The apparent digestibility coefficients (ADCs) of amino acids (AA), protein, and energy in six feed ingredients (Brazilian fish meal, soybean meal, corn gluten meal, alcoholic yeast, corn, and wheat bran) we evaluated for pacu juveniles. In general, all ingredients showed high digestibility values for all AA, and differences among ADCs of individual amino acid were detected (P < 0.01). Corn gluten, soybean, and fish meals had the highest ADCs of AA. The ADCs of protein in fish meal, yeast, and corn gluten meal should not be used as AA digestibility indicators, because those showed differences up to 6.7% between the ADCs of protein and AA. All ingredients had lower ADCs of energy than corn gluten meal (P < 0.01). Lysine was the first limiting amino acid in fish meal, corn gluten meal, wheat bran, and corn, and the second limiting amino acid in soybean meal, as methionine was the first limiting amino acid in soybean meal and yeast. However, the soybean meal was the best quality protein source, as it had the highest digestible essential amino acid index. This demonstrated that digestible amino acid values can be used to formulate practical diets for pacu, preventing potential deficiencies or excess that might cause environmental and economic losses.

Due to lack of information on the use of non‐protein energy sources in diets for pacu (Piaractus mesopotamicus), a 2 × 2 × 3 factorial experiment was conducted to evaluate the performance and digestibility of 12 diets containing approximately two crude protein (CP; 220 and 250 g kg−1), two lipid (40 and 80 g kg−1) and three carbohydrate levels (410, 460 and 500 g kg−1). The pacu juveniles‐fed diets containing 220 g kg−1 CP did not respond (P > 0.05) to increased dietary lipid and carbohydrate levels, but the fish‐fed diets containing 250 g kg−1 CP showed a better feed conversion ratio. There were interactions in weight gain (WG), specific growth rate (SGR), crude protein intake (CPI) and feed conversion rate (FCR) dependent on dietary carbohydrate and lipid levels, showing positive effects of increasing carbohydrate levels only for fish‐fed diets containing 80 g kg−1 lipid level. However, when the diets contained 40 g kg−1 lipid, the best energy productive value (EPV) results were obtained at 460 g kg−1 carbohydrate. A higher usage of lipids (80 g kg−1) reduced CPI and was detrimental to protein [apparent digestibility coefficient (ADC)CP] and energy (ADCGE), but did not affect growth. The ADCGE improved proportionally as dietary carbohydrate levels increased (P < 0.05), increasing the concentration of digestible energy. In addition, the WG, CPI, ADCGE results showed best use of the energy from carbohydrates when dietary protein level was 250 g kg−1 CP. The utilization of 250 g kg−1 CP in feeds for juvenile pacu for optimal growth is suggested. Therefore, the optimum dietary lipid and carbohydrate levels depend on their combinations. It can be stated that pacu uses carbohydrates as effectively as lipids in the maximization of protein usage, as long as it is not lower than 250 g kg−1 CP or approximately 230 g kg−1 digestible protein.

To determine the digestible lysine requirement for pacu juveniles, a dose–response feeding trial was carried out. The fish (8.66 ± 1.13 g) were fed six diets containing the digestible lysine levels: 6.8, 9.1, 11.4, 13.2, 16.1 and 19.6 g kg−1 dry diet. The gradual increase of dietary digestible lysine levels from 6.8 to 13.2 g kg−1 did not influence the average values of the parameters evaluated (P > 0.05). The increase of dietary digestible lysine level to 16.1 g kg−1 significantly improved weight gain (WG), specific growth rate (SGR), protein productive value (PPV), protein efficiency rate (PER), and apparent feed conversion rate (FCR), but was not different from fish fed diets containing 19.6 g kg−1 lysine. Fish fed diets containing 16.1 and 19.6 g kg−1 digestible lysine showed lower body lipid contents than fish in the other treatments. The digestible lysine requirement as determined by the broken‐line model, based on average WG values, was 16.4 g kg−1. The other essential amino acid requirements were estimated based on the ideal protein concept and the value determined for lysine.