Reducing Carbon Input Improved the Diversity of Bacterial Community in Large-Scale Biofloc Shrimp Culture Facilities

Abstract

In this study, a group of large-scale biofloc shrimp culturing facilities were designed. The bioflocs were domesticated by continuously reducing the ratio of carbon to feed. The bacterial community diversity on the 20th, 40th, 60th, and 80th days was analyzed by high-throughput sequencing technology. The results showed that the biofloc content (0~16.8 ± 4.3) mL/L, TSS concentration (0~247.46 ± 27.3) mL/L, total ammonia nitrogen concentration (0~0.28 ± 0.052) mg/L, nitrite nitrogen concentration (0~4.13 ± 1.42) mg/L, nitrate nitrogen concentration (108.57 ± 19.6) mg/L were all within the safe concentration range of Litopenaeus vannamei. With the progress of reducing carbon input, the Chao1 index, the number of operational taxonomic units, and the Shannon index increased significantly. The number of OTUs of B80 (572.36 ± 13.26) was significantly higher than that of B60 (489.69 ± 12.97), B40 (423.35 ± 18.46) and B20 (407.67 ± 15.65) (p < 0.05). The Chao1 index of B80 (768.58 ± 36.96) was significantly higher than that of B60 (646.8 ± 52.53), B40 (569.7 ± 46.53) and B20 (516.3 ± 21.35) (p < 0.05). The Shannon index of B80 (5.63 ± 0.16) was higher than that of B60 (4.85 ± 0.13), B40 (4.68 ± 0.21) and B20 (3.65 ± 0.22), with significant difference (p < 0.05). At the end of the experiment, the domestication formed a micro-ecosystem with Proteobacteria as the carrier (46.98 ± 15.82%), Chloroflexi as the skeleton (2.2 ± 0.36%), Nitrospirae (1.35 ± 0.26%) as the main water treatment functional bacteria, and other bacteria as auxiliary nitrogen and phosphorus removal; At the genus level, unclassified_f_Rhodobacteracea (22.97 ± 3.82%), Ruegeria (10.35 ± 1.26%), Muricauda (5.73 ± 0.61%), Algoriphagus (3.75% ± 0.85%) and Nitrospira (1.56 ± 0.56%) are the dominant bacteria. Under the synergistic effect of the above bacteria, the biofloc system remains relatively stable. The survival rate and unit yield of shrimp were (65.32 ± 6.85)% and (4.15 ± 1.58) kg/m³, respectively.

1. Introduction

China is the largest shrimp farming country in the world, and its shrimp farming output reached 2.018 million tons in 2020 [1]. The corruption and decomposition of shrimp bait, feces, and corpses result in the deterioration of water quality and shrimp disease [2]. The traditional culturing mode mainly dilutes the concentration of harmful nitrogen pollutants through large-scale water change, which discharges a large amount of culturing wastewater into the surrounding environment, increasing energy consumption, culturing water consumption, and culturing costs. In addition, this treatment leads to the continuous accumulation of waste in the environment, which seriously damage the structure and function of the offshore marine ecosystem [3].

The zero-exchange culturing system based on biofloc technology and biosafety can solve the above problems [4]. By adding carbon sources to the aquaculture water, the biofloc culture system can regulate ammonia nitrogen concentration under the condition of a very low water change or even no exchange. Biofloc can convert ammonia nitrogen into bacterial protein, purify water quality, reduce bait coefficient, disturb the quorum sensing of pathogenic bacteria, cut off the transmission path of the virus, produce antagonists, and inactivate toxic signal molecules [5,6,7,8]. At present, biofloc technology has been widely used in the culturing of Oreochromis niloticus, Macrobrachium rosenbergii, Ctenopharyngodon idella, and Aristichthys nobilis [9,10,11,12].

