A new approach for quantifying phosphorus requirement in an Amazonian fish using CT-scanning

Phosphorus (P) is an essential mineral for fish growth, as it plays pivotal roles in skeletal development and energy transfer reactions. However, the dietary requirement of this mineral is variable among fish species and the growth stages. Thus, this study aimed to determine the digestible P (dP) requirement for tambaqui in the initial growth stage (± 17 to 150g) using growth data, mineralization of the whole body, vertebrae and scales, as well as blood chemistry as response parameters. A total of 192 tambaqui juveniles of approximately 17 ± 0.85 g were stocked into a water recirculation system. Fish were assigned to 24 70L-tanks using randomized block design (two floors) with six treatments (1.3, 2.4, 4.8, 6.3, 7.8 and 8.8 g kg−1 dP) and four replicates. Fish were fed six semipurified diets with increasing levels of dP for 90 days. The dietary requirement of P was estimated using regression models (P < 0.05). Duncan and SNK multiple range tests were used when regression models were not fitted. No mortality or apparent signs of P deficiency were observed. All performance variables were improved with increasing levels of dP in the diet. Based on weight gain, the P requirement was 6.3g kg−1 diet while for increased carcass mineral deposition was 6.6g kg−1 diet and for adequate mineralization of vertebrae the requirement was 4.75 g dP kg−1 diet. The blood chemistry parameters were greatly affected by the dietary P level, except for serum calcium. Thus, the dietary dP requirement for tambaqui juveniles in the early stage was 6.3 g kg−1 diet based on growth and bone mineralization.


48
Phosphorus (P) is the first limiting nutrient for algae growth in freshwater 49 ecosystem. Additionally, its association with nitrogen overload could induce the 50 eutrophication of water bodies. Thus, P discharge by farmed fish might contribute to the 51 pollution of water reservoirs and reduce the sustainability of the aquaculture production.

52
In this context, Brazil is responsible for 6% of the worldwide phosphorus discharge to the 53 aquatic environment [1] which poses a significant impact for the expected growth in 54 aquaculture predicted for the following decades [2] once the soluble P removal from the 55 water is complex and still inefficient [3]. Therefore, finding ways to reduce P discharges 56 by aquaculture is a key point for the future of freshwater fish production [4] [3]. 57 One of the suggested ways to efficiently reduce the P discharge from fish farming 58 is by strictly adjusting the P intake to the specific requirement of a fish species in each 59 developmental stage once during fish growth the nutrient requirement changes according 60 to the physiological status and environmental changes. For instance, the P requirement 61 tends to reduce according to the growth stage mainly due to the limitation of bone tissue 62 growth as fish reaches the adult stage [5]. Thus, defining the minimum dietary P 63 requirement in different growth stages is one of the first steps to reduce the P discharge 64 in aquaculture production. 65 However, the P requirement in animals diverge depending on the response 66 parameters evaluated and the models used to estimate the requirement [6]. For instance, 67 P requirement based on weight gain are lower than those estimated based on bone 68 mineralization [7][8][9][10]. The use of P requirement estimates based on WG have led to 69 several bone deformities in salmonids imposing a threat to the welfare of fish and the 70 quality of the final product [10][11][12]. This has been used as role for most of the 4 71 aquacultured fish species, at least for Nile tilapia and carp aquaculture [8,13]. However, 72 limited data on other fish species are available and further extrapolations should be made 73 with care. 74 Tambaqui (Colossoma macropomum) is the second most farmed species in Brazil 75 and possess an important role on food security for Latin American countries [14]. 76 Additionally, the high P discharge in tambaqui aquaculture might be a serious problem 77 for the Amazonian aquatic ecosystem where most of the tambaqui farms are located. 78 Therefore, the precise P formulation in tambaqui diets is of utmost importance for 79 protecting this ecosystem. 80 Although the importance of this species for aquaculture, information on its 81 nutrient requirements are limited and, to the best of our knowledge, just one study have 82 reported the phosphorus requirement for tambaqui in the grow-out stage (~ 150g) [7]. 83 Therefore, we designed a trial to determine the P requirement for tambaqui juveniles 84 (from 15 to 150g) using different response parameters. Additionally, for the first time in 85 a requirement study, we provide evidence that the P requirement for whole-body 86 mineralization might be similar to the estimate based on weight gain while the 87 requirement for bone mineralization are markedly low in fish with dense mineralized 88 bones. supplementing the basal diet with potassium phosphate to achieve the following levels: 99 3.0, 4.5, 6.0, 7.5 and 9.0 g kg -1 ( Twenty fish were randomly sampled at the beginning of the trial for determining 145 the initial whole-body chemical composition and for calculation of the nutrient retention.

146
At the end of the trial (90 days), all fish were starved for 24h, weighed and tissues were 147 sampled. Two fish per tank were sampled at the end of the trial for whole-body chemical 148 composition.

149
Diets and fish samples were analysed in duplicates for chemical composition.

150
Moisture was determined in an oven at 105 ºC for 24h; protein (nitrogen * 6.25) was 151 determined using microKjeldahl method; total lipid was determined using a Soxhlet 152 extractor; and ash was determined by incineration in a furnace at 550 ºC for 5 h [16].

