Lysine requirements for maintenance and utilization eciency of Tambaqui (Colossoma macropomum) at different body weights

ECO-CAL2K). The analysis of the total amino acid (TAA) content of corn and soybean meal was performed by high performance liquid chromatography (HPLC). To obtain the digestible essential amino acids (DAA), the values of TAA were converted using the digestibility coecients of corn and soybean meal for


Background
Tambaqui (Colossoma macropomum) is a species of sh found widely in the Amazon basin and is one of the most cultivated shes in Brazil and other countries of South and Central America, owing to its great robustness, ease of reproduction in captivity and the organoleptic characteristics of its meat [1,2].
Despite the great importance of tambaqui in these regions, diet formulations for this species generally use available data on the nutritional requirements of other tropical sh species, which may not ensure satisfactory performance due to the physiological particularities existing in each species.
Lysine stands out as one of the rst limiting essential amino acids in practical sh diets, especially when formulated with ingredients of plant origin [3]. In addition, lysine is used almost exclusively for protein synthesis, being found at a higher proportion in muscle tissues [4 6]. Therefore, precise estimation of requirement this amino acid is necessary to formulate e cient and low-cost diets.
The methodology for determining the lysine requirement that has been gaining popularity in studies with sh is the factorial method, which estimates the requirements on the principle that sh need lysine for the maintenance of vital processes and growth. The lysine requirement for maintenance can be de ned as the balance between lysine intake and body protein (nitrogen) excretion, with no change in body protein content. The e ciency of lysine utilization for growth can be determined by the slope coe cient of a straight line between consumption of lysine and its retention [7 9].
The factorial method allows the elaboration of models capable of predicting the nutritional requirements of sh, as it contemplates differences in weight, body composition, growth potential, in different breeding conditions. Body weight can have a major in uence on the requirement for maintenance and e ciency of utilization of lysine in sh [ 10 12].
However, the application of the factorial method depends on the determination of the parameters that express the maintenance requirements, and the e ciency of utilization of the amino acids in the diet, and this information can be obtained from dose-response studies, speci cally planned. Given that there are no studies with tambaqui, the objective was to determine the requirement for lysine for maintenance and the e ciency of utilization for tambaqui with different body weights.

Chemical analysis of food and experimental diets
To formulate the experimental diets, analysis of corn, soybean meal, corn starch and soy oil were performed ( Table 1). The crude protein (CP) of the food was determined by the standard method [13], and the percentage of nitrogen was obtained by the Kjeldahl method after acid digestion. Gross energy (GE) content was measured by combustion in a Parr bomb calorimeter (model: ECO-CAL2K). The analysis of the total amino acid (TAA) content of corn and soybean meal was performed by high performance liquid chromatography (HPLC). To obtain the digestible essential amino acids (DAA), the values of TAA were converted using the digestibility coe cients of corn and soybean meal proposed by [14] for tambaqui.  The experimental diets were formulated using the "dilution" technique [15]. Initially, a reference diet (RD) was formulated containing 18.06% crude protein (CP) and 0.901% digestible lysine, based on corn and soybean meal (Table 2). Subsequently, this was diluted with another protein-free diets (PFD) based on corn starch, containing the same levels of energy, vitamins, and minerals, enabling levels of 0.225, 0.450, 0.675 and 0.901% digestible lysine (Table 2). After preparing the experimental diets, the total amino acids were analyzed and their contents con rmed with the formulation targets.
To con rm that lysine was the rst limiting nutrient in experimental diets, a fth treatment (CD: control diet) was added, with ve replications, to which synthetic lysine was added at the rst level (0.225%) until reaching the second level (0.450%). Thus, it is possible to con rm that the response obtained in the study occurred due to the variation in lysine and not due to the variation between crude proteins (4.51-18.06%) of the experimental diets.
In experimental diets, the ratios of methionine plus cystine / lysine, threonine / lysine and tryptophan / lysine were maintained above the ideal protein ratio proposed by [16 18], respectively, for tambaqui (Table 2), in order to prevent another amino acid from becoming limited for each level of digestible lysine evaluated.  3 Values calculated based on the digestibility coe cients determined by [19,20]. 4 Values calculated based by [21]. 5 Values calculated based on the digestibility coe cients determined by [14].
For the extrusion process of experimental diets, the ingredients of PDF and RD were nely ground (Trf 60, Trapp®) and weighed individually and mixed ( Each experimental trial lasted 20 days, and ve days before the beginning of the experimental period, the sh were acclimated to the experimental conditions. In each experiment, 100 tambaqui with different initial mean weights of 121 ± 1.35; 235 ± 1.23 and 596 ± 47.57 g were used, distributed in a completely randomized design, consisting of four treatments (0.225; 0.450; 0.675 and 0.901% of digestible lysine in the diets) and ve replicates, with 5 sh per experimental unit. A fth treatment (CD: control diet) was added, with ve replications, to con rm that the response obtained was due to the variation of digestible lysine and not to the variation between crude proteins (4.51 -18.06%).
During the experimental period, the sh were kept in polyethylene boxes, with a capacity of 1000 L each, equipped with individual water supplies, drainage systems, and aeration systems, the water supply for the boxes was derived from an artesian well, with ow rate of 40 L h -1 per box.