Nitrifying biofloc technology is a high-order technology that cultivates nitrogen oxidizing bacteria through domestication, continuously improves the proportion of nitrogen oxidizing bacteria, and stimulates the growth of nitrifying bacteria. Compared with the traditional biofloc technology, it has the advantages of lower culturing cost, less oxygen consumption, low decarburization pressure, and so on. In 2011, Ray et al. [13] first proposed a chemoautotrophic shrimp biofloc culture system and described the differences in chemical kinetics between it and heterotrophic bioflocs. In 2014, Ray et al. [14] carried out a comparative experiment on the culture of Litopenaeus vannamei with four different types of bioflocs: chemical autotrophy, sucrose heterotrophic, molasses heterotrophic, and glycerol heterotrophic. The results showed that the 5-day biochemical oxygen demand (BOD₅) and total suspended solids concentration of the chemical autotrophic group were significantly lower than those of the three heterotrophic groups, and the average concentration of nitrate nitrogen was 162 mg/L, significantly higher than other treatments. In 2017, Tan [15] successfully transformed the heterogenous bioflocs into nitrifying flocs in the shrimp farming system by continuously reducing the input of organic carbon sources; Ferreira et al. [16] reported that toxic nitrogen compounds (ammonia and nitrite) were kept at a low level without adding carbohydrates through nitrifying bioflocs.

The biofloc shrimp farming systems of the above scholars have achieved success, but they were limited to a small tank. Most of the small biofloc systems use a ring tank, which has good hydrodynamic characteristics and a relatively simple structure, but the construction cost is relatively high, and the power consumption is large. If biofloc shrimp farming can be realized in a square tank with a simpler structure and a certain scale, it will be conducive to the promotion and application of this technology. Root blowers are widely used in small water bodies for water mixing and oxygenation, while the significant increase in the area of aquaculture water has increased the difficulty of water body mixing evenly. Based on this, the experiment envisages ensuring the full mixing of water bodies by introducing waterwheel-type aerators and surge machines.

To verify the above assumption, a group of large-scale biofloc shrimp culture systems with an area of 350 m² was designed in this experiment. Anhydrous glucose was selected as the carbon source. The biofloc content was controlled through the filtration and circulation device, and the bacterial community structure of the system was analyzed by high-throughput sequencing technology, so as to further explore the microbial diversity of the large-scale biofloc culture system.

2. Materials and Methods

2.1. Construction of Shrimp Culturing Facilities

The large-scale biofloc shrimp culturing facility was built in the fishery equipment and engineering pilot base of the institute of fishery machinery and instruments, Chinese Academy of Fishery Sciences, with 3 replicates. As shown in Figure 1, a cement tank with an area of 350 m² and a water depth of 80 cm covered with glass fiber-reinforced plastics was selected. 64 microporous aerators with a diameter of 1 m were set. Polyvinyl chloride (PVC) aeration pipes with a length of 20 m, spacing of 50 cm, aperture of 1.5 mm, and diameter of 20 mm were laid at the bottom of the tank. The aeration pipes were connected with a 2.2 kW root blower through PVC pipes and inflated continuously for 24 h. A 0.9 kW surge machine was set in the tank, two 0.37 kW waterwheel aerators and one 1.5 kW four-wheel waterwheel aerator was equipped with a set of polyvinyl chloride tanks for 24 h continuous addition of inputs and two sets of floc removal devices at the tank side. Moreover, 1.5 kW high-speed aerators were equipped with stainless steel protective covers to prevent their impellers from killing prawns.

2.2. Experimental Method

On 18 July 2020, post-larvae of Litopenaeus vannamei were put into the tank at the density of 500 per square meter. The shrimp harvest time was 6 October 2020. The raw water for culturing was taken from the nearby river course and enters the original tank after preliminary filtration. It was disinfected with 20 mg/L of effective chlorine preparation and aerated at the same time. After 5 days, the residual chlorine was digested with 5 mg/L sodium bicarbonate. The PVC bucket of water was pumped into the shrimp tank through the 24 h continuous adding device of input products to maintain the 25‰ salinity unchanged. When the content of bioflocs exceeds 15 mL/L, started the water pump to filter and remove some bioflocs. Shrimp biomass depends on the average body mass of shrimp sampled every month and the actual survival number of shrimp. During the whole experiment period, no water was changed, and no drugs were added. The water samples collected on the 20th, 40th, 60th, and 80th day of the experiment process were named B20, B60, B60, and B80, respectively, for microbial community analysis. The water will not be changed during the whole experiment period, and only the water lost due to leakage, evaporation, and sampling will be supplemented.

