Digestive enzyme responses of tambaqui (Colossoma macropomum) fed on different levels of protein and lipid

The adaptation of digestive proteases, amylase and lipase was studied in tambaqui fed with 350, 253, 301 and 205 g kg^?1 of crude protein, and 49, 81, 113 and 145 g kg^?1 of lipid, in isocaloric diets. Digestive protease increased when dietary protein increased in stomach, where the highest specific activity was observed. Unspecific protease activity in intestine was very low. Lipase was observed throughout the gastrointestinal tract, and higher activities were observed in the stomach. However, pyloric caecum and the anterior and posterior intestines were the responsive sections to the dietary lipid. Amylase was detected throughout the gastrointestinal tract, but pyloric caecum was the most relevant amylase producer. Positive correlations were observed between anterior intestine lipase versus dietary lipid and pyloric caecum amylase versus dietary protein. Tambaqui is responsive to the food composition adapting the main digestive enzymes to the best profile.     

Eugenol as an efficacious anaesthetic for tambaqui, Colossoma macropomum (Cuvier)

Anaesthetics are important in fish culture to reduce handling stress and mortality. Eugenol is a promising anaesthetic because of its low cost, efficacy, safety margin for fish and lack of toxicity to humans. The goal of this study was to establish a protocol using eugenol as a fish anaesthetic for tambaqui Colossoma macropomum (Cuvier), and provide information for regulating authorities on establishing safety dosage protocols for its use. Juvenile and sub‐adult tambaqui were first individually exposed to doses of 35, 50, 65, 85, 100 or 135 mg L^?1 eugenol for 10 min. A second experiment examined the effect of the duration of exposure to eugenol on the time required for recovery and survival of tambaqui. A eugenol dose of 65 mg L^?1 was adequate to induce fish of both sizes into a surgical anaesthetic state, and recovery time was similar for dosages up to 100 mg L^?1. Exposure to the ideal dose (65 mg L^?1) for up to 30 min did not cause fish mortality. Fish blood glucose values were similar for all the tested eugenol doses as well as with the benzocaine control. The results show that eugenol is an efficient and safe anaesthetic for tambaqui.     

Digestible protein requirements and muscle growth in juvenile tambaqui (Colossoma macropomum)

Several investigations have been carried out to improve the productivity of tambaqui, an economically important fish species in Brazil and other Latin American countries. This study determined the digestible protein (DP) requirements in juvenile tambaqui by assessing their productive performance and nutritional efficiency. It also evaluated the effects of different dietary DP levels on the morphology and cellularity of skeletal fast muscle fibres. The 1750 tambaqui tested (6.53 ± 0.43 g body weight, 7.58 ± 0.18 cm length) were randomly distributed into 35 tanks. Fish were fed one of the seven isocaloric diets, which contained 140, 170, 200, 230, 260, 290 or 320 g/kg DP. The DP requirement, calculated by segmented (broken line) regression of weight gain data, was 290 g/kg. An increase in diet DP to 290 g/kg significantly improved final weight, weight gain, feed intake, specific growth rate and crude protein gain, and changed fibre diameter in deep muscle. Muscle fibres were randomly distributed into a mosaic pattern, characterized by fibres with different diameters. Treatments with 290 and 320 g/kg DP increased the frequency of large‐diameter fibres (>50 µm), indicating hypertrophic growth of skeletal muscle during the juvenile phase, which occurred in conjunction with hyperplasia.     

Growth,feed utilization and body composition of Asian catfish (Pangasius hypophthalmus) fed at different dietary protein and lipid levels