Water quality parameters
The water temperature was measured daily at 7:00 and 16:00 h with a mercury bulb thermometer graduated from 0 to 50 °C. Controls for pH and the content of dissolved oxygen and ammonia in the water were measured every three days using a pH meter (HI 8424, Hanna®), oximeter (HI 9146, Hanna®) and commercial kit (Arcor®) for toxic ammonia test, respectively.

Carcass preparation and analysis
The comparative slaughter technique was used to study maintenance requirements and e ciency of utilization of lysine of tambaqui. At the beginning of each experiment, 15 sh from the same population used in each experiment were euthanized (benzocaine 100 mg L -1 , after fasting for 24 h) and frozen in an ultra-freezer with temperature -70 °C for later determination of the initial body protein content. At the end of the study, after a 24 h fast, all sh in each experimental unit were weighed, euthanized (benzocaine 100 mg L -1 ) and frozen in an ultra-freezer with temperature -70 ° C to determine the nal body protein content.
The crude protein of whole body tambaqui was determined by the standard method [13], and the percentage of nitrogen was obtained by the Kjeldahl method after acid digestion.
The analysis of lysine content of whole body of tambaqui was performed using high performance liquid chromatography (HPLC) at the CBO laboratory (Valinhos, Brazil).
For the calculate weight gain (%) the following formula was used: For the calculate body protein retention (BPR) were used analyzes of the initial body protein content (initial sh samples) and analyzes of the nal body protein content, calculated using the following formula: -BPR (mg kg -0.7 day -1 ) = ( nal body protein (%) × nal weight (g)) -(initial body protein (%) × initial weight (g) / experimental period / nal weight (kg -0.7 ).
Maintenance lysine requirements were determination based on digestible lysine intake (mg kg -0.7 day -1 ) that provided zero protein retention (mg kg -0.7 day -1 ).The e ciencies for using lysine for growth were obtained by the slope coe cient of the line between the digestible lysine consumption (mg kg -0.7 day -1 ) as a function of body lysine retention (mg kg -0.7 day -1 ) [8,12].

Statistical analysis
The data obtained, in each experimental trial, were submitted to an analysis of variance. For those with signi cant effects as a function of digestible lysine levels, the means of each treatment were compared using the Duncan test. Linear regression analysis of protein and body lysine deposition as a function of digestible lysine consumption was also performed.
Additionally, to compare the responses of the different body weights, the body protein retention (BPR) and body lysine retention (BLR) data obtained in the experimental tests were compared using the parallelism test [22] between equations two by two. The weight was a categorical variable and DLI (mg kg -0.7 day -1 ) was a covariate in the model: -BPR / BLRij = β 0 + BWi + β 1 *DLIij + Σiβ 2 i (BW* DLI)ij + e ij ; Where BPR / BLRij = body protein retention / body lysine retention corresponding to the j body weight observation i; BWi = effect of body weight; DLIij = digestible lysine intake; β0, β1, and β2i = regression parameters; (BW*DLI)ij = effect of the interaction between the categorical variable and covariate; e ij = random error associated with observation j of body weight i.
In this case, the hypotheses tested were: a) H0: Gi = 0 for all i, there is no effect on body weight; H1: Gi ≠ 0 for at least one i; there is an effect of body weight; b) H0: β1 = 0, general slope is zero, non-signi cant regression; H1: β1 ≠ 0, the overall slope differs from zero, signi cant regression; c) H0: β2i = 0, the inclination of body weight i does not differ from the average inclination.
All analyses were performed with the aid of SAS software (Statistical Analysis System, version 9.0) considering a signi cance level of up to 5%.

