Effects of Different Carbon Sources on the Growth and Production of Rotifer ( Brachionus plicatilis ) in a Zero-Water Exchange Bio ﬂ oc Culture System

Brachionus plicatilis is considered an indispensable ﬁ rst live feed for many ﬁ sh and crustacean larvae; the demand for the species has increased globally. The mass production of the rotifer involves quality microalga and a standard diet; this culture is expensive and needs a skilled workforce. The hatchery ’ s incubators are likely to have limited resources leading to sudden rotifer culture crashes that ultimately disrupt the larvae production. More recently, improved sustainable rotifer production has been achieved through bio ﬂ oc technology (BFT) that uses ﬁ sh wastes and wheat ﬂ our. However, various carbon sources, which are typically used in BFT-based systems need to be explored and tested for their ef ﬁ cacies. A 4-day rotifer, B . plicatilis batch culture, was conducted in BFT systems by adding four carbon sources: molasses, rice bran, maize starch, and palm kernel expeller versus a control (without any carbon source). Fifteen 125L containing polyethylene tanks with a water volume of 100 L were used for this experiment, and each tank was stocked with 5 × 10 6 rotifer (50 rotifers mL − 1 ). Different carbon sources in triplicates including a control were tested as treatments. The carbon: nitrogen ratio in the study was maintained at 10 :1. The rotifers were fed with Baker ’ s yeast at 1.0, 0.50, and 0.25g million - − 1 rotifers for the ﬁ rst, second, and third day and continued after that. Total ammonia – nitrogen (TAN) and pH values were found to be signi ﬁ cantly ( p < 0 : 05) lower in all four treatments of the BFT system than in the control. Signi ﬁ cantly higher ( p < 0 : 05) settleable solids were obtained in the molasses and rice bran treatments than those in the maize starch or palm kernel expeller. Likewise, the signi ﬁ cantly ( p < 0 : 05) higher density of B . plicatilis and their speci ﬁ c growth rate were obtained in the molasses and rice bran-adding treatments, followed by those in palm kernel expeller, maize starch, and the control. This study indicates that molasses and rice bran as carbon sources when added to BFT-based systems enhance B . plicatilis production.


