Effect of dietary protein level on growth, food utilization, food conversion and survival rate of giant trevally (Caranx ignobilis) [version 1; peer review: 2 not approved]

Background: Proper feed formulation is required for successful fish farming activities. Therefore, it is necessary for fish feed to provide optimal growth so that the cultivation business generates profits. Currently, there is very limited information about the appropriate feed for Caranx ignobilis, causing problems with its development. This study aims to provide feed with different protein levels to C. ignobilis. Methods: We will examine the protein levels’ effects on the daily growth rate (DGR), specific growth rate (SGR), absolute growth rate (AGR), feed conversion ratio (FCR), feed efficiency (FE), and survival rate (SR). This research was conducted for 35 days, from June to October 2017, at the Center Brackiswater Aquaculture Development (BPBAP) Ujung Batee, Ministry of Marine Affairs and Fisheries, Aceh Besar, Indonesia. This study used a completely randomized design method, with five treatment levels (30%, 40%, 50%, 60%, and 70% protein feed) and four replications. Results: The results showed that feeding with different proteins on C. ignobilis had a significant effect on the mean values of DGR, SGR, AGR, FCR, FE, and SR. The 50% protein feed gave the best results for C. ignobilis, with a mean DGR value of 0.267 ± 0.005 g / day, a mean SGR of 1.722 ± 0.030% / day, a mean AGR of 0.081 ± 0.003 cm/day, a mean FCR of 1.290, a mean FE 77.755% and a mean SR was 86.667%. Conclusions: Furthermore, feed treatment with increased protein Open Peer Review


Introduction
Giant trevally (Caranx ignobilis) is a reef fish species with a high economic value 1 because it is in very high demand in domestic and international markets, with a price range of IDR 42,000-80,000 / kg 2,3 . Furthermore, C. ignobilis does not have muscles between muscles, so it is easy to consume 4 . Based on these advantages, C. ignobilis is exploited for consumption, economy, and recreation in various ways, including the use of destructive fishing gear (include e.g. bottom trawling, cyanide fishing, dynamite fishing, and ghost fishing) 5 . Massive coral damage due to environmental damage and the effects of global warming also puts pressure on the species population 6 . Therefore, it is necessary to conserve C. ignobilis through the domestication process so that numbers caught in nature can be reduced by diverting fishermen to C. ignobilis cultivation activities.
Fish that are eligible to become cultivated commodities must meet criteria such as economic value, desirability, fast growth, resistance to disease, and ease of domestication [7][8][9][10] . C. ignobilis has several advantages, such as being euryhaline so that it can live in a salinity range of 15-32 ppt 11 , it can live in conditions of high density(up to 150 individuals / m 2 12 ), fast growth 13 , economic value and no history of disease 9 . Based on these advantages, C. ignobilis can be cultivated in ponds and floating net cages andseveral fishermen in Aceh, Indonesia, have started cultivating this species in with this method 14 . However, there are several obstacles in developing cultivation, such as a lack of availability of commercial feed with the appropriate protein content. Therefore, we need a study on feed protein that can provide optimal growth in this fish species.
In cultivation activities, growth is a determining factor for success. The problem that is often faced by fish farming is the low value of protein, so cultivating fish is not optimal. According to Subandiyono and Hastuti 15 , protein has various roles and functions, including brave protein as a body structure, such as collagen, which is a fibrous connective tissue and has a solid structure to form fish muscles. Fish can grow optimally if protein needs are met. Protein has a role in determining the fish growth process because most of the fish's body (45-75% dry weight) consists of protein 16 . Also, protein is an efficient source of energy for aquatic animals, especially carnivorous fish 17 . We obtained 300 C. ignobilis juveniles from Lancang Barat Village, Dewantara District, Aceh Utara Regency. The fish were acclimatized for two weeks at the BBAP Ujung Batee, Aceh Besar Indonesia. They were placed in 20 rearing nets of 1 × 1 meter in size, which were placed in a concrete pond containing 40 tons of water (range of salinity value 29.19-29.44‰), each containing 15 C. ignobilis juveniles.