Water quality
For body weights of 121, 235 and 596 g, the minimum and maximum water temperature values observed during the experimental tests were 25.3 ± 0.6, 25.8 ± 0.6, and 25.6 ± 0.3°C in the morning There was no mortality during the experimental periods, nor were any external pathological signs observed, even in animals fed with at the low-lysine level.
The animals submitted to the control diet (CD) were superior to those submitted to a diet with the rst level of lysine (0.225%) for all variables analyzed in the experiments (different body weights), with the exception of FI for body weights from 235 to 596 g. The lowest weight gains were observed in sh fed with the rst level of digestible lysine (0.225%).
In addition, the mean nal weight and weight gain of tambaqui were in uenced (P < 0.05) by the increasing digestible lysine levels in the diets for all body weights evaluated (Table 3), with the highest nal weight and weight gain obtained by tambaqui fed the highest digestible lysine level evaluated (0.901%). Regarding the body weights evaluated, the greatest weight gain was obtained by the 121 g animals, with a value of 26.06%. The values presented correspond to the mean (± standard error) of 25 sh for each body weight evaluated in each treatment; Means with different superscripts in the same column, for each experiment, differed by the Duncan test (P < 0.05).
Digestible lysine levels in uenced feed intake (P < 0.05) only for sh weighing 121 g, where there was an increase in diet intake up to 0.450% of digestible lysine. As expected, digestible lysine intake increased in all body weights groups evaluated (P < 0.05) because the diets presented increasing digestible lysine levels. The values presented correspond to the mean (± standard error) of 25 sh for each body weight evaluated in each treatment; Means with different letters in the same column, for each experiment, differed by the Duncan test (P < 0.05).
Maintenance requirements and e ciency of utilization of lysine