This study examined the effect of dietary protein and lipid levels on growth, feed utilization and body composition of Asian catfish Pangasius hypophthalmus reared in cages. Eight test diets were formulated at four protein (340, 380, 420 and 460 g kg⁻¹ crude protein) and two lipid (50 and 90 g kg⁻¹ crude lipid) levels. Fish (initial weight 4.7 g fish⁻¹) were fed the test diets for 8 weeks. Final body weight, weight gain (WG), feed intake (FI), feed conversion ratio (FCR), contents of crude protein, lipid and energy in whole body were dependent on both dietary protein and lipid levels, while specific growth rate (SGR), hepatosomatic index and body moisture content were dependent on dietary lipid level. The WG and SGR increased with the increase in either dietary protein level (at the same lipid level) or lipid level (at the same protein level). The FI and FCR decreased with the increase in dietary protein level (at the same lipid level) or lipid level (at the same protein level). Protein sparing action occurred in case dietary lipid level increased. Fish fed the diet containing 453 g kg⁻¹ crude protein and 86 g kg⁻¹ lipid had the highest WG and SGR, but the lowest FI and FCR, among the diet treatments. There were no significant differences in the protein retention efficiency (PRE) and energy retention efficiency (ERE) among the diet treatments, although PRE and ERE were relatively high in fish fed the diet containing 453 g kg⁻¹ crude protein and 86 g kg⁻¹ lipid. At the end of the feeding trial, body protein content increased, while body lipid content decreased, with the increase in dietary protein content at the same lipid level. Our results suggest that dietary levels of 450 g kg⁻¹ crude protein and 90 g kg⁻¹ lipid are adequate to support fast growth of P. hypophthalmus reared in cages.     

Growth,feed utilization and body composition of juvenile Manchurian trout,Brachymystax lenok (Pallas) fed different dietary protein and lipid levels

A 10‐week feeding trial with four dietary protein levels (400, 450, 500 and 550 g kg^?1 crude protein) and two dietary lipid levels (80 and 160 g kg^?1 crude lipid) was conducted to assess optimum dietary protein and lipid levels for the growth, feed utilization and body composition of juvenile Manchurian trout (initial weight 11.80 ± 0.15 g). Fish were fed twice daily (08:30 and 16:30 h) to apparent satiation. The results showed that fish fed the diet with 500 g kg^?1 protein and 80 g kg^?1 lipid had the highest growth and feed efficiency. However, fish fed the diet with 450 g kg^?1 protein and 160 g kg^?1 lipid showed comparable growth to that of the fish fed diet 5 (500/80) and had higher protein efficiency ratio (PER), nitrogen retention (NR) and energy retention (ER) than other groups (P < 0.05). Growth, PER, NR and ER of fish fed the 160 g kg^?1 lipid diet was significantly higher (P < 0.05) than that of fish fed the 80 g kg^?1 lipid diet at 400 and 450 g kg^?1 protein diet, whereas these values showed an opposite trend at 500 and 550 g kg^?1 protein diet, and the lowest PER, NR and ER was found by fish fed the 400 g kg^?1 protein diet with 80 g kg^?1 lipid. Fish fed diets with 400 g kg^?1 protein had lower feed intake (FI) than that of other groups. Feed intake of fish fed 80 g kg^?1 lipid level was significantly lower than that of fish fed 160 g kg^?1 lipid diet at 400 g kg^?1 protein (P < 0.05), while no significant differences were observed at 450, 500 and 550 g kg^?1 protein‐based diets. Contrary to moisture content, lipid content of whole body and muscle increased significantly (P < 0.05) with increasing lipid levels. The results of this study indicated that the diet containing 450 g kg^?1 protein and 160 g kg^?1 lipid, with a P/E ratio of 23.68 g protein MJ^?1 would be suitable for better growth and feed utilization of juvenile Manchurian trout under the experimental conditions and design level used in this study.     

Tolerance response of tambaqui Colossoma macropomum (Cuvier) to water pH

Tambaqui Colossoma macropomum (Cuvier) is a fish of primary importance in Amazon aquaculture. It has been described as an acid‐resistant species that moves seasonally between white (muddy) water and black water rivers and enters the extremely dilute acidic areas of flooded jungle to feed during the rainy season. To analyse the pH tolerance of this species, tambaqui were exposed to three water pH levels for 40 days (pH 4.0, 6.0 and 8.0). The water was acidified slowly over 3 h, allowing the fish to acclimate. A similar protocol was used to adjust water pH to 8.0. No mortality was observed during the exposure period. Several haematological parameters were significantly changed in alkaline‐exposed animals, with significant decreases in haematocrit (20%), haemoglobin concentration (8%) and red blood cells (12%). Tambaqui showed severe blood variations when exposed to alkaline pH. Fish final weight, condition factor and specific growth rate (SGR) was inversely proportional to a pH increase, and SGR were higher for fish reared in acidic water. The relative insensitivity of tambaqui to low pH confirms its acid tolerance and is in accordance with its natural occurrence in black water habitats.     