Feed preparation
The experimental feed was made from fish meal, rebon shrimp flour, rice flour, soybean flour, cornflour, blood meal, coconut oil, CaCO 3 , Isolesin, L-Tryptophan, DL-Methionine, and premix with a feed protein content of 30-70%. All ingredients were mixed and analyzed for protein content (Table 1).
Feeding procedure Juvenile C. ignobilis were selected randomly. Their average weight was 11.41 ± 0.40 g and their average initial total length was 7.70 ± 0.76 cm. Furthermore, the fish are distributed into 20 nets measuring 1 × 1 meter. The net is placed in a concrete pond with a water volume of 40 tons (15 fish/net). The fish were given experimental feed twice a day at 7 am and 5 pm for 35 days, as much as 3% per body weight.

Research parameters
The parameters measured in this study are DGR, specific growth rate (SGR), AGR, feed conversion ratio (FCR), FE, and survival (SR). DGR and SGR were analyzed based on the formula by Muchlisin et al. 18,19 . AGR was analyzed based on the Jones 20 formula, FCR, and FE were analyzed based on the formula of Shapawi et al. 21 ; SR was analyzed based on the formula of Hseu et al. 22 . Note: Wo is the initial weight of fish when researched (g); Wt is the weight of the fish at the end of the study (g); t is the study duration (days). L1 is the length of the fish at the start ofraising (cm), L2 is the length of the fish at the end of raising (cm), Δt is the length of rearing (days). Nt is the number of test fish at the end of maintenance, and No is the number of test fish at the beginning of rearing.

Data analysis
Data of DGR, SGR, AGR, FCR, FE, and SR were analyzed statistically by one way ANOVA (Analysis of Variance) with a 95% degree of confidence (P < 0.05) using SPSS version 20 software. Significant data were tested further with Duncan's multiple ranges test.

Discussion
The analysis showed that C. ignobilis growth occurred optimally with 50% protein feed based on the mean values of DGR, SGR, and AGR. Similar results were also found in its sister, C. melampygus; feed with about the same protein content (45%) provided optimum growth in these fish 23 . Furthermore,  this study revealed that the increased feed protein content had a positive correlation to fish growth, but this only occurred up to 50% protein feed, while fish growth would decrease if the feed protein content was >50%. This is probably because C. ignobilis is only able to absorb feed protein optimally by ≤ 50%. Similar results occurred in Oreochromis niloticus, which was given feed treatment with 17%, 30%, and 35% protein, which showed that the optimum growth results occurred in 30% protein feed treatment, but the 35% protein feed in O. niloticus had no significant effect on fish growth 24 . Furthermore, another study by Yang et al. 25 revealed that Bidyanus bidyanus, which was given feed treatment with protein levels of 13%, 19%, 25%, 31%, 37%, 43%, 49%, and 55%, showed that fish growth continued to increase in size from 13%-37% protein feed. However, fish growth was no longer significant in the range of 43%-55% protein feed. Other research results, also by Yang et al. 26 on Spinibarbus hollandi, which received protein feed of 13%, 19%, 25%, 31%, 37%, 43%, 49%, and 55%, showed that fish growth continued to increase from 13%-31% protein feed, but fish were no longer grew significantly in the 37%-55% protein feed. This review shows that the fish's ability to absorb feed protein differs by species and that fish have a limited ability to absorb protein in feed at certain periods and conditions.
Protein is the main macromolecular component fish need 27 . The function of protein is prioritized for the synthesis of new proteins according to the fish's nutritional needs. At the same time, carbohydrates and lipids supply energy availability for fish 28 . Feed protein will be hydrolyzed in the digestive tract, and only free amino acids can be absorbed by the intestine for tissue and organ purposes as body protein 29 . However, feed with high protein content may not be used optimally by fish as in this study. Feed with 60% and 70% protein given to C. ignobilis actually decreased the ability to grow, the feed efficiency (FE), and fish survival. In addition, the feed treatment with 60% and 70% protein given to C. ignobilis also decreased the average FE and SR values but increased the FCR value. High FCR will increase feed consumption so that fish farming production costs tend to increase as well. Furthermore, the SR values in protein feed treatment of 60% and 70% only reached 68.34% and 65.00%, probably because the ability to absorb feed (FE) decreases, so that the quantity of fish feces increases and causes the fish to die due to poisoning.
The mean optimum FE value of C. ignobilis reached 77.755%, which resulted in the 50% protein feed treatment. This result is better than the best average FE value of C. hippos, which reached 19.01% in the combination treatment of 50% fish waste and 50% artificial feed 30 . Furthermore, another study by Rostika et al. 31 also produced the best average FE value for Caranx sp. lower than this study with a value of 30.60% in the treatment of 100% trash fish feed. The different results may be due to different test fish species and feed treatments, so that the fish's ability to absorb feed nutrients is also different. Another factor that plays a role in FE is the biological value contained in feed protein. Protein in feed with high biological value will stimulate the accumulation of body protein greater than that of low biological value 32-34 . Protein is a nutrient required in large quantities in fish feed formulations because it needs to be given continuously with sufficient quality and quantity. Buwono 35 states that, biologically, the quality of artificial feed shows the nutritional value of the protein contained in the feed. Thus, the feed quality is assumed to be protein quality. The feed protein quality depends mainly on its essential amino acid content; the lower the essential amino acid content, the lower the protein quality 36-39 .
The best SR value in this study was shown in the 50% protein feed treatment with a mean value of 86.67%. This result is lower than the studies conducted by Rostika et al. 31 and Rostika et al. 30 regarding the combination of pelleted feed and trash fish to Caranx sp. and C. hippos, which obtained an SR value of 100% in each treatment. Furthermore, the mean SR value of C. melampygus in the study of Suprayudi et al. 23 is also higher than this study, with a value reaching 100% with the best treatment of 45% protein feed. Furthermore, another study by Rombenso et al. 40 showed that C. latus, C. crysos, and  C. hippos reared in floating net cages near the coast of Brazil by feeding sardines produced an average SR value of 90%. Therefore, further research is needed to increase the SR value of C. ignobilis in the future so that it can increase fish production.
The best FCR value produced in this study was in the 50% protein feed treatment with a mean value of 1.29. The lowest value was shown was in the treatment of 70% protein feed (mean FCR 1.96). This result is better than the research conducted by Rombenso et al. 40