Introduction
Rotifera is one of the smallest metazoa, comprising over 2,300 species that consist of a head, including a corona, a trunk that contains the organs, and the foot [1,2].They are planktonic organisms, typically free-swimming (but sedentary and colonial forms are also known) and wheel-bearing animalcules, with from 50 to 2,000 µm; ubiquitously distributed throughout the world [2,3].The rotifers are unsegmented, pseudocoelomate, bilaterally symmetrical invertebrates, most of which are found in freshwaters; some of these do inhabit brackish and marine waters, and limnoterrestrial habitats, for example, mosses, lichens, liverworts [1][2][3].They are filter feeders, eating various foods, including dead organic material, algae, and other microscopic living organisms, multiplying by parthenogenesis and sexual reproduction under distress conditions [1,2].
A production system using biofloc technology (BFT) is environmental friendly as it maintains good water quality by removing nitrogenous metabolites and provides an additional food source in the form of an added organic carbon source [4][5][6][7][8].The BFT-based systems are often zero-exchange systems, which congregate potentially consumable "biofloc," composed of various microorganisms, such as bacteria, fungi, microalgae, and zooplankton [5][6][7].Various zooplankton communities, such as ciliates, flagellates, rotifers, copepods, and nematodes, naturally occur in a BFT system.The rotifers are second dominant species after ciliates, including a variety of genera, for example, Brachionus, Euchlanis, Lecane, Colurella, Anuropsis, Gastropus, Habrotrocha, and Philodina, that are usually preyed upon by the cultured larvae [5][6][7].The rotifers also contain an essential portion of microorganism biomass in a BFT system or an unique ecosystem [2,9,10].The biochemical composition of bioflocs includes protein, lipids, carbohydrates, amino acids, fatty acids, and antioxidants reflected by the presence of various microorganisms [4,11,12].The biofloc composition is known to result in higher growth, production, survival, and spawning rates, besides increasing the immunity of the cultural species [7,8].
To create a BFT system, the suitable carbon source with an optimal nitrogen ratio is one of the prerequisites (C : N 10-20 : 1).For instance, the different simple and complex carbon sources have been used in BFTs for many species of fish, prawns, shrimps, crayfish, and live feeds.Various carbon sources, such as acetate, corn starch, glycerol, molasses, etc., have typically been added [5,13].These simple carbon sources facilitate the removal of nitrogenous toxicants faster than complex ones.At the same time, some of them for example, glucose and molasses, are known to suppress Vibrio's, known to increase the total bacteria [13,14].Among the complex carbon sources, rice bran is more widely administered than others in microbial-based systems, the addition of 24 hr of fermented rice bran is more effective to create high-quality bioflocs and it leads to a higher animal production [5,12,13,15].
The mass rotifer culture depends on supplementing highquality unicellular microalgae and/dried algae, yeast, enriched diets [31][32][33].For example, a defatted microalgal meal (Haematococcus pluvialis) has significantly enhanced the growth and reproduction performance of B. plicatilis and its total carotenoid and astaxanthin contents [34].Furthermore, adding high-quality unicellular microalgae in the rotifer production system has enhanced rotifer population density, along with fatty acid contents, including EPA (eicosapentaenoic acid) and DHA (docosahexaenoic acid) [33,35].On the other hand, dried algae are seldom used to rotifer production systems in hatcheries due to their commercial unavailability and being expensive [1] and resulting in reduced population growth [36][37][38].The supplementation of commercially available yeasts can maintain good growth of B. plicatilis, which is also considered cheaper; however, yeast-fed rotifers are often deficient in highly unsaturated fatty acids (HUFA) [35].Meanwhile, HUFA is essential for marine fish and crustacean larval nutrition, growth, survival, and overall health status [26,39].High-quality B. plicatilis mass production can be achieved using a recirculation aquaculture system (RAS) to supplement highly mechanized artificial diets requiring a skilled workforce [40].
A laboratory-grown bacterial supplementation including Pseudomonas, Moraxella, Micrococcus, Alteromonas, unidentified gram-negative strain-B3 to the axenic culture of B. plicatilis has increased its reproduction, leading to increasing rotifer numbers and resting eggs [41][42][43].Similarly, the addition of probiotic bacteria with microalgae (Chlorella/Nannochloropsis oculata) alone, a combination of microalgae and yeast, and or an artificial diet with yeast have improved B. plicatilis production [44][45][46] and suppressed Vibrio's [45,47].Additionally, the waste-grown phototrophic bacteria were shown to increase rotifer growth and biomass [48,49].However, despite this progress, each supplementation can have different drawbacks, including disrupting rotifer production in hatcheries and a high demand on resources.Moreover, continuous rotifer mass culture requires a vast amount of high-quality microalgae, a heavy burden for many hatcheries [1,36].A microbial or BFT-based system could be an alternative technology as it proliferates microorganisms that are composed of natural probiotic bacteria, flagellates, ciliates, etc., which are likely ingested by B. plicatilis further leading to higher rotifer production [15,50].
Several studies have examined the potential application of the BFT to rotifer culture.For example, tilapia BFT system grown floc associated bacteria was sieved using 20 μm net and fed to B. angularis, in which the rotifer densities were 18.07 and 15.97 Ind. mL −1 for 21 and 25°C on day 60, respectively [50].Adding fish waste and wheat flour has significantly enhanced B. rotundiformis density (1,188 Ind. mL −1 , on day 13) than those fed fish wastes or microalgae C. vulgaris only [51].Panigrahi et al. [15] cultured B. plicatilis using BFT based on the aquamimicry and added microalgae Nannochloropsis sp. and filtered fermented juice when rotifers reached a minimum density of 150 Ind. mL −1 .However, no studies have assessed the effect of different carbon sources on the growth and production of a rotifer culture in a BFT system.Thus, this study aimed to evaluate the effect of four different carbon sources such as molasses (MO), rice bran (RB), maize starch (MS), and palm kernel expeller (PKE) on B. plicatilis mass production in a zerowater exchange BFT system.Dickson, Negeri Sembilan, Malaysia.The stock 25‰ saline water was prepared in a 2.5-tonne fiberglass tank by adding 5 ppm chlorine, and then, vigorous aeration was used to bubble off the chlorine before use [4].Approximately 25 million rotifers, B. plicatilis, were acquired from the I-AQUAS hatchery.These rotifers were stocked at 50 Ind.mL −1 in the 1-tonne tank with a water volume of 500 L. The rotifers were fed with Baker's yeast at 1.0, 0.50, and 0.25 g million −1 rotifers for the first, second, and third days, respectively [38;, 52]; these were added twice a day at 9:00 and 18:00 hr.On day 4, the rotifers were harvested for experimental use.Additionally, 50 L of N. oculata (approximately 2 × 10 6 cell mL −1 ) was added daily.The microalgae were cultured with fertilizers at a rate of 1 mL L −1 daily (Table 1), using 30 L containers with 25 L water volume, aerated using two airstones (size: 15 mm × 42 mm).