2.3. Feed Feeding and Glucose Addition

In the early stage, shrimp slices and Artemia were fed, and the feeding frequency of Artemia decreased by 15% with the increase in the number of culturing days. After one week, the shrimp seedlings were domesticated to a pure artificial formula feed. During the experiment, a commercial formula feed of Litopenaeus vannamei with a crude protein content of 42% was fed, and each day was fed at 5:30 a.m., 10:00 a.m., 14:30 p.m., 18:30 p.m., and 22:30 p.m. respectively. In order to accurately grasp the feeding amount, three feeding tables were placed in each culturing tank, and the feeding table was checked after one hour. According to the feeding conditions, weather conditions, shrimp growth, and seasonal changes, the feed dosage should be adjusted in time. Input anhydrous glucose and baking soda were shown in

Table 1. Feeding amount, glucose, and baking soda addition during the experiment.

Culture Days	Daily Glucose Addition/Daily Feeding Amount	Daily Baking Soda Addition/Daily Feeding Amount
1~20	150%	15%
21~40	100%	15%
41~60	50%	15%
61~80	0%	15%

.

2.4. Determination of Water Quality Indicators

The temperature, pH value, dissolved oxygen, and salinity were measured on-site with a YSI professional plus water quality analyzer. Total ammonia nitrogen concentration was determined by the sodium hypobromate oxidation method (GB12763.4). Nitrite nitrogen concentration was determined by the diazo azo method (GB12763.4). Total nitrogen concentration was determined by potassium persulfate digestion UV spectrophotometry (GB11894-89); Water quality indicators were measured once every 10 days. Determination of total suspended particulate matter (TSS): take a certain amount of experiment water sample, use the dried Whatman Gf/f glass fiber filter membrane with a diameter of 47 mm to filter, dry the filter membrane with suspended solids to a constant mass at 110 °C, and weigh the dry mass.

The body weight (W) was measured by electronic balance, and the body length (SL) was measured by a vernier caliper

Survival rate (SR) = number of live shrimp at the beginning of fishing/number of seedlings released.

2.5. Sequencing and Analysis of Microbial Diversity

2.5.1. DNA Extraction and PCR Amplification

Microbial community genomic DNA was extracted from biofloc samples using the E.Z.N.A.^® soil DNA Kit (Omega Bio-tek, Norcross, Gwinnett County, GA, USA) according to the manufacturer’s instructions. The DNA extract was checked on 1% agarose gel, and DNA concentration and purity were determined with NanoDrop 2000 UV-vis spectrophotometer (Thermo Scientific, Wilmington, NC, USA). The hypervariable region V3–V4 of the bacterial 16S rRNA gene was amplified with primer pairs 338F (5′-ACTCCTACGGGAGGCAGCAG-3′) and 806R(5′-GGACTACHVGGGTWTCTAAT-3′) by an ABI GeneAmp^® 9700 PCR thermocycler (ABI, Vernon, CA, USA). The PCR amplification of the 16S rRNA gene was performed as follows: initial denaturation at 95 °C for 3 min, followed by 27 cycles of denaturing at 95 °C for 30 s, annealing at 55 °C for 30 s, and extension at 72 °C for 45 s, and single extension at 72 °C for 10 min, and end at 4 °C. The PCR mixtures contain 5 × TransStart FastPfu buffer 4 μL, 2.5 mM dNTPs 2 μL, forward primer (5 μM) 0.8 μL, reverse primer (5 μM) 0.8 μL, TransStart FastPfu DNA Polymerase 0.4 μL, template DNA 10 ng, and finally ddH₂O up to 20 μL. PCR reactions were performed in triplicate. The PCR product was extracted from 2% agarose gel and purified using the AxyPrep DNA Gel Extraction Kit (Axygen Biosciences, Union City, CA, USA) according to the manufacturer’s instructions and quantified using Quantus™ Fluorometer (Promega, Madison, WI, USA).

2.5.2. Illumina MiSeq Sequencing

Purified amplicons were tanked in equimolar and paired-end sequenced on an Illumina MiSeq PE300 platform/NovaSeq PE250 platform (Illumina, San Diego, CA, USA) according to the standard protocols by Majorbio Bio-Pharm Technology Co. Ltd. (Shanghai, China). The raw reads were deposited into the NCBI Sequence Read Archive (SRA) database (Accession Number: SRX10945114, SRX10945115, SRX10945116, SRX10945117).