Growth of Australian shortfin eel (Anguilla australis) elvers given different dietary protein and lipid levels

The Australian shortfin eel, Anguilla australis is a potential candidate for intensive aquaculture. The present study was undertaken to evaluate the growth of elvers (5.4 g ± 0.1 initial weight) fed with diets of varying protein and lipid content, and to assess the potential of using soya‐bean meal as a dietary ingredient. A 10 week experiment was conducted at 24 (±1.0) °C by rearing fish, in 60 L conical fibre glass tanks using a closed recirculation system. Diets having protein concentrations of 40 or 50% (by dry weight) were tested at three lipid levels (15, 20, 25%); diets being designated P₄₀L₁₅, P₄₀L₂₀, P₄₀L₂₅, P₅₀L_15, etc. All these diets contained 5% soya‐bean meal. In addition P₅₀L₂₀ diets were formulated to contain 10 and 20% soya‐bean meal in the diet (Diets S1 & S2). Shortfin eel grew best on the P₅₀L₁₅ diet, with an average specific growth rate of 2.26%. Food conservation ratio (FCR) and Protein efficiency ratio (PER) ranged from 1.21 (P₅₀L₁₅) to 2.12 (P₄₀L₂₅), and 0.92 (P₅₀L₂₅) to 1.65 (P₅₀L₁₅), respectively. Based on all criteria the best growth performance of shortfin eel was on the P₅₀L₁₅ diet, followed by P₄₀L₂₀ and P₄₀L₁₅. At both protein levels fish reared on diets with 25% lipid performed poorly. The performance of shortfin eel was not affected by the amount of soya‐bean meal in the diet, up to a maximum of 20% dietary inclusion. No significant differences in muscle protein were evident in shortfin eel reared on different dietary treatments, nor was the lipid content of muscle related to dietary lipid level.     

Growth performance and body composition of white seabream (Diplodus sargus) juveniles fed diets with different protein and lipid levels

An 83‐day feeding trial was carried out to determine the effect of different dietary protein and lipid levels on the growth performances and carcass composition of white seabream. Juveniles (10.7±0.2 g) were fed to satiation on four diets, varying in protein (15% and 28%) and lipid (12% and 16%) levels. The best growth performance was observed in fish fed on diets with higher protein level. Dietary lipids did not affect growth performance. Voluntary feed intake decreased with a increasing dietary protein level at both dietary lipid levels. Feed conversion ratio improved with the increase in dietary protein and lipid levels. Carcass composition remained unaltered by dietary protein levels (P>0.05). Carcass protein content tended to decrease, while lipid content tended to increase in groups fed on 16% lipid, compared with the 12% lipid groups. Additionally, protein retention was higher in fish fed on low‐protein and low‐lipid levels, compared with the high‐protein and high‐lipid group (29% vs. 19%). Lipid retention increased significantly with dietary protein level (P<0.001). Energy retention improved with dietary protein, but was not affected by dietary lipid levels. On the basis of our results, feeding white seabream on 15% dietary protein had a negative effect on growth and feed utilization. Dietary lipid did not induce a protein‐sparing action in Diplodus sargus juveniles.     

Effect of different dietary protein and lipid levels on growth and survival of juvenile Australian redclaw crayfish, Cherax quadricarinatus (von Martens)

This study determined the effect of different dietary protein and lipid levels on growth and survival of juvenile redclaw Cherax quadricarinatus. Nine practical test diets were formulated to contain three crude protein (CP) levels [260, 310 and 360 g kg^?1, equivalent to 225, 260 and 296 g kg^?1 digestible protein (DP) respectively] at three crude lipid (CL) levels (40, 80 and 120 g kg^?1, equivalent to 38, 67 and 103 digestible lipids respectively), with digestible protein : digestible energy (DP : DE) ranging from 14.6 to 22.6 mg protein kJ g^?1. Three replicate groups of 15 crayfish (initial weight mean ± SD, 0.71 ± 0.13 g) per diet treatment were stocked in 40 L tanks, at 28 °C for 60 days. The highest mean weight, specific growth rate and biomass, with values of 7.0 g, 3.67% day^?1, and 370.2 g m^?2, respectively, were achieved by feeding a diet with P : L = 310 : 80 (P < 0.05). The treatments resulted in a survival rate of 80–91%, feed conversion ratio of 1.08–1.61 and protein efficiency ratio of 2.24–3.08. Results indicated that a diet containing 270 g kg^?1 DP (equivalent to 320 g kg^?1 CP), 75 g kg^?1 digestible lipid (DL) with a DP/DE of 18.4 mg protein kJ^?1, and 0.031 g protein per animal per day was optimum for juvenile C. quadricarinatus under the tested experimental conditions.     

