (NSBM), fish meal (FM), poultry meal (PM), squid hydrolysis (SQH), scallop hydrolysis (SCH), flash dried yeast (FDY), two batches of Ulva meal (UMF and UMS) and bacteria biomass (BB). A basal diet was formulated and produced along with the experimental diets which included 300 g/kg of each ingredient and 700 g/kg of the basal diet. Juvenile shrimp (initial mean weight: 12 g, six shrimp/tank, n = 3) were stocked in a recirculation system. Apparent dry matter, protein and energy digestibility coefficients ranged from ?40.11% to 78.51%, 15.17% to 97.03% and 13.33% to 82.56% among different protein sources, respectively. In general, protein and energy digestibilities in soy sources (77.6% to 97.03% and 62.77% to 82.56%, respectively) are higher than the tested animal protein (51.39% to 71.41% and 45.29% to 69.77%, respectively) and single‐cell protein sources (15.17% to 53.47% and 13.33% to 40.39%, respectively). Among the three soybean sources, TSBM showed highest protein and energy digestibility. Apparent individual amino acid digestibility coefficients were also variable among different types of ingredients, and there was a reasonable correspondence to protein digestibility. The most digestible feed ingredients for Pacific white shrimp in this study were conventional soybean meal (SBM) and NutriVance soybean meal (NSBM), which indicated that these ingredients are good protein and amino acid sources for Pacific white shrimp. Resultant digestibility data may provide useful information to commercial shrimp feed industry.     

Replacement of fishmeal in diets for Australian silver perch Bidyanus bidyanus (Mitchell). II. Effects of cooking on digestibility of a practical diet containing different starch products

Abstract The interactive effects of cooking and starch type on the dry matter, gross energy and protein digestibility of diets fed to fingerling silver perch were examined. Nine experimental diets were made: a reference diet either uncooked or cooked, six other diets comprising uncooked reference diet supplemented with 30% uncooked or cooked wheat, maize or potato starch and one diet comprising uncooked reference diet supplemented with 15% pregelatinized maize starch. Cooking involved autoclaving at 121 °C for 15 min. The indirect method was used to calculate apparent digestibility coefficients, and chromic oxide (1%) was used as the inert indicator. Both ingredient type and cooking influenced dry matter and gross energy digestibility of diets. All diets had significantly different digestibility coefficients in the following order (from most to least digestible): reference > wheat > maize > potato, for both dry matter and gross energy. Cooking significantly increased the apparent digestibility of diets for both dry matter and gross energy. There was no significant effect of ingredient or cooking or their interaction on protein digestibility. Dry matter and energy digestibility of the diet containing pregelatinized maize starch were significantly higher than for diets containing cooked or uncooked starch products. Ingredient digestibility followed the same trend as diet digestibility. The results from this study indicate that the successful use of starch in practical diets for silver perch is dependent on the origin and processing of the starch. The significant benefits of cooking in the reference diet suggest that pelleted diets should be cooked (e.g. steam conditioned or extruded).     

Partial replacement of fishmeal with biofloc meal in the diet of postlarvae of the Pacific white shrimp Litopenaeus vannamei

This study evaluated the potential for the replacement of fishmeal (FM) by biofloc meal (BM) in the diet of Litopenaeus vannamei postlarvae. Four isonitrogenous (437.42 g kg^?1) and isocaloric (19.84 kJ g^?1) diets were formulated, in which FM was replaced with BM at 0% (T0), 7.5% (T7.5), 15% (T15) and 30% (T30). A commercial diet (CD) was used as reference. The study was conducted over 42 days in 50 L tanks connected to a water recirculation system. All tanks were stocked with three postlarvae (initial weight 0.0023 g) per litre. Shrimp survival, weight gain (WG), final weight (FW), specific growth rate (SGR) and protein efficiency ratio (PER) were measured. Survival rates were high (>91.1%) in all treatments. As the T30 treatment returned a significantly (<0.05) better performance in WG, FW, SGR and PER than all other treatments, partial replacement of FM with BM does not appear to affect productivity. In fact, a regression analysis indicated that a FM to BM replacement level of over 20% may actually improve shrimp growth. In addition to providing a cost‐effective alternative ingredient for L. vannamei postlarvae diets, the production of meal also represents a more sustainable way of disposing of the excess of biofloc produced by BFT systems.     

Partial replacement of fishmeal by soybean meal in diets for juvenile cobia (Rachycentron canadum)

An 8‐week feeding experiment was conducted in floating cages (1.5 × 1.0 × 2.0 m) to determine the potential use of defatted soybean meal (roasted and solvent‐extracted) as a partial replacement of fishmeal in the isonitrogenous (approximately 450 g kg^?1 CP [crude protein]) diet for juvenile cobia with an initial average weight of about 8.3 g. Diets were formulated to include 0, 100, 200, 300, 400, 500 and 600 g kg^?1 (diets D0, D10, D20, D30, D40, D50 and D60, respectively) of fishmeal protein being substituted by defatted soybean meal without methionine supplementation. The results showed that weight gain rate decreased significantly when the replacement level of fishmeal protein was increased from 400 g kg^?1 to 500 g kg^?1, and the D60 diet was the lowest in all groups. These results indicate that up to 400 g kg^?1 of fishmeal protein can be replaced by defatted soybean meal without causing significant reduction in growth. Feed conversion ratio (FCR) and protein efficiency ratio (PER) were significantly affected by the replacement level of fishmeal protein being substituted by defatted soybean meal, when the replacement level of fishmeal protein was 200 g kg^?1 (diet, D20), FCR was the lowest and PER was the highest. There were no significant differences in the moisture, lipid, crude protein and ash content in whole body and muscle, while lipid content in liver increased as the dietary soybean meal replacement levels increased. There were significant differences in haemoglobin, haematocrit, red blood cell, plasma glucose and triglyceride concentration in fish fed diets with different soybean meal replacement levels. Results of this trial indicated that the optimum level of fishmeal protein replacement with defatted soybean meal, determined by quadratic regression analysis was 189.2 g kg^?1, on the basis of maximum weight gain.     

Effects of fish meal replacement with animal protein blend on growth performance,nutrient digestibility and body composition of juvenile Chinese soft‐shelled turtle Pelodiscus sinensis

An 8‐week feeding trial was conducted to evaluate the potential of partial replacement of fish meal with rendered animal protein blend (APB) [meat and bone meal (MBM): expanding blood meal (EBM) = 4 : 1] in the diet of juvenile Chinese soft‐shelled turtle Pelodiscus sinensis. Seven isonitrogenous and isoenergetic practical diets replacing 0%, 10%, 20%, 30%, 40%, 50% and 60% fish meal protein by APB were formulated. Weight gain, feed efficiency, protein efficiency and the apparent digestibility coefficients (ADCs) of dry matter and gross energy in the experimental diets all exhibited no significant difference (P > 0.05) when dietary fish meal was replaced for 0–40%, but decreased significantly (P < 0.05) with further increasing substitution levels. However, relative feed intake, ADC of lipid as well as the contents of whole‐body moisture, protein, lipid and ash all showed little differences (P > 0.05) among all the treatments. Nitrogen retention and the ADCs of protein and most amino acids all decreased significantly (P < 0.05) when fish meal was substituted by more than 30%. The results indicated that fish meal in the practical diet of juvenile Chinese soft‐shelled turtle could be replaced by 30–40% using a combination of MBM and EBM without negative effects on growth performance, nutrients digestibility and body composition.