2.5.3. Processing of Sequencing Data

The raw 16S rRNA gene sequencing reads were demultiplexed, quality-filtered by fastp version 0.20.0 (Chen, Shenzhen, China) [17], and merged by FLASH version 1.2.7(Magoc, Baltimore, MD, USA) [18] with the following criteria: (i) the 300 bp reads were truncated at any site receiving an average quality score of <20 over a 50 bp sliding window, and the truncated reads shorter than 50 bp were discarded, reads containing ambiguous characters were also discarded; (ii) only overlapping sequences longer than 10 bp were assembled according to their overlapped sequence. The maximum mismatch ratio of the overlap region is 0.2. Reads that could not be assembled were discarded; (iii) Samples were distinguished according to the barcode and primers, and the sequence direction was adjusted, exact barcode matching, 2 nucleotide mismatches in primer matching.

Operational taxonomic units (OTUs) with a 97% similarity cutoff [19,20] were clustered using UPARSE version 7.1(Edgar, Tiburon, CA, USA) [19], and chimeric sequences were identified and removed. The taxonomy of each OTU representative sequence was analyzed by RDP Classifier version 2.2(Wang, ELansing, MI, USA) [21] against the 16S rRNA database (e.g., Silva v138) using a confidence threshold of 0.7.

3. Results and Discussion

3.1. Biofloc Deposition

Figure 2 shows the changing trend of bioflocs during the experiment. No bioflocs appeared before the 13th day, and a small amount of bioflocs was detected on the 15th day. Then, the amount of bioflocs increased rapidly, reaching (15.5 ± 3.2) mL/L on the 31st day, and the biofloc removal device was started. From the 32nd day to the end of the experiment, the amount of bioflocs remained between (15.8 ± 3.5) and (16.8 ± 4.3) mL/L. The change trend of TSS was consistent with the change in biofloc content. TSS was very low at the beginning of the experiment. From the 13th day to the 31st day, TSS increased rapidly, reaching (51.56 ± 3.8) mg/L on the 20th day and (228.32 ± 10.56) mg/L on the 31st day. After that, due to the operation of the biofloc removal facility, TSS was basically stable, reaching a peak of (247.46 ± 27.3) mg/L on the 42nd day. Bioflocs were formed by organisms and inorganic substances dominated by aerobic microorganisms in aquaculture water. The amount of carbon source added directly affects the formation time and effect of bioflocs [22]. In this experiment, the first 30 days were heterotrophic bioflocs, and the addition of sufficient organic carbon sources significantly increases the content of bioflocs in the system. At the end of the experiment, the amount of carbon source decreased to zero, and the rising rate of biofloc began to slow down after the 60th day of nitrification biofloc domestication. Researchers conducted a biofloc culture experiment in a 0.64 m³ Litopenaeus vannamei culture tank, Obvious Brown flocs were formed on the 7th day, and then rose rapidly, and to more than 5 mL/L after 28 days [23]. Zhao [24] used the biofloc technology to culture Penaeus japonicus in a 30 m³ cement tank, a small amount of floc was detected on the 8th day, and then rose to 12.6 mL/L. Compared with the results of these researchers, the change rule of flocs in the middle and later period of this experiment was similar, but the formation time of bioflocs in the early stage was relatively long, which may be caused by the large area of shrimp culturing facilities (350 m²) in this experiment.

