Isolation and Characterization of Cellulolytic Marine Bacteria for Litopenaeus Vannamei Aquaculture Using Sugarcane Bagasse as Carbon Source

Background: In aquaculture system, it is essential to adjust the inherent disadvantage of C/N ratio by adding a lot of additional carbon sources, such as sugarcane molasses, organic acids, and organic acid salts, which will greatly increase the cost of shrimp aquaculture. Herein, we aimed to isolate cellulolytic marine bacteria to hydrolyze sugarcane bagasse (SB) for reducing the cost of addition of external carbon sources in industrial Litopenaeus vannamei aquaculture. Results: A total of 97 cellulolytic marine bacterial strains belonged to 6 genera were isolated from 2,585 indigenous bacteria, indicating that seagrass bed can be used as an important place for screening the cellulolytic bacteria. The hydrolysis capacity (HC) of 58 cellulolytic marine bacterial strains was ranged from 1.1–4.0. MW-M5 displayed the largest HC value, followed by MW-M10 and MW-M14. The cellulase contents of 30 strains were more than 3 U/g in the supernatant of fermentation broth after 24 h, which was signicantly higher than that of commercial cellulose. 26 cellulolytic marine bacteria with HC greater than 2 were safe for L. vannamei. MW-M19 with the lowest multiple antibiotic resistance index, 0.1, had a highest SB enzyme activity, 4.14 U/mL. The SB decomposition rates of CFW-C18 and MW-M15 were up to about 63.81% and 48.57% after fteen days, respectively. Conclusions: These results provide valuable information for further construction of a shrimp aquaculture system based on low-cost external carbon sources using cellulolytic bacteria, and even for other biotechnological applications. SB: sugarcane bagasse; HC: hydrolysis capacity; MAR Index: multiple antibiotic resistance index; FPase: lter paper activity; SBase: SB enzyme activity; CMC: carboxymethylcellulose; DNS: 3,5-dinitrosalicylic acid; SCS: South China Sea; REC: relative enzyme content; L. vannamei: Litopenaeus vannamei; L. pentosus: Lactobacillus pentosus; Trichoderma harzianum: T. harzianum.