Effects of aflatoxin B 1 on performance and health of tambaqui fingerlings ( Colossoma macropomum )

The detection of mycotoxins in feeds and their ingredients in aquaculture gained prominence due to losses caused in production and animal health, mainly the occurrence of aflatoxin (AFB1). The aim of this study was to evaluate the effects of AFB1 on the performance of tambaqui fingerlings (Colossoma macropomum). Four hundred tambaqui were used. Four different treatments were evaluated: treatment T1, considered as the control treatment (CT) with 3.84 μg kg−1; treatment T2, treatment T3 and treatment T4 with 500, 1000 and 2000 μg kg−1 of AFB1, respectively. The AFB1 of the samples (muscle, liver and kidney) was detected by high-pressure liquid chromatography. Four fingerlings from each treatment for histological analysis were examined. Moreover, the performance parameters (weight gain, feed conversion and feed intake) were studied. The levels of toxins used in T2, T3 and T4 represent a reduction in the growth of 14%, 35% and 45%, respectively. The T3 and T4 showed the lowest weight gain (78%) and the worst feed conversion. Aflatoxin B1 in muscle (3.28 μg kg−1) and kidneys (8.8 μg kg−1) in the T3, as well as liver (4.4 μg kg−1) and kidney (4.08 μg kg−1) in T4, was detected. Histopathological changes in liver and kidney tissues of fingerlings were more pronounced in T3 and T4. Fingerlings that consume feed contaminated with AFB1 in concentrations higher than 500 μg kg−1 present decreases in growth, reduction in weight gain and feed intake with increased feed conversion. The consumption of feed contaminated with 1000 and 2000 μg kg−1 of AFB1 caused severe deterioration in the hepatic and renal tissues.     

Non-linear release of Bohr protons with haemoglobin-oxygenation in the blood of two teleost fishes; carp (Cyprinus carpio) and tambaqui (Colossoma macropomum)

Some teleost fishes exhibit a non-linear release of H⁺ with haemoglobin oxygen saturation (SO₂) in whole blood that may be related to the Root effect and a low apparent cooperativity in oxygen binding. To further investigate this correlation, the relationship between red cell pH (pH_i) and SO₂ was evaluated in two teleost fishes, the carp (Cyprinus carpio) and tambaqui (Colossoma macropomum). Carp possess a relatively small Root effect within the physiological pH range, while that in tambaqui is relatively large. It was therefore hypothesized that both species would possess a non-linear release of H⁺ with SO₂ under in vivo conditions, but that the degree of non-linearity would be less pronounced in carp than in tambaqui. Red cell suspensions of tambaqui showed a marked non-linear relationship between pH_i and SO₂ at in vivo pH values (pH_e=7.66), where the majority of Bohr protons were released between 50 and 100% SO₂. In whole blood of carp, the relationship between pH_i and SO₂ was almost linear at normal resting extracellular pH values (pH_e=8.10) where Hills n ₅₀ (cooperativity) was 1.3. Under acidic conditions (pH_e=7.11), n ₅₀ decreased to 0.9 and the release of H⁺ with SO₂ became slightly non-linear. The reduction in pH in carp blood was associated with the onset of the Root effect, and the oxygen tension at half saturation (P₅₀) increased from 3.8 to 38.2 mmHg. The relationship between total CO₂ and SO₂ in carp whole blood changed from being almost linear at pH_e 8.1 to being non-linear at pH_e 7.1, consistent with the relationship between pH_i and SO₂. Thus, possession of a Root effect could be a prerequisite for a non-linear release of Bohr protons with oxygenation, but the expression in whole blood may depend upon a species-specific cooperativity threshold. The non-linear release of Bohr protons with oxygenation may be a general phenomenon in teleosts with important implications for gas transport and acid-base homeostasis.     