3.2. Concentration Total Ammonia Nitrogen, Nitrite Nitrogen, and Nitrate Nitrogen

The concentration of total ammonia nitrogen was in the rising stage from 0 to 31 days. On the 31st day, the concentration of total ammonia nitrogen reached the highest (0.28 ± 0.052) mg/L. From the 32nd to 64th day, the concentration of total ammonia nitrogen maintained a slow downward trend. From the 65th to 80th day, it first decreased rapidly to (0.11 ± 0.05) mg/L, and then gradually tended to a stable level (Figure 3a). With the continuous increase of feed input, nitrite nitrogen gradually increased, and increased to (0.1 ± 0.02) mg/L on the 15th day. At this time, the amount of carbon source input decreased slightly, and nitrite nitrogen began to rise gradually, increasing to (0.79 ± 0.38) mg/L on the 35th day. With the further substantial reduction of carbon source addition, nitrite nitrogen soared significantly, reaching the experiment peak of (4.13 ± 1.42) mg/L on the 48th day, indicating that ammonia-oxidizing bacteria in the facility have matured at this time, which can quickly convert ammonia nitrogen produced by shrimp metabolism into nitrite nitrogen. while nitrifying bacteria proliferate slower than ammonia-oxidizing bacteria, and the construction of nitrification function lags behind the nitrification process, so nitrite nitrogen accumulation occurs. From the 49th day to the 67th day, nitrite nitrogen rapidly decreased to (0.2 ± 0.1) mg/L, autotrophic nitrifying flocs were basically domesticated at this time. From the 67th day to the end of the experiment, nitrite nitrogen remained below 0.15 mg/L. (Figure 3b). The concentration of nitrate nitrogen remained low in the first 15 days and rose sharply after 16 days until the peak value of nitrate nitrogen reached (108.57 ± 19.6) mg/L at the end of the experiment (Figure 3c). After the 67th day, this experiment evolved the ecological chain of nitrogen cycling microorganisms. At the beginning of the experiment, the feed was transformed into ammonia nitrogen, and nitrosate bacteria were slowly produced. At this time, nitrosate bacteria had not yet developed and matured, so the nitrite nitrogen in the tank accumulated up to (8.25 ± 3.42) mg/L. With the increasing abundance of nitrobacteria, ammonia nitrogen stabilized, nitrite nitrogen decreased, and nitrate nitrogen increased. Samocha et al. [25] and Tan et al. [15] both have nitrate concentrations higher than 200 mg/L in small-scale biofloc facilities, which is significantly different from the peak value of this experiment (108.57 ± 19.6) mg/L. The reason may be that the culturing cycle (80 d) of this experiment was lower than that of the above two scholars.

3.3. 16 S rDNA Sequencing Analysis

3.3.1. Alpha-Diversity

Table 2. Analysis of the alpha diversity index of the bacterial community in four periods.

Alpha-Diversity Index	Treatment
	B20	B40	B60	B80
Number of OTUs	407.67 ± 15.65 c	423.35 ± 18.46 c	489.69 ± 12.97 b	572.36 ± 13.26 a
Chao 1 index	516.3 ± 21.35 d	569.7 ± 46.53 c	646.8 ± 52.53 b	768.58 ± 36.96 a
Shannon index	3.65 ± 0.22 c	4.68 ± 0.21 b	4.85 ± 0.13 b	5.63 ± 0.16 a
Coverage/%	99.9 ± 0.00	99.85 ± 0.00	99.88 ± 0.00	99.92 ± 0.00

OTU: operational taxonomic units; Different letters after peer data indicate significant differences.

shows the alpha diversity index changes of dominant bacterial communities in four different periods of large-scale biofloc shrimp culturing facilities. The coverage of each treatment was greater than 98%, indicating that the bacterial sequence obtained in this experiment had good coverage, and its sequencing depth was enough to analyze the composition of bacterial community structure and diversity. It can be seen that the number of OTUs of B80 (572.36 ± 13.26) was significantly higher than that of B60 (489.69 ± 12.97), B40 (423.35 ± 18.46), and B20 (407.67 ± 15.65) (p < 0.05). The number of OTUs of B60 was significantly higher than that of B40 and B20 (p < 0.05). The number of OTUs of B40 was not significantly different from that of B20 (p > 0.05). The Chao1 index of B80 (768.58 ± 36.96) was significantly higher than that of B60 (646.8 ± 52.53), B40 (569.7 ± 46.53) and B20 (516.3 ± 21.35) (p < 0.05). The Shannon index of B80 (5.63 ± 0.16) was higher than that of B60 (4.85 ± 0.13), B40 (4.68 ± 0.21) and B20 (3.65 ± 0.22), with significant difference (p < 0.05). The Shannon index of B60 and B40 was significantly higher than that of B20, with a significant difference (p < 0.05). The Shannon index of B60 was not significantly different from that of B40 (p > 0.05).

Qin et al. [26] analyzed the microbial diversity of shrimp biofloc facilities through high-throughput sequencing. The results showed that the Chao1 index, OTU number, and Shannon index of the experimental facilities increased significantly with the increase of culturing cycle. In this experiment, with the progress of the experiment, reducing carbon input treatment increased the bacterial diversity index of bioflocs. The overall number of bacterial communities in the B20 period was weak, significantly lower than that in B60 and B80 periods. The number of bacterial communities in bioflocs increased with the increase in feeding.