Background
Litopenaeus vannamei, also known as Paci c white shrimp or King prawn, which has become one of the pillar industries in the agricultural economy of most coastal areas [1,2]. It is also one of main prawn species in China.
Due to its high economic value, delicious meat, rich nutrition, low-feed conversion rate, high survival rate, low water consumption, and short growth cycle, L. vannamei is considered a good alternative [3]. With the change of aquaculture environment and facility and the improvement of aquaculture technology, the mode of L. vannamei aquaculture experienced a gradual change from extensive aquaculture to intensive aquaculture. However, the intensive aquaculture has been restricted because of the eutrophication of water [4], such as nitrogen and phosphorus enrichment and organic deposition [5,6], which indirectly led to harmful algal blooms [7][8][9][10]. In order to pursue high yield and income, some operations blindly increased the scale of aquaculture and put in bait and fertilizer containing a large amount of nitrogen and phosphorus, which led to water deterioration and frequent disease. In recent years, a large amount of nitrogen and bait remained in the aquaculture system couldn't be used in the process of L. vannamei aquaculture, and then the accumulated biological excreta and inedible feed in the aquatic environment will be decomposed into high concentrations of nitrogen compounds, such as ammonia and nitrite [11][12][13]. The increase of total nitrogen content in water leads to the decrease of C/N ratio and the lack of carbon content, which makes it di cult for microorganisms to survive and inhibit the assimilation of heterotrophic microorganisms [3,14]. Ammonia nitrogen and nitrite nitrogen will decrease with the increase of C/N, and the denitrogen rate of bacteria will furtherly increase with the increase of C/N [15].
Currently, biological treatments in tail water of aquaculture have become a research hotspot. Some researchers reported that Bacillus could improve water quality, increase the survival rate and health of prawns, and prevent the outbreak of pathogenic Vibrio [16]. The bio-oc technology provides a method to solve the shortage of protein feed resources and the persistent organic pollution. There is no need to exchange water or a small amount of water exchange in the process of intensive aquaculture, thereby improving sustainability of the aquaculture system, biological safety, and production of aquaculture [17]. In this system, carbon source is supplemented according to the characteristics of microorganisms to maintain water quality, which can promote the growth of heterotrophic bacteria, remove inorganic nitrogen from the water, and maintain water balance [18,19]. In addition, the feed containing cellulase has been paid more and more attention in the industrial aquaculture in recent years. The nitrogen that can be used by prawns is actually very limited in the process of prawn aquaculture, and the nitrogen of the high-protein bait deposited in the bottom mud will make the water turn black. However, the inorganic nitrogen in the experimental shrimp ponds can be reduced by adding carbon source. The process mainly transforms nitrogen into protein required by microorganisms as the source of feed protein for L. vannamei to avoid the accumulation of inorganic nitrogen in water body. The exogenous carbon sources can promote the growth of microorganisms in aquaculture and provide the required carbon source to form bio-ocs, which will further meet the ingestion of prawns and reduce the accumulation of toxic inorganic nitrogen compounds [20][21][22]. It is a practical and cheap method to control inorganic nitrogen by controlling C/N ratio in aquaculture system. If the C/N ratio in water is too high, a part of nitrogen will be used by bacteria to synthesize protein. Especially in the later stage of cultivation, the additional carbon sources, such as sucrose and molasses, can increase the C/N ratio, promote the growth of heterotrophic bacteria, enrich the diversity of bacteria, improve the aquaculture environment, and increase the total production of aquaculture [18,19,22,23].
Therefore, it is an effective method to improve the yield of L. vannamei and solve the water pollution by adding exogenous carbon sources in the aquaculture system to adjust the C/N ratio. The carbon sources supplemented in aquaculture mainly include sucrose, glycerin, glucose, and sodium acetate [24][25][26]. However, these carbon sources are expensive and easy to cause environmental pollution. Accordingly, it is urgent to nd cheap carbon sources for long-term feeding in aquaculture. Sugarcane is a widely grown sugar crop in Hainan, China, and its bagasse is the main wastes of sugar industry. Sugarcane bagasse (SB) as major by-product of sugar industry, in general, about 260-280 kg wet SB produced from per ton sugarcane, with total production of more than 279 MMT tons annually worldwide [27,28]. About 40-50% of dried SB is cellulose, while it is very di cult for this polysaccharide to decompose because of its stable chemical linkages inside a monomeric cellulose chain, as well as the crystalline structure formed by multiple cellulose micro brils that are interconnected by hydrogen bonds [29]. Cellulose is the most abundant organic matter on the earth and characterized by cheap and easily available carbon source. Microbial cellulolytic enzymes, called cellulase, are complex enzymes that consist of endoglucanases, exoglucanases and β-glucosidases, which synergistically work to hydrolyse the β-1,4 glycosidic bonds of cellulose [30]. Most cellulolytic enzymes with high activity used in commercial applications are produced from fungi [31]. By contrast, bacteria have higher growth rate and enzyme production rate than that of fungi. More importantly, bacteria have higher thermal stability and genetic stability than that of fungi [32]. In addition, the oceans, which covers more than three quarters of the Earth's surface, is an open ecosystem. It has been estimated that at least 50%, and potentially more than 90%, of all marine species are undescribed [33][34][35][36]. Marine microorganisms have been recognized as potential sources of novel enzymes with better biocatalytic properties than their terrestrial counterparts. These properties, such as salt tolerance, hyperthermostability, barophilicity, alkali-resistance, and low optimum temperatures, are necessary for e cient bioprocesses exploitation [33][34][35][36]. Hainan is a big maritime province in China and rich in marine resources, but the knowledge of cellulolytic microbes isolated from this area is limited. At present, it is necessary to add a lot of additional carbon sources to adjust the inherent disadvantage of C/N ratio in aquaculture system, such as sugarcane molasses, organic acids, and organic acid salts [23, 37, 38], which will greatly increase the cost of shrimp aquaculture. To overcome these problems, the marine bacteria demonstrating SB cellulolytic performance were isolated from the alongshore of the Hainan Province, China, which were suitable for tropical marine aquaculture. The purpose was to determine competent cellulolytic marine bacteria for SB decomposition as the low-cost external carbon source in L. vannamei mariculture system.