Mohammad Bodrul Munir
Aquatic Resource Science and Management, University Malaysia Sarawak, Kota Samarahan, Sarawak, 94300, Malaysia The MS "Effect of dietary protein level on growth, food utilization, food conversion and survival rate of giant trevally (Caranx ignobilis)" is not written well. For example, writing style of the background in the abstract was found to have grammatical errors. The methods in this section are expressed as future tense.
There was no feed proximate analysis found; however it is necessary to validate the desired protein percentage.
How many biological and technical replicates did you take?
Superscripts in Table 2 indicated the significance differences among the rows; but the researchers did not express this, therefore it is quiet difficult to understand. Please revise it.
In the Table 2 Expression of superscripts in Table 3 have the same problem as Table 2, please re-write. I felt confused the FCR data. Was there any relation of 3% body weight feed provided to fish? How did you calculate this 3% body weight?
Overall not at the standard for indexing using the present format.

Is the study design appropriate and is the work technically sound? No
How did the authors produce the feed? Home-made or produced by manufacturers? ○ Where did the authors get the ingredients? Home-made or purchased? ○ Do the fish feed pellet float or sink? Does the diet fit the preference of fish?

○
The design seems like the authors were testing the suitability of blood meal (I prefer "meal" instead of "flour") rather than testing the effects of protein levels. I did a quick check. There are many studies to replace fishmeal with blood meal. For omnivorous fish, 50% replacement is suggested ( The diets were tested using juveniles and that should be reflected on the title.

5.
The authors suggested fish mortality of groups D and E were related to feces accumulation and poisoning. However, the authors didn't mention the depth of the experimental pond. In the pond used in the experiment, the authors suggested nets were used. Could the fish feed accumulate inside the cage and kill the fish because of that? Did the fish show any sign of intoxication? Also, is that pond equipped with any aerators? Is there any water treatment facility?
6. Amino acids and proximate compositions of the diets: The authors calculated the protein content of fish feeds. Did the authors measure the exact protein concentration? Would it be possible that the high level of blood meal resulted in inferior growth, because of insufficient amino acid(s)? In addition to protein, other proximate compositions i.e. lipid, ash, moisture and carbohydrate contents should also be measured and presented.