3.3.2. Component Abundances at the Phylum Level

The changes in the level of dominant bacteria in large-scale biofloc shrimp culturing facilities were shown in Figure 4. From this figure, it can be seen that the development of biofloc bacterial community structure in different periods has different responses to different microecological environments. In the B20 period, the bacterial species contained in bioflocs were Proteobacteria, Bacteroides, and Actinobacteria, followed by Chloroflexi, Cyanobacteria, Plantomycetes, Gemmatimonades, and Verrucomicrobia; with the culturing process, the order of the percentage of dominant bacterial groups in bioflocs changed. In the B40 period, Proteobacteria, Bacteroides, and Chloroflexi were dominant, Parcubacteria, Actinobacteria, and Cyanobacteria were next, and Planctomycetes, and Verrucomicrobia were weak; In the B60 period, Proteobacteria, Bacteroides and Actinobacteria were dominant, followed by Chloroflexi, Cyanobacteria and Planctomycetes, and Planctomycetes, Verrucomicrobia and Nitrospirae were weak; In B80 period, Proteobacteria and Bacteroides were dominant, followed by Actinobacteria and Chloroflexi, and Nitrospirae, Cyanobacteria, Planctomycetes, and Verrucomicrobia were weak.

The abundance values of Proteobacteria in B20, B40, B60 and B80 were (46.51 ± 8.63%), (46.06 ± 9.67%), (42 ± 11.73%) and (46.98 ± 15.82%) respectively, with no significant difference (p > 0.05); Proteobacteria is a symbiotic bacterium in biofloc facilities. It can use fecal residues to complete nitrogen and phosphorus removal while degrading organic matter and plays an important role in the nitrogen cycle chain [27,28]. At the same time, it provides a carrier and symbiotic environment for the survival and development of other bacteria. The relative abundance of Bacteroidetes in the B60 and B80 periods was significantly higher than that in the B20 and B40 periods (p < 0.05), showing a trend of increasing with the cycle of biofloc culture. Bacteroidetes have a very strong ability to metabolize complex organics, proteins, and lipids [29]. Because the standardized biofloc shrimp culturing facilities contain many macromolecular organics that were difficult to decompose, Bacteroidetes continue to expand. Actinobacteria also account for a large proportion, and the abundance values in the four periods were (9.46 ± 1.15%), (5.18 ± 0.53%), (11.59 ± 1.86%, (6.34 ± 1.09%), showing unstable changes. Some bacteria in Actinobacteria can secrete extracellular substances beneficial to shrimp, and were important functional bacteria in biofloc facilities [30]. The relative abundance of Chloroflexi increased first and then decreased steadily in the four periods; Chloroflexi often exists in the form of floc skeleton, which not only provides skeleton support for the structure of large-scale bioflocs, but also plays the role of biological phosphorus removal [31].

The abundance value of Planctomycetes first decreased significantly and then stabilized. Some bacteria of Planctomycetes can use NH₄⁺-N and NO₂⁻-N as electron donors and receptors respectively to react to produce gaseous nitrogen, so as to biodenitrify. They were important functional bacteria in the tank [32]. The relative abundance of Nitrospirae [33], which can oxidize nitrite in the soil to nitrate, was significantly higher in B80 (1.35 ± 0.26%) and B60 (1.08 ± 0.23%) than in B20 and B40 treatments, which proved that there was a strong degree of nitrification process in B80 and B60. It explained the change law of three-state nitrogen in the four periods of this experiment from the perspective of microecology and indirectly supported the speculation that the nitrification function of bioflocs was mature at about the 67th day.

3.3.3. Component Abundances at the Genus Level

In order to deeply analyze the influence of different light intensities on the bacterial community structure of bacteria and algae biofilm, according to the data of flora community genus pair abundance in four periods of B20, B40, B60, and B80, through comparison with Silva (ssu123) database, a total of 538 microbial genera were identified. Figure 5 shows the genus with a bacterial percentage greater than 1%. It can be seen from the figure that the bacteria with the highest relative abundance in the B20 period was Ruegeria (15 ± 1.73%). While the bacteria with the highest relative abundance in B40, B60, and B80 periods were unclassified_f_Rhodobacteracea (34.07 ± 5.69%, 20.69 ± 6.53%, 22.97 ± 3.82%). The secondary dominant bacteria in the B60 and B80 stages of acclimation completion were Ruegeria.