Results
2.1. Isolation, screening, and identi cation of cellulolytic marine bacteria A total of 2,585 marine bacteria with dissimilarly morphological colonies were isolated from four sites in Hannan (Table 1). Among them, ninety-seven bacterial isolates were de ned as cellulolytic bacteria because they exhibited the cellulolytic zone around their colonies on 2661E agar after Congo red staining, which were belonged to 6 genera ( Fig. 1 and Table S2). The largest genus was Bacillus (64%), followed by Vibrio (18%). The remaining genera were shared by Microbulbifer (10%), Pseudomonas (6%), Tenacibaculum (1%), and Muricauda (1%). We found that the cellulolytic bacteria were mainly from the natural environments (seagrass beds of Mangrove and Dongjiao Coconut Forest). However, there were only 6 cellulolytic bacteria isolated from aquaculture environments (two shrimp cultural bases). As shown in Fig. 1, Bacillus was accounted for the most in natural environment and aquaculture environment, followed by Vibrio. Interestingly, Vibrio, Pseudomonas, and Microbulbifer were only isolated from the natural environment, while Tenacibaculum and Muricauda were only isolated from the aquaculture environment.

Cellulolytic activity of cellulolytic marine bacterial strains
At present study, there were 97 marine bacterial strains showing FPase activity (Table S2). MW-C57 belonging to Bacillus had the highest FPase activity, 2.325 U/mL. The FPase activities of other marine bacteria from this study areas were ranged from 0.018 to 0.597 U/mL. Of the 97 marine bacterial strains, only 58 strains had HC values more than 1 ( Table 2). Among them, MW-M5 displayed the largest HC value (3.951), followed by MW-M10 (HC value=3.752) and MW-M14 (HC value=3.505). In addition, there are 6 strains with HC value greater than 3, 7 strains with HC values ranged from 2.5 to 3, 15 strains with HC values ranged from 2 to 2.5, and 29 strains with HC values ranged from 1 to 2. Based on the linear relationship ( Fig. S1 and S2) between HC value and content of commercial cellulase from Aspergillus niger (TCI-Tixi Ai (Shanghai) Chemical Industry Development Co., Ltd., Shanghai, China), the cellulase production from 58 strains were quanti ed ( Table 2). In this study, of the 28-test cellulolytic marine bacteria except CFW-C9 and MW-C47 were harmless to shrimp as no mortality was observed by subjecting the test animal, L. vannamei to high density (10 -7 CFU mL -1 water) immersion challenge of cellulolytic marine bacteria for consecutive 7 days (Table 3).

Determination of the SB decomposition of the safe cellulolytic bacteria
In this study, SB powder with diameter of 0.05-0.1 mm was used as the sole carbon source, and then evaluated the ability of SB decomposition of cellulolytic marine bacterial strains in order to lay the foundation for providing cheap carbon sources to aquaculture in the future. As shown in Table 3, MW-M19 has the highest SBase (4.14 U/mL). The SBases of MW-C77, MW-C44, MW-C79, and MW-M17 were 3.746 U/mL, 2.152 U/mL, 1.952 U/mL, and 1.75 U/mL, respectively. 26 cellulase producing strains were inoculated with SB as the sole carbon source, and the SB decomposition rate of each strain was also determined after 15 days. As shown in Table 3, the SB decomposition rates of 8 cellulolytic marine bacteria had no signi cant difference compared with that of control.
The SB decomposition rates of 14 cellulolytic marine bacteria were more than 30%. Among them, the SB decomposition rates of CFW-C18 and MW-M15 were up to 63.81% and 48.57%, respectively.