Growth performance of Colossoma macropomum(Cuvier) juveniles at different feed rations

Abstract Growth performance, feed utilization and changes in body composition of Colossoma macropomum (Cuvier) juveniles — in the weight range 2 to 200g — fed an artificial high protein diet, have been analysed against feed ration and fish weight. A polynomic equation is developed which permits prediction of growth and radon parameters in this weight range. The results are compared with the values of other cultured common freshwater fish species. Colossoma macropomum must be considered a very fast grower with good feed utilization.     

Supplementation of citric acid in plant protein‐based diets for juvenile tambaqui,Colossoma macropomum

We evaluated citric acid supplementation (0, 10, 20, and 30 g/kg) in plant protein‐based diets for juvenile tambaqui, Colossoma macropomum. In Experiment 1, fish (n = 160; 27.56 ± 2.73 g) were distributed in 310 L tanks (n = 16) for 80 days. Zootechnical performance, hematological parameters, and mineral composition (muscle and bones) did not differ significantly (p > .05). The inclusion of citric acid‐reduced pH of diets (p = .001) and stomach (p = .035). In Experiment 2, fish (n = 60; 48.00 ± 1.98 g) were distributed in fecal collection aquaria (200 L) for 21 days. The following were observed: a linear effect for the crude protein (p = .0004) and crude fat (p = .0037) digestibility coefficients; a quadratic effect for crude energy (p = .0037) and dry matter (p = .0004); an optimal supplementation level of 18.5 g/kg. Calcium (p = .0060) and magnesium (p = .0222) showed a linear effect, while phosphorus (p = .0001) had a quadratic effect with the 20 g/kg optimum supplementation level. To conclude, the supplementation of this organic acid in diets does not negatively affect fish nutritional status, health or welfare, and increases nutrient and mineral availability.     

Growth,body chemical composition and trypsin activity of South American catfish,surubim (Pseudoplatystoma sp.) juveniles fed different dietary protein and lipid levels

To evaluate protein and lipid requirement of South American catfish surubim (Pseudoplatystoma sp.) juveniles, nine semi‐purified diets containing three levels of protein (40%, 45% and 50%) and three levels of lipid (12%, 16% and 20%) were tested. After 8‐week feeding trial, body weight increase averaged 2124.3 ± 295.7%. Growth performance was significantly affected by dietary level of protein (P < 0.05). At the 40% protein level, increasing level of dietary lipid had a positive effect on final individual mean weight (protein sparing effect). Whole body protein and moisture contents were affected by the dietary level of lipid (P < 0.05). Whole body lipid content positively correlated with the level of dietary lipid (P < 0.05). Cannibalism related mortality was observed despite rearing fish in 24 h dark. Fatty acid composition of fish was affected by the dietary lipid level (P < 0.05). Polyunsaturated fatty acids increased with the increasing level of dietary lipid while saturated fatty acids and monounsaturated fatty acids decreased. Trypsin activity in the digestive tract of surubim was influenced by dietary levels of protein and lipid (P < 0.05). Our preliminary results suggest that the optimum protein/lipid ratio might be close to 45/16% for surubim juveniles.     

Growth,feed utilization and body composition of juvenile golden pompano Trachinotus ovatus fed at different dietary protein and lipid levels

A feeding trial was conducted to determine the suitable dietary protein and lipid levels for juvenile golden pompano Trachinotus ovatus reared in net pens. Ten test diets were formulated at five levels of crude protein (330, 370, 410, 450 or 490 g kg^?1) and two levels of crude lipid (65 or 125 g kg^?1). Golden pompano fingerlings (initial body weight 4.7 g ind^?1) were fed the test diets for 8 weeks. Weight gain (WG), specific growth rate (SGR), feed conversion ratio (FCR), nitrogen retention efficiency (NRE), energy retention efficiency (ERE), condition factor (CF), hepatosomatic index (HSI), body protein content and total nitrogen waste (TNW) were dependent on both dietary protein and lipid levels. Feed intake (FI) and viscersomatic index (VSI) were dependent on dietary protein level, while body lipid content was dependent on dietary lipid level. Weight gain increased with increasing the dietary protein level (at the same lipid level) but was lower at the dietary lipid level of 65 g kg^?1 than at 125 g kg^?1 (at the same protein level). Fish fed at the dietary protein levels of 460–490 g kg^?1 had higher WG and lower FCR than at 330–410 g kg^?1. Energy retention efficiency tended to increase with increasing the dietary protein level from 330 to 410 g kg^?1, while no significant difference was found in nitrogen retention efficiency between the dietary protein levels (at the same lipid level). Results of this study suggest increasing the dietary lipid level from 65 to 125 g kg^?1 could not induce protein‐sparing action in golden pompano, and the suitable dietary protein and lipid levels for juvenile golden pompano reared in net pens should be 450–490 and 65 g kg^?1.     