153
Eight fish per treatment were dissected for scales and vertebrae samplings.

154
Vertebrae were washed with deionized water to remove all remaining connective tissue, 155 then lipids were removed by extraction in a chloroform to methanol solution (1:1) for 6h.

156
The fat free vertebrae and scales were dried in an air oven at 65 ºC for 24h, ground and 157 prepared for acid digestion [18].

158
Mineral content of diets, whole-body, vertebrae and scales were determined after 159 digestion with a nitropercloric solution (3:2) for 3 h at 300 ºC. After digestion, the mineral 160 extracts were diluted in 50 mL deionized water and minerals were determined.

226
The analyzed total phosphorus levels in the diets were similar to the calculated 227 levels in the formulation (Table 1). Fish accepted well the semi-purified diets and showed 228 a normal feeding behavior throughout the experiment. All growth performance 229 parameters were affected by dietary phosphorus (P<0.05) ( Table 2). The maximum 230 growth rate was estimated to be at 6.3 dP g kg -1 diet. Fish fed diets containing dP levels 231 lower than those had a significant reduction on feed intake, specific growth rate and 232 impaired feed conversion ratio. Protein efficiency ratio and protein retention linearly 233 increased according to the dietary dP levels. On the other hand, dietary P was better 234 utilized by fish fed the lowest dP levels. The estimated requirement based on growth 235 parameters varied from 3.7 to 7.1 and different models best fitted to the parameters (Table   236 3). No mortalities were observed throughout the experiment. Whole-body lipid significantly decreased while the ash content linearly increased 252 in response to the dietary dP levels (Table 4). Regardless for Ca and P, dietary dP levels 253 did not affect the whole-body protein and moisture content (Table 4). P content in and 254 scales showed a quadratic effect in response to dP supplementation (Table 6). Estimated 255 requirement based on carcass gross nutrient composition was around 6.6g kg -1 diet, while 256 for vertebrae mineralization was between 4.75 and 6.75 g kg -1 diet (  Serum Pi levels were highest in fish fed diets supplemented with 4.8, 6.3 and 7.8 296 g kg -1 dP, while the lowest values were recorded for fish fed the two lowest dP levels 297 ( Fig. 2A). Serum albumin concentration linearly increased up to 3.18 g kg -1 and then 298 plateau (Fig. 3B), while the plateau value for ALB:GLOB ratio was 3.92 g kg -1 according 299 to the increase on dietary dP levels. Serum lipid profile of tambaqui was profoundly 300 affected by the dietary dP levels (Fig. 4). Generally, TAG and VLDL were higher in fish 301 fed the three lowest dP levels (Fig 4A and 4E, respectively), while LDL showed a linear 18 302 increase with a plateau at 4.76 (Fig. 4D). The highest total cholesterol level was recorded 303 in fish fed diets with the two highest dP levels (Fig. 1B). Similarly, the enzyme activity 304 was affected by dietary dP levels (Fig. 5). ALK activity linearly decreased up to 3.08 g 305 kg -1 and then plateau in response to dietary dP levels (Fig. 5A).  Tambaqui juveniles reached the maximum growth rate with 6.3 g kg -1 diet of dP.

337
This value is higher than a previous study with tambaqui that reported a dP requirement structures and the total bone volume was 2.90 cm 3 , while fish fed adequate P levels (6.3 371 g Kg-1 ) showed a great increase on total bone volume (13.34 cm 3 ). Thus, the assumption 372 that the requirement based on the bone mineralization is higher than those based on 21 373 growth might not be true for all fish species. Additionally, the CT-scanning could predict 374 the P requirement for bone mineralization using the vertebrae as the bone tissue.

375
In this view, the evaluation of bone density could be an important tool to assess 376 and/or properly interpret the data on P requirement studies, since bone mineralization is 377 reduced by P deficiency with a concurrent reduction on bone density [33]. However, few 378 studies have used this technique to evaluate P adequacy in fish and, to the best of our 379 knowledge, this is the first study to report the use of CT-scan to determine the mineral

395
The increase on fatty body deposition is a commonly reported effect of P 396 deficiency in fish [4, 7, 10, 13, 24-26, 30, 35]. In this study, we have similarly observed further studied in fish since the mechanism used to explain its action is based on higher

436
The use of serum Pi as a biomarker to assess the proper dietary P levels in fish has 437 been criticized by several studies because the serum P usually reflects the effect of a 438 recent meal rather than P status of a fish [7,35,42]. However, we found serum Pi to be a 439 responsive biomarker of P status in this study since its results correlates well with all the 440 other parameters used to assess the dietary P adequacy. This effect is supported by the 441 serum albumin levels, which showed the same trend as the Pi levels. Albumin is 442 responsible for the transport of several molecules through the body, phosphate included 443 [7]. Therefore, in P-deficient animals the low Pi levels are usually followed by reduced 24 444 albumin levels. This effect might have been prominent in our study because the fish were 445 fasted for 24h before the blood sampling, thus reducing the effect of the last meal.

446
Taking altogether, our study demonstrated that feeding tambaqui with 6.3 g kg -1 447 digestible P on the beginning of the growth stage is adequate based on several parameters 448 of growth and welfare. Additionally, we have demonstrated for the first time that CT-scan 449 might be an adequate and non-invasive method to assess the P requirement or bone health