Isolation, screening, and identi cation of cellulolytic marine bacteria
The South China Sea (SCS) is vast, covering more than 2 million square kilometers, which is one of the largest marginal seas [39] with rich microbial resources in the world. There are a lot of seagrass ecosystems in SCS.
Seagrass ecosystems are considered major blue carbon sinks, which occupy less than 0.2% of the area of the world's oceans but are estimated to bury roughly 10% of the yearly estimated organic carbon burial in the oceans [40,41]. Seagrass bed is also a marine ecosystem with high microbial diversity, and the cellulose is one of main organic matter [42]. Therefore, we reason that bacteria with highly cellulolytic activates are likely to exist in these marine ecosystems. However, the knowledge of cellulolytic marine bacteria from Hainan seagrass beds is limited.
Herein, our aim is to isolate cellulolytic marine bacteria for L. vannamei aquaculture from these areas to reduce its cost by hydrolyzing SB. 91 cellulolytic marine bacteria of 97 cellulolytic isolates were from seagrass bed ecosystems, indicating that seagrass beds can be used as an important place for screening the cellulolytic bacteria. These results indicated that Bacillus and Vibrio were the dominant cellulolytic bacteria in seagrass ecosystems. In addition, although we only obtained 6 cellulolytic bacteria from aquaculture environments, there were 4 strains belonging to Bacillus, indicting that Bacillus was also the dominant cellulolytic bacteria in aquaculture ecosystems.
Most cellulolytic enzymes isolated from mangrove were from bacteria belonging to the genera Micrococcus, Bacillus, Pseudomonas, Xanthomonas, and Brucella [43,44]. The cellulolytic bacteria isolated from different areas have different properties. The cellulolytic bacterium from farm at Zhanjiang (Guangdong Province, China), Lactobacillus pentosus AS13, can effectively enhance the growth performance, feed utilization, digestive enzymes and disease resistance of L. vannamei [45]. Geobacillus thermodenitri cans IP_WH1 from North West Himalayas could produced a thermotolerant cellulase [46]. Therefore, these cellulolytic marine bacteria from the tropical marine environment may be also characterized by potential use.
So far, more researches on mesophilic and thermophilic cellulases have been reported, while fewer works were on low-temperature cellulase. Low-temperature enzymes can not only ensure high-e ciency enzyme reactions at lowtemperature, but also inactivate enzymes at low-temperature, thereby saving energy and cost in the production process. The probability of developing low-temperature cellulase from terrestrial organisms is much lower than that from marine organisms. The marine environment is very unique. Marine microbial enzymes are usually identi ed by cold tolerance and salt-alkali resistance, and have unique application prospects compared with terrestrial microbial enzymes. Therefore, the cellulolytic marine bacteria from marine environments may be have more potential applications than those from terrestrial, particularly on mariculture.