Preparation of Biofloc Inoculum and Study Protocols.
The biofloc inoculum was prepared by modifying the methodology of Hapsari [53], and Hosain et al. [5].To prepare four carbon-sourced biofloc inoculums, four 7.5 L containing cylindrical polyethylene buckets were used.Each bucket was filled with 5 L treated saline water at 25%.Each inoculum was added with 0.1 g yeast, 50 mg ammonium sulfate, 2.1 g carbon source, and 0.1 g commercial probiotics (total bacteria ≥ 1.0 × 10 9 CFU g −1 , PondPlus ® , Novozymes).The mixture of these inoculums was vigorously aerated for 24 hr using an airstone.Then, ammonia-N was measured, and it was not detected.After that, each inoculum (1 L) was added to respective carbon source treatment tanks only once at the beginning of this study.
This study was conducted in an enclosed greenhouse room at I-AQUAS, UPM, for a 4-day mass rotifer culture.There was a total of five treatments: control (transparent water system and no addition of carbon source) and four different tested carbon sources; MO, RB, MS, and PKE.Livestock feed-grade MO was dried at 60°C in an oven before the using it in the experimental.Commercially available MS (Cap Bintang corn starch, Korea) was added.Raw RB and PKE were purchased from a livestock feed ingredient store in Serdang, Selangor; these were then warmed up for an hour at 50°C using an oven before grinding.After that, these were hammer milled to a fine powder and passed through a sieve of 200 μm mesh size.Before rotifer culture, 15 tanks were filled with 100 L treated saline water at 25%.The experimental tanks were set up with a completely randomized design for five treatments, each of which was triplicated.Each experimental tank was stocked at approximately 5 million rotifers (50 Ind. mL −1 ).The average length (AESE) of each rotifer was 167.40 AE 2.12 µm.
The rotifers were fed with Baker's yeast at 1.00, 0.50, and 0.25 g million −1 for the first, second, and third day and continued after that [38,52].Carbon sources were added daily to maintain the carbon-to-nitrogen ratio at 10 : 1 [54].Schryver et al. [55] calculated the daily additional quantity of carbon sources, where approximately 50% of the most organic carbon sources on a dry matter basis was carbon.The daily amount of nitrogen was determined based on the daily calculation of yeast adding and its total protein level (yeast contained 40% protein, USDA Nutrient Database).The daily yeast whole protein nitrogen level was calculated [56], where approximately 16% of the protein was nitrogen.According to the aforementioned empirical methodology, the daily carbon sources from each carbon source tank were adjusted and weighted.Each estimated carbon source and yeast were mixed with 200 mL water using an airstone in a 250 mL glass beaker.These mixtures were added to each carbon source tank at 10:00 and 18:00 hr.This was done to help ensure that none of the carbon sources and yeast would clump in the culture tanks.