Effect of varying dietary lipid and protein levels on growth and reproductive performance of female swordtails Xiphophorus helleri (Poeciliidae)

We conducted a study to determine the effect of dietary protein and lipid levels on the growth and reproductive performance of a freshwater ornamental species, the swordtail (Xiphophorus helleri). Two protein levels (20% and 30%) with four lipid levels (8%, 12%, 16%, 20%) within each protein levels were tested through formulation of practical diets labelled as 20P8L, 20P12L, 20P16L, 20P20L, 30P8L, 30P12L, 30P16L and 30P20L respectively. Results showed that dietary protein level significantly influenced final weight, weight gain and specific growth rate, while dietary lipid did not influence any of these growth parameters. More specifically, increasing dietary lipid levels from 8% to 12–16% in both 20% and 30% protein levels significantly improved swordtail growth performance. Dietary protein levels positively influenced gonadosomatic and hepatosomatic indexes. Both dietary protein and lipid significantly influenced female muscle protein content. Significantly highest fry production was obtained from diets 30P12L and 30P16L, respectively, while diet 20P8L resulted in the lowest fry production. This present study indicates the dietary protein and lipid requirements for female swordtails for optimized growth and reproductive performances to be at 30% and 12% respectively.     

Lippia alba (Verbenaceae) hydrolate as sedative of tambaqui (Colossoma macropomum) juveniles in simulated transport conditions

Lippia alba hydrolate (LAH) is produced as a secondary product in essential oil extraction. The objective of this study was to evaluate the use of LAH in a simulation of the transport of tambaqui (Colossoma macropomum) juveniles. A series of 17 hr simulations were performed using load densities of 30, 60 and 90 juveniles per L and four LAH concentrations of 0.4%, 1%, 2% and 5%, with a control of just water. Fish survival, temperature, dissolved oxygen, conductivity, pH, ammonia excretion, alkalinity water hardness and net Na⁺, K⁺, Cl^‐ fluxes, were all evaluated before and after the transport simulations. Mucous cell densities in the gills were also determined microscopically, post simulation. The results showed that addition of 5% LAH achieved better survival during the simulation than the water control. Dissolved oxygen was significantly lower with an increased load density. Also, conductivity, NH₄ excretion and net ionic fluxes (K⁺ and Cl^?) were lower with 5% LAH, although mainly where higher load densities were tested. Mucous cell density was lower in LAH concentrations of 2% and 5%, again with higher load densities. In view of these findings, it is suggested that 5% LAH is an effective sedative for use in the transport of tambaqui, mainly with high load densities. This work is the first study of LAH and highlights its potential applications in aquaculture management.     

Growth,feed utilization and body composition of African bonytongue,Heterotis niloticus,fingerlings fed diets containing various protein and lipid levels

In order to evaluate the effects of dietary protein and lipid levels on the growth, feed utilization and body composition of Heterotis niloticus fingerlings, a factorial experiment with three replicates was conducted. Six experimental diets containing three crude protein levels (28%, 32% and 36%) and two crude lipid levels (6% and 13%) were tested. Heterotis niloticus (2.34 g) were fed with the diets to apparent satiation, twice a day. For 56 days, weight gain (WG), specific growth rate (SGR), feed efficiency (FE) and protein retention (PR) were significantly affected by dietary protein and dietary lipid levels respectively (P<0.01). The highest WG, SGR and FE were observed for fingerlings fed the diet containing 36% protein and 6% lipid, but no significance difference was found between groups fed with the following diets: P28L13 (28% protein and 13% lipid), P32L6, P32L13 and P36L13. A significant interaction between dietary protein and lipid was observed for WG, SGR, FE and PR. The whole‐body protein, lipid, moisture and ash content were not significantly affected by dietary lipid levels, but body protein and lipid content were significantly affected by dietary protein. The dietary protein‐sparing effect was clearly demonstrated when the dietary energy of lipid increased from 17 to 19.6 kJ g^?1 at 28% crude protein on H. niloticus.