Cellulolytic activity of cellulolytic marine bacterial strains
Cellulases are mainly divided into endoglucanases, exoglucanases and β-glucosidases, and the three enzymes act synergistically in cellulose hydrolysis [30]. The cost of cellulolytic enzymes is a major factor in the hydrolysis of lignocellulosic materials to fermentable sugars [47]. Currently, various cellulolytic microorganisms were from fungi, such as Aspergillus [48,49], Penicillium [50,51], Trichoderma [30, [52][53][54], and Talaromyces [55][56][57], which required long production cycles. Compared with fungi, bacteria not only have a higher growth rate, but also produce cellulase in a short time. At the present study, the highest FPase activity of Bacillus MW-C57, 2.325 U/mL, was 18-fold higher than that of NAB37 from Haryana [58]. Unfortunately, MW-C57 didn't display HC. The FPase activity of other marine bacteria from this study areas was similar to that from Haryana [58]. The FPase activities of our marine bacteria, 0.018 to 0.597 U/mL, were lower than that of fungi, for example, the FPase of T. harzianum belonging to fungi was in the range of 0.445-2.7 U/mL [30].
In this study, a total 28 cellulolytic marine bacterial strains with HC value greater than 2 were classi ed as high enzyme activity strains, which were used for further screening the safe cellulolytic marine bacteria for L. vannamei ( Table 3). The cellulolytic marine bacterial with the largest HC value were isolated, namely MW-M5 (HC value=3.951), MW-M10 (HC value=3.752), and MW-M14 (HC value=3.505). The HC values of cellulose degrading bacteria from mangrove soils were ranged from 1.25 to 2.5 [43], which were very similar to that from ower stalksvegetable waste co-composting system (HC values were ranged from 0.4 to 2.1) [59]. HC values of cellulolytic bacteria from oil palm were ranged from 1.56 to 4.14 [60].
It is worth noting that there were 30 strains with the relative enzyme content (REC) more than 3 U/g in the supernatant of fermentation broth after 24 h according to the linear relationship between HC value and content of commercial cellulase (2.4 U/g), which were signi cantly higher than that of commercial cellulose. The method for quantitatively analyzing the detected enzymic production of the strains according to the standard curve of known enzyme is more intuitive, faster, and more effective, which will provide a new idea for the method of measuring enzymic activity of cellulolytic strains.
3.3. Screening of the safe cellulolytic marine bacteria for L. vannamei aquaculture For the aquaculture of L. vannamei, the use of probiotics as a group of live bacteria that added directly to water has bene cial effects on organisms in culture, such as water bioremediation, improvement of digestion, intensi cation of the immune response, and inhibition of the growth of pathogenic bacteria [61]. FAO has suggested the use of probiotics as a major means for the improvement of aquatic environmental quality [62]. People usually use probiotics to replace antibiotics in the process of aquaculture of aquatic animals, which can not only restrain the multiplying of pathogen and enhance the immunity of cultured animals, but also improve the aquaculture environment. However, there is no report about the cellulolytic probiotics from the tropical marine environment in Hainan. Therefore, the indigenous cellulolytic probiotics from the Hainan tropical aquaculture system and the marine environment of SCS that are bene cial for L. vannamei aquaculture will not only ensure the safety and improve the production of white shrimp, but also reduce the cost of aquaculture by using the cheap SB as the only carbon source.

Safety of cellulolytic marine bacteria to L. vannamei
In order to be considered as a probiotic, the strain has to be non-toxic to the host. 26-test cellulolytic marine bacteria were safe for L. vanname (Table 3), which were agreed with the earlier reports where no mortality has been detected by Macrobrachium rosenbergii and L. vannamei with L. pentosus [45,63]. In aquaculture water, the changes of salinity will affect the physiological function of microorganisms, which will eventually hinder their growth, survival, and food consumption [64]. In the present study, we also detected the salinity tolerance and pH tolerance of these 26 cellulolytic marine bacterial strains (data not shown). There were 11 strains suitable for fresh In addition, all of these bacteria can't grow under over-acidi cation condition (pH<4), while some strains can grow under alkaline condition (pH 10). There were 13 strains with a wide range of acid-alkali resistance. These results will provide important information for the adaptation of these cellulolytic marine bacteria to industrial aquaculture applications.