Discussion
In the comparison by means test, the animals submitted to the CD differed from all levels and presented similar responses to those fed with a diet containing 0.450% digestible lysine in all analyzed variables (Tables 3 and 4), with the exception of FI for body weights from 235 to 596 g, con rming that lysine was the rst amino acid limiting in experimental diets.
The highest weight gain was obtained for tambaqui fed 0.901% lysine in the diet, for all weight ranges evaluated, evidencing the essentiality of this amino acid for the species.
In addition, digestible lysine intake provided su cient intervals to obtain protein retention values and negative, near zero, and positive body lysine, allowing estimation of maintenance requirements without extrapolation.
Normally, in trials to determine the amino acid requirement for maintenance, animals are expected to lose weight because of the use of diets extremely de cient in the amino acid under study, especially at lower levels [8,23]. In the present study, this effect was observed only for tambaqui with a 235 g body weight, fed 0.255% level of digestible lysine in the diet. Similar effects were observed by [8] in studies to determine the requirement of maintenance of lysine and methionine for Nile tilapia (Oreochromis niloticus).
However, weight gain is not a good parameter for determining amino acid requirements for maintenance because sh gain weight through protein, fat, and mineral deposition [24], and there is no demand for amino acids to maintain lipid and mineral reserves. Thus, protein (nitrogen) or amino acid deposition under study is a more accurate indicator for achieving amino acid requirements for maintenance than weight gain [7].
Studies of requirements maintenance of digestible lysine for tambaqui are lacking in the literature. The values obtained in the present study, for body weights from 121 to 235 g (82.03 mg kg − 0.7 day − 1 ) and 596 g (106.85 mg kg − 0.7 day − 1 ), are above the recommendations described for other sh species already studied.
In studies with juvenile and adult Nile tilapia, tropical and omnivorous species, with initial mean weights of 20.70 and 165 g, values of 16.90 and 68 mg kg − 0.7 day − 1 of digestible lysine were obtained for maintenance, respectively [8]. For Atlantic salmon (Salmo salar), with initial mean weights of 62.80 and 95.70 g, the values obtained were of 21.70 and 18 mg kg − 0.7 day − 1 of digestible lysine, respectively [9,25]. Fish species may present differences in protein turnover [26], as well as differences in the metabolism of amino acid utilization for maintenance [3].
In addition, breeding programs with Nile tilapia and Atlantic salmon seek to maximize weight gain over a shorter period and improve feed e ciency [27,28], may have reduced the requirement of amino acids for the maintenance of these species compared to that tambaqui. Thus, this explains the higher requirement for maintenance lysine for tambaqui observed in the present study.
The requirement for maintenance lysine was positively correlated with tambaqui body weight, i.e., there was an increased requirement for maintenance lysine as the sh grew. This effect was also reported by [8], who observed that adult tilapia (165 g) had a higher demand for maintenance lysine compared to that of juveniles (20.7 g), with values of 68.80 and 16.90 mg kg − 0.7 day − 1 , respectively.
In young animals, amino acid requirements for body maintenance represent a small fraction of the total requirement, but this situation is reversed as the animal increases in size [29]. In the early stages of animal life, protein deposition occurs at a high rate compared to that of adult animals; therefore, caution is needed when determining the requirement for amino acids for maintenance early in life because the relationship between amino acids destined for maintenance and growth change rapidly with the growth rate of the animal [10].
The length of the experimental period is another factor that interferes with the variation of the maintenance values of an amino acid. Very long trial period (greater than 30 days) metabolic adaptation and changes in maintenance conditions can occur [10]. In this study, each experiment lasted 20 days, being enough to guarantee changes in the deposition of body protein content (retained nitrogen) and to determine the requirement to maintenance of lysine for tambaqui.
The e ciency of lysine utilization for sh has received great attention in the formulation of diets, especially when formulated with a protein of plant origin. The lysine utilization e ciency obtained for tambaqui was 55% for body weight 121 to 235 g and 40% for body weight 596 g, which indicated a negative relationship with tambaqui body weight. Thus, there was a reduction in the e ciency of use with the increasing body weight of tambaqui, this demonstrated the direct in uence of body weight on this variable.
The results obtained in this study indicated a reduction in amino acid utilization e ciency for body protein deposition and a simultaneous increase in amino acid utilization for replacement of inevitable protein losses (maintenance) as tambaqui grows. Moreover, in young animals, the growth rate and protein turnover are higher compared to that of older animals [10].
The results obtained in the present study corroborated those reported by [8] for Nile tilapia, where the e ciency of lysine utilization was 72 and 52% for animals with initial mean weight of 20.70 and 165 g, respectively. [12], with tilapia of different body weights fed the same diet, also observed a reduction in the e ciency of lysine utilization with increasing body weight, with values of 68, 63 and 47% for the mean body weights of 10, 58 and 248 g, respectively.
The low e ciency of lysine utilization by tambaqui (55 and 40%), compared to that of other species, may be related to its slow growth. Slow-growing sh probably use a higher proportion of dietary amino acids for body tissue maintenance than do fast-growing sh, which use amino acids for body protein deposition. Thus, this increases the requirement of lysine for maintenance, which is an amino acid used almost exclusively in protein synthesis.
However, because of the limited amount of information available in the literature on the e ciency of lysine utilization and that of other amino acids by different species, it is di cult to properly assess the in uence of species on amino acid utilization e ciency until there have been a su cient number of species studied.
In addition to the differences among species and body weights, other factors may in uence the e ciency of lysine utilization, such as dietary lysine concentration [25]. Low values of e ciency of utilization of lysine are obtained when using high concentrations of lysine in experimental diets. Probably due to the reduction in the rate of lysine absorption [34] or the need for excretion of excess amino acids in diets, which requires energy expense.
The source of protein used to formulate experimental diets also in uences the e ciency of lysine utilization [31,33], when a source of low biological value protein is used and ingested by sh, there may be a reduction in the body protein website due to the limitation of essential amino acids.
The values obtained in the present study for lysine maintenance requirements and utilization e ciency for tambaqui are crucial for understanding the basal metabolic needs of lysine, allowing the production of tambaqui at a minimal cost. Additionally, they will serve as beginning parameters for the elaboration of an effective factorial model, able to predict the lysine requirements for tambaqui under several environmental and/or physiological conditions, and will thereby serve as a guide for further studies.

Conclusions
Body weight is a variable that in uences the maintenance requirement and the utilization e ciency of lysine, and must be taken into account in mathematical models that estimate the lysine requirement by the factorial method.
The requirement for digestible lysine for tambaqui maintenance was 82.03 and 106.85 mg kg 0.7 day -1 for body weights 121 to 235 and 596 g, with 55 and 40% utilization e ciencies, respectively.  Figure 2 Caption found in gure.