Water Quality Variables, Settleable Solids, Rotifer
Populations, and Size.Water quality parameters such as temperature, pH, salinity, and dissolved oxygen (DO) were determined daily at 09:00 hr in each tank with a multiparameter meter (YSI Model 556, YSI Incorporated, Yellow Springs, Ohio, USA).Total ammonia-nitrogen (TAN), nitrite-nitrogen (NO 2-N), and nitrate-nitrogen (NO 3-N) were measured using API ® commercial test kits (API ® Aquarium Pharmaceuticals, North America) daily.The settleable solids (mL L −1 ) were estimated with an Imhoff cone daily, according to Romano et al. [57].The Imhoff cone was filled with 1 L of culture water and allowed 30 min to settle the solids; after that, the volume was recorded as settleable solids.A 50 mL rotifer sample was collected from each tank daily.The daily rotifer population abundance was calculated using Sedgwick Rafter and profile projector with a 1-mL sample from each tank in triplicate.Based on this calculation, the quantity of daily yeast and carbon sources was estimated, as described previously.On day 4, the average size of rotifers was measured from each tank, and 30 rotifers were randomly measured using a compound microscope (Olympus Model BX41-CCD) and a USB digital microscope camera facility with a set magnification (10x).The specific growth rate of rotifer culture was estimated following the equation used by Suantika et al. [40]: Specific growth rate ðSGRÞ : ln where N t = rotifer density after a culture period t (individuals mL −1 ), N 0 = initial rotifer density (individuals mL −1 ), and t = culture period (day).

Results
3.1.Water Quality Variables and Settleable Solids.Temperature did not differ significantly among the five treatments during 4 days of this study (Table 2 and Figure 1(a)).pH values in the four other carbon source biofloc groups were significantly (p<0:05) decreased from day 2 to day 4 than control group (Table 2 and Figure 1(b)).Similarly, DO significantly (p<0:05) decreased at day 2 to day 4 in the four different carbon sources than the control (Table 2 and Figure 1(c)).TAN concentration was significantly (p<0:05) increased over the period in the control, while this was not substantially increased during the study among four carbon source biofloc groups (Table 2 and Figure 1(d)).Nitrite-N concentrations was similar among the five different treatments (Table 2 and Figure 1(e)).Significantly (p<0:05) higher Nitrate-N level was detected in the MO biofloc group than the control, which was similar to other carbon sourced bioflocs groups (Table 2 and Figure 1(f)).Settleable solids was significantly higher in the MO and RB biofloc groups than those in the control and MS biofloc group (Table 1 and Figure 2).

Rotifer B. plicatilis Density, Specific Growth Rate and Size.
Significantly (p>0:05) higher B. plicatilis density and specific growth rate were in the MO and RB biofloc treatments, followed by PKE, MS, and control, respectively (Table 3 and Figure 3).Moreover, B. plicatilis density was significantly higher in the MO, RB, and PKE adding biofloc groups at the day 2, 3, and 4 when compared to control and MS biofloc group (Figure 3).In terms of B. plicatilis length, it was significantly more prominent in the control and MS adding biofloc group than those of RB and palm kernel adding bioflocs treatments (Table 3).