Antibiotic susceptibility assay
In addition to the detection of animal safety, antibiotic susceptibility analysis is also an effective method for identifying potential probiotics. The abuse of antibiotics will destroy the balance of the original microbial ora in the breeding environment, resulting in drug resistance. At the present study, the main purpose was to screen the highly safe strains with low drug resistance. Most test cellulolytic marine bacteria were susceptible to chloramphenicol. Chloramphenicol is a broad-spectrum antibiotic, which has a strong inhibitory effect on Grampositive and Gram-negative bacteria (Goldfarb, Doi, Rodriguez, 1981), and has been used in aquaculture industry since 1980s (Cravedi, Choubert, Delous, 1987; Reide, Siegmund, 1989; Rijkers, Teunissen, Van Oosterom, Van Muiswinkel, 1980). According to our results, the animals in the study environments had been threatened by penicillin, piperacillin, carbenicillin, cefalexin, streptomycin, spectinomycin, tetracycline, erythromycin, and azithromycin, while some antibiotics with relatively lower MAR index could be reasonably used in these environments, such as cefoxitin, ceftriaxone, chloramphenicol, and furazolidone. More importantly, Vibrio sp. WM-M19 displayed a lowest MAR index (0.1), which was resistant to only three antibiotics, indicating that this strain was safe and non-drug resistance for further use. Of course, antibiotic resistance pattern may vary depending on the geographical locations and selective pressure and these patterns change rapidly from time to time [65].

Determination of the SB decomposition of the safe cellulolytic bacteria
Although high-density aquaculture can bring high pro ts, the accumulation of high-protein bait and excrement will aggravate the deterioration of the aquaculture water [4]. Residual bait and feces are the main sources of nitrogen and phosphorus during the pressing of aquaculture, while the relative lack of carbon will lead to water deterioration [3,[11][12][13][14]. However, the high cost of adding carbon source arti cially limits its application of largescale popularization. Therefore, it has become a research hotspot to nd cheap carbon sources. Sugarcane, a tropical crop, requires special environment, such as, an optimum high temperature, availability of sunshine and higher rainfall for its growth. The extraction of juice from sugarcane generates a brous residue, SB. Normally one ton of sugarcane produces 280 kg of SB [66]. SB is rich in fermentable components and usually composed of 25% lignin, 25% hemicelluloses and 50% cellulose [66,67]. Sugarcane is a widely grown sugar crop in Hainan, China, and its SB is the main wastes of sugar industry. However, there were few reports on SB decomposition of cellulolytic marine bacterial strains in shrimp aquaculture. The cellulase producing strains from seagrass beds of mangrove displayed relatively higher SBase, such as MW-M19 (4.14 U/mL) and MW-C77 (3.746 U/mL), we reasoned that the cellulolytic bacteria isolated from the seagrass beds in Wenchang could induce the production of cellulase with high enzyme activity by SB, further indicating that it will be more instructive for shrimp aquaculture to isolate cellulolytic strains from the environment similar to the culture environment. Generally, the larger the contact area of SB with microorganisms, the more easily the SB can be degraded into available carbon sources by microorganisms, which can promote the growth of microorganisms and improve water quality.
CFW-C18 and MW-M15 showed the excellent SB decomposition capabilities, and the SB decomposition rates of both were up to 63.81% and 48.57%, respectively. Interestingly, MW-M19 with lowest MAR Index and highest SBase also displayed the relatively high SB decomposition rate (up to 29.52%). Overall, CFW-C18 and MW-M19 with relatively high SBase and SB decomposition rate were obtained in this study, which will lay a foundation for the development of cellulase producing marine microbial agents for L. vannamei aquaculture based on the arti cial addition of SB as the carbon source.

Conclusions
In conclusion, we reported the cellulolytic marine bacteria from the tropical marine environment in SCS. Seagrass bed was an important place for screening the cellulolytic bacteria, and the Bacillus and Vibrio were dominant cellulolytic bacteria in seagrass ecosystems. We successfully obtained the safe cellulolytic bacteria with high hydrolysis capacity and de nite antibiotic resistance pro le, such as MW-M19 with lowest MAR Index and highest SBase, and CFW-C18 and MW-M15 with relatively higher SB decomposition rate, which will make them pro cient candidates for L. vannamei aquaculture based on low-cost external carbon sources, and even for other biotechnological applications.