Unclassified_f_Rhodobacteracea in B40 period (34.07 ± 5.69%) was significantly higher than that in B60 period (20.69 ± 6.53%), B80 period (22.97 ± 3.82%) and B20 period (6.73 ± 1.08%) (p < 0.05). It showed a trend of the rapid rise and then stabilized in the experiment. Unclassified_f_Rhodobacteracea contains denitrifying bacteria [34]. These bacteria can denitrify with nitrate nitrogen and nitrite nitrogen as substrates under aerobic conditions [35] to achieve the purpose of biological denitrification. This result was similar to the research results of flora diversity of three bioflocs in Litopenaeus vannamei culturing water body by Zhang et al. [19]: the relative abundance value of unclassified_f_Rhodobacteracea in the later stage of the experiment was stable at about 20%. In B20 stage (15 ± 1.73%), the genus of Ruegeria was significantly higher than that in B60 (7.16 ± 0.86%), B80 (10.35 ± 1.26%) and B40 stage (1.1 ± 0.53%) (p < 0.05). Ruegeria had great changes in the early and middle stages and was relatively stable in the late stage. Ruegeria is one of the common genera in the marine environment [36]. The bacteria of this genus can often be isolated from marine water bodies, sediments, and marine organisms [37,38]. After the rapid decline of abundance in the B40 period, Ruegeria may use the density sensing pathway to compete for survival, so that it can occupy a certain advantage and stabilize in the B60 and B80 periods in such a high-density environment of large-scale biofloc facilities [39].

Muricauda maintained an advantage in four periods, of which the relative abundance values in B20 (5.55 ± 1.06%), B60 (5.82 ± 0.93%), and B80 (5.73 ± 0.61%) were significantly higher than those in B40 (2.43 ± 0.39%). Muricauda can produce a variety of extracellular enzymes to decompose and utilize complex carbon sources [40], while simple carbohydrates produced by hydrolysis can provide carbon sources for other flora to remove excess N and P in water [41]. Algoliphagus was not detected in the B20 period and remained relatively weak in B40 and B60 periods. Its relative abundance increased rapidly to (3.75 ± 0.85%) in the B80 period. Algoliphagus belong to Bacteroidetes, Cytophages, and it is a Gram-negative aerobic bacillus, which can cause microalgae to flocculate and crack, and has strong algae killing ability [37], which may be an important reason for the gradual reduction of Cyanobacteria in the later stage of the experiment.

In the rotating biological contactor, trickle filter, bead filter, and biological fluidized bed nitrification filter of the mariculture system, Nitrosomonas and Nitrospira have a considerable proportion [42,43]. Yue et al. [44] studied the microbial community structure of submerged biofilter and found that the main microorganism of nitrite-oxidizing bacteria in the filter was Nitrospira. In this experiment, the relative abundance of Nitrospira at the genus level in the B60 and B80 periods was (1.15 ± 0.33%) and (1.56 ± 0.56%), which was significantly higher than that in the B20 and B40 periods. The main microorganism of NOB was Nitrospira, which was similar to the research results of submerged biofilters. The reason may be that the micro ecological environment of large-scale biomass flocs was closer to that of submerged biofilters.

3.4. Growth Performance of Shrimp

It can be seen from

Table 3. Prawn growth and culture effect of large-scale biofloc shrimp culture facility system.

30 Day-Age		60 Day-Age		80 Day-Age		Survival Rate/%	Unit Output/kg·m−3
Weight/g	Length/cm	Weight/g	Length/cm	Weight/g	Length/cm
0.93 ± 0.12	4.78 ± 0.43	6.18 ± 0.65	8.02 ± 0.65	12.96 ± 2.85	12.26 ± 2.56	65.32 ± 6.85	4.15 ± 1.58

that at the age of 30 days, the average body length of shrimp in the large-scale biofloc shrimp culture facility was (4.78 ± 0.43) cm, and the average weight was (0.93 ± 0.12) g; At 60 days old, the average body length of shrimp was (8.02 ± 0.65) cm, (6.18 ± 0.65) g; At 80 days of age, the average body length was (12.26 ± 2.56) cm and the average body weight was (12.96 ± 2.85) g. The survival rate and unit yields were (65.32 ± 6.85)% and (4.15 ± 1.58) kg/m³, respectively. This experiment achieved good shrimp culture results. Zhang et al. [45] used a square cement tank with a side length of 6 m to carry out the circulating water culture experiment of Litopenaeus vannamei. The culture cycle was 85 days, the survival rate of shrimp (74.58 ± 1.74%), and the yield (3.91 ± 0.49 kg/m³). Deng [46] used 0.42 m³ to carry out the closed culture experiment of bioflocs of Litopenaeus vannamei. The results showed that the survival rate of 300/m² group was 83.8% and the yield was 3.66 ± 0.36 kg/m² in the 84 d culture period. In this experiment, the survival rate (65.32 ± 6.85%) and yield (4.15 ± 1.58 kg/m³) of shrimp were similar in the culturing cycle close to those of the above researchers. On the other hand, due to the reduction of aisles and reserved operation space, under the same shrimp culturing land area, compared with small tank shrimp culturing facilities, large-scale biofloc shrimp culturing facilities can increase the culturing water surface by 5–10%.