Discussion
Developing a zero-exchange system with BFT provides microbial feeds associated with small particles and ensures good water quality conditions for cultured animals [5,8,13].A BFT-based system can remove nitrogenous toxicants and improve the overall health status of cultured animals [13,58].For instance, in this study, four carbon source-based BFT systems have ensured a better water quality status than control over the entire experimental period.In this study, TAN concentration was within suitable limits in the four carbon sources-based biofloc systems, which were more than four times lower than the control.Furthermore, the TAN concentration in the control group gradually increased over the period; in contrast, the TAN spikes decreased on days 3 and 4 in all four carbon-sourced biofloc systems than on day 2.These indicate that heterotrophic bacteria require time to establish.However, during the 4-day study, TAN increased in the control group, and inadequate heterotrophic bacterial numbers could not remove TAN due to the absence of any carbon source.On the contrary, there was two times higher nitrite-N and more than four times higher nitrate-N concentrations in the four groups of the biofloc system than in the control.These indicate that adding four carbon sources inoculums has rapidly facilitated the establishment of heterotrophic bacteria, which has contributed to removing TAN and nitrite-N and accumulating nitrate-N as an end-product [4,5,59].
In a biofloc system, the key strategy is the proliferation of microbial aggregation "biofloc" [7].This biofloc includes bacteria, plankton, and even inert or dead particles, estimated as settleable or total suspended solids.In this study, the settleable solids were significantly higher in the RB and MO carbon source BFT groups over the periods than in the PKE, MS, or control group.This higher settleable solid in the RB and MO carbon source BFT groups are likely supported to produce the higher number of rotifer B. plicatilis grown in those treatments, indicating that MO and RB suit to B. plicatilis production in biofloc systems.
Furthermore, these carbon sources have provided a favorable culture environment that might include rotifers, feeds, or other facilities.For example, Pekkoh et al. [60] stated that the addition of MO and RB mixture in a BFT system had  ), (d) ammonia-nitrogen (mg L −1 ), (e) nitrite-nitrogen (mg L −1 ), and (f ) nitrate-nitrogen (mg L −1 ) in control and four different carbon sources using biofloc systems for the culture of Brachionus plicatilis.CON: control; MO: molasses; RB: rice bran; MS: maize starch; and PKE: palm kernel expeller.enhanced the bacterial-algal performance that provides the most compact biofloc structure and better settleable solids, including larger biofloc particle/substrates.For instance, in this study, the addition of MO and RB with yeast and commercial probiotics facilitated in increasing biofloc volume/ settleable solids or provided the excellent consumable bacteria-algae-yeast substrates that have been grazing by rotifers and resulted in the higher number of B. plicatilis in these two treatments.The rotifer B. plicatilis is notably known to feed on bacteria, algae, yeast, and even smaller sizes of inert particles; these contribute to the higher production of rotifers in an optimal culture condition.
In this study, the higher B. plicatilis densities of 1,058.67,886.34, and 584.0 in Ind. mL −1 were found after the 4-day culture in the MO, RB, and PKE carbon sourced biofloc systems, respectively.The specific growth rates of 0.76, 0.72, and 0.61 were obtained for MO, RB, and PKE carbon sourced biofloc system.Compared to well-established rotifer culture studies, these three carbon sources can be considered a viable B. plicatilis batch production technique.For example, the maximal density of B. plicatilis was found to be 1,162 Ind. mL −1 on day 5 when rotifers were fed with frozen Nannochloropsis (Japanese) at the rate of 1.75 g (dry weight) to 10 6 rotifers per day [61].Green algae (Chlorella saccharophila) fed B. plicatilis showed a higher density and specific growth rate of 324 Ind. mL −1 and 0.19, respectively.It decreased in the order of Isochrysis galbana (191 Ind. mL −1 , 0.14) > Tetraselmis suecica (168 Ind. mL −1 , 0.14) > Saccharomyces cerevisiae (150 Ind. mL −1 , 0.13) > Thalassiosira pseudonana (104 Ind. mL −1 , 0.11) [62].Density and specific growth rate (257.6 Ind. mL −1 and 0.29) of B. plicatilis were maintained in a semicontinuous rotifer mass culture, in which rotifers were fed with Selco Sparkles (INVE S.A., INVE, Ghent, Belgium) using an automatic feeder [15].More recently, Bhosale and Mugale [63] have maintained a   [30] reported in a review that supplemental feeding of Baker's yeast to rotifer culture could be considered a feasible and easy technique, in which rotifer densities can be reached at 150 Ind. mL −1 , but under specific culture conditions, it could go up to 1,500-2,000 Ind. mL −1 .
Typically, the total length of B. plicatilis complex fluctuates between 100 and 400 μm [64], while neonate females do increase their length from birth to adulthood [65,66].In this study, the average length of B. plicatilis was 172.07 µm that was found in the maize starch that was approximately 3.53% longer than the 165.98 µm in rice bran biofloc group.The study exhibited that the increase in rotifer density was likely related to decrease in length in the RB, PKE, and MO biofloc system than the control and MS groups.It appears that the small length of B. plicatilis at day 4 in the RB, PKE, and MO was due to the abundance of neonate females that doubled within a day, and its density was 11 and 9 times higher in the MO and RB than the control.Various sizes of rotifers Brachionus sp. were used in incubators for fish and crustacean larvae culture, 130-340 μm and 150-250 μm were in Chlorella using hatcheries, 150-350 μm in Baker's yeast using hatcheries [25]; approximate body size 150-250 μm was in N. oculata in hatcheries [67].For instance, in this study, B. plicatilis cultivated in the MO (147-193 μm), RB (148-193 μm), and PKE (146-193 μm) would be an excellent live feed for farmed marine fish such as neon goby Elacatinus figaro, torktail blenny Meiacanthus atrodorsalis, false clownfish Amphiprion ocellaris, and spotted seahorse Hippocampus kuda with their suit mouth gape around 350, 307, 300, and 260 μm, respectively [68].On the other hand, enriched rotifers are often used in hatcheries according to the nutritional requirements of larvae [25][26][27][28],.However, more studies should be warranted using biofloc-based systems produced by rotifers with enrichments of larval nutritional requirements and their subsequent effects on larval zootechnical performance and overall health status.
Currently, the high density of rotifer culture strategy has been practiced field to meet the higher demand at fish, shrimp, and crab hatcheries in developed countries such as Japan, Canada, Taiwan, Turkey, and USA [32,[69][70][71][72].These high-density or ultrahigh-density rotifer mass culture systems are supplemented with the concentrated microalgal paste (Chlorella and Nannochloropsis) and artificial rotifer diets [35,72,73].These rotifer mass culture systems are usually shown to have increased nitrogenous toxicants and decreased pH values [35,73].In a higher density rotifer mass production in a batch culture or recirculation aquaculture system, B. plicatilis was stocked at 250 Ind. mL −1 [73].The authors recorded the maximum rotifer density and growth rate of 860 Ind. mL −1 and 0.31 for a batch culture after 4 days, 8,000 Ind. mL −1 , and 0.35 for an RAS with a 500% daily recirculation rate after 8 days.From this standpoint, mass culture of rotifer in a biofloc system could be an excellent option because in the current study, the growth rate of B. plicatilis was two times higher at 0.76 and 0.72 in the MO and RB carbon source using biofloc system after the 4 days of culture.However, more studies are warranted to optimize the batch and semicontinuous high or ultrahigh-density mass culture of B. plicatilis.Additionally, these studies should examine the optimal carbon-to-nitrogen ratio, the nutritional values of B. plicatilis cultured in a BFT system, and their subsequent effect on fish, shrimp, and crab larvae or post larvae.

Conclusion
The results obtained from the biofloc system could be considered an improved mass-production technology of rotifer (B.plicatilis) using the readily available and cheaper carbon sources of MO or RB.However, further research is recommended to optimize the rotifer production in biofloc systems and their subsequent effect on fish, shrimp, crab larvae, post larvae, or as an inoculum in a biofloc-based system.

2. 2 .
Stock B. plicatilis Culture.This experiment was carried out at the International Institute of Aquaculture and Aquatic Sciences (I-AQUAS), Universiti Putra Malaysia (UPM), Port

TABLE 1 :
[11]a composed of the fertilizers and chemicals was administered to an unicellular marine microalgae Nannochloropsis oculata culture tank; this microalgae was added daily during B. plicatilis stock culture[11].Statistical analysis was done using the computerized SPSS version 25.Before analysis, data were checked for normality and homogeneity of variances using Levene's test.Differences in water quality parameters and the rotifer population density, SGR, and rotifer length data of five treatments were performed using analysis of variance (ANOVA).When significant differences were detected (p<0:05), Tukey's multicomparison test was used for post hoc compare the mean among different treatment groups.

TABLE 2 :
Water quality parameters in control and the various carbon sources using biofloc treatments during a 4-day rotifer, B. plicatilis culture.Mean values with the same superscript letters in the same row were not significantly different (p>0:05), while different letters did significantly differ (p<0:05).CON: control; MO: molasses; RB: rice bran; MS: maize starch; and PKE: palm kernel expeller.

TABLE 3 :
The rotifer B. plicatilis density, specific growth rate (SGR), and length in control and various carbon sources using biofloc treatments after 4 days of culture.Mean values with the same superscript letters in the same row were not significantly different (p >0:05), while different letters did significantly differ (p <0:05).Daily mean (AESE) population density (Ind.mL −1 ) in control and different carbon sources using biofloc systems for the culture of Brachionus plicatilis.CON: control; MO: molasses; RB: rice bran; MS: maize starch; and PKE: palm kernel expeller.Aquaculture Research rotifer (B.calyciflorus) density at 381 Ind. mL −1 in the RAS, wherein the rotifers were fed with C. vulgaris, and 30% of them were harvested daily.Lubzens et al.