4. Conclusions

(1)

With the Reducing carbon input treatment, the Chao1 index, OTU number, and Shannon index of the bacterial community in large-scale biofloc shrimp culturing facilities increased significantly. The number of OTUs of B80 (572.36 ± 13.26) was significantly higher than that of B60 (489.69 ± 12.97), B40 (423.35 ± 18.46) and B20 (407.67 ± 15.65) (p < 0.05). The Chao1 index of B80 (768.58 ± 36.96) was significantly higher than that of B60 (646.8 ± 52.53), B40 (569.7 ± 46.53) and B20 (516.3 ± 21.35) (p < 0.05). The Shannon index of B80 (5.63 ± 0.16) was higher than that of B60 (4.85 ± 0.13), B40 (4.68 ± 0.21) and B20 (3.65 ± 0.22), with significant difference (p < 0.05). The development of community structure in four periods had different responses to different microecological environments, and there was a strong degree of nitrification process in the B80 and B60 periods. At the end of the experiment, the domestication formed a micro-ecosystem with Proteobacteria as the carrier (46.98% ± 15.82%), Chloroflexi as the skeleton (2.2% ± 0.36%), Nitrospirae (1.35% ± 0.26%) as the main water treatment functional bacteria, and other bacteria as auxiliary nitrogen and phosphorus removal; At the genus level, unclassified_f_Rhodobacteracea (22.97% ± 3.82%), Ruegeria (10.35% ± 1.26%), muricauda (5.73% ± 0.61%), Algoriphagus (3.75% ± 0.85%) and Nitrospira (1.56% ± 0.56%) were the dominant bacteria. Under the synergy of the above bacteria, bioflocs can effectively regulate and control water quality, maintain a low level of ammonia nitrogen and nitrite nitrogen without drainage, and maintain a relatively stable shrimp culturing ecosystem.

(2)

Large-scale biofloc culture facilities have strong feasibility in shrimp culture and production. biofloc content (0~16.8 ± 4.3) ml/L, TSS concentration (0~247.46 ± 27.3) mL/L, total ammonia nitrogen concentration (0~0.28 ± 0.052) mg/L, nitrite nitrogen concentration (0~4.13 ± 1.42) mg/L, nitrate nitrogen concentration (108.57 ± 19.6) mg/L were all within the safe concentration range of Litopenaeus vannamei; The survival rate and unit yield of shrimp were (65.32 ± 6.85)% and (4.15 ± 1.58) kg/m³ respectively, which were similar to the reported yields of small-scale Litopenaeus vannamei circulating water and biofloc culture experiments. However, due to the reduction of the aisle and reserved operation space, this facility has the advantage of land saving.

(3)

The facilities of the biofloc shrimp culture system need to be further optimized. The uniformity of the water and the power allocation of the system are not necessarily the optimal solutions, which need further research and maturation. We need to develop new equipment to control the total amount of bioflocs and utilize them as resources. During the domestication of biological flocs, the current difficulties are the high frequency of water measurement and the long domestication cycle. No standardized domestication technology has been formed yet. It is necessary to research the accurate control of ammonia nitrogen concentration and the corresponding key technologies of bacterial nutrition; There are still many unknowns about the ecological function and development mechanism of biofloc microorganisms. Denitrifying biofloc is a new type of denitrifying biofloc system in which aerobic denitrifying bacteria are added to autotrophic nitrifying biofloc. It can convert 82% of feed protein into gaseous nitrogen and leave the aquaculture system. It is the most advantageous way to achieve “zero water exchange”. At present, it is in its infancy. How to select special nutrients and suitable carbon sources, and the mechanism of aerobic nitrogen removal need further research.