Description of sampling sites and sample collection
The shrimp samples and the water samples were extracted by digging 0.5 m deep around 1 m away from aerator of the aquaculture ponds, and the muddy sediment samples were collected around 3 cm away from seagrass. All of the samples were from four sites: shrimp cultural base in Haiwei Town, Changjiang Li Autonomous County (19°26′39.02″N, 108°50′11.23′′E), shrimp cultural base in Huiwen Town, Wenchang City (19°27′28″N, 110°45′13″E), seagrass beds around mangrove in Huiwen Town, Wenchang City (19°28′11.66″N, 110°47′41.22′′E), and seagrass beds in Dongjiao Coconut Forest, Huiwen Town, Wenchang City (19°31′28.81″N, 110°52′0.45″E). The shrimp samples and the mud samples were putted in the sterile zipper plastic bags, and the water samples were stored in the sterile sampling bottles. All of the samples were stored at 4°C for further use.

Isolation and puri cation of marine bacteria
The samples were serially diluted with sterile normal saline solution (0.85% NaCl) within 24 h of collection to obtain 1:10, 1:10, 1:100, and 1:1,000 dilutions. One hundred microlitres of each diluted sample was spread-plated on marine 2216E agar and incubated at 30°C for 18 h. The agar plates were investigated in terms of colony morphology including shape, margin, elevation and pigmentation. Morphologically dissimilar colonies were selected and streak plated on marine 2216E agar to obtain pure colonies.

Screening and identi cation of cellulolytic bacteria
Screening of the cellulolytic bacteria was conducted by carboxymethylcellulose (CMC) agar plates and Congo red staining method. Two microlitres of overnight growth culture in the marine 2216E of each bacterial isolate was spot plated on CMC agar (0.2% NaCl, 0.5% CMC sodium salt, 0.67% Na 2 HPO 4 , 0.13% (NH 4 ) 2 SO 4 , 0.05% MgSO 4 ·7H 2 O, and 1.7% agar). The agar plates were incubated at 30°C for 96 h and then ooded with 1mg/mL Congo red for 30 min at room temperature. The cellulolytic isolates were detected by the cellulolytic zone around the colonies after Congo red staining. The HC value determined the cellulolytic activity was calculated from the ratio between the diameter of the cellulolytic zone and the diameter of the bacterial colony. In this study, we only selected colonies with a ratio greater than 1. The selected cellulolytic isolate was identi ed by molecular genetic analysis. The PCR ampli cation and 16S rDNA sequence analysis were previously described [44] using a set of primers as follows: forward primer 27F: GAGTTTGATCATGGCTCAG and reverse primer 1492R: CGGTTACCTTGTTACGACTT. All molecular genetic analyses including PCR ampli cation, 16S rDNA sequence analysis, and homology similarity analysis were carried out by Guangzhou Sangon Biotech Co., Ltd., Guangzhou, China.

Cellulolytic activity assay
The selected isolate was grown in CMC broth (0.2% NaCl, 0.5% CMC sodium salt, 0.5% tryptone, and 0.1% yeast extract) at 30°C for 24 h. Bacterial cells were removed from the culture broth by centrifugation at 5,000 rpm for 15 min at 4°C. The cell-free supernatant obtained after centrifugation served as a crude enzyme solution, which was analyzed for cellulolytic activity assay.
Herein, lter paper activity (FPase) was used to estimate the cellulolytic activity and concentrations of the produced reducing sugars were estimated by the dinitrosalicylicacid (DNS) method using glucose as the standard.
Brie y, cellulolytic activity was measured by incubating 0.5 mL of enzyme solution with 1 mL of substrate solution (a 1×6 cm strip of Whatman No.1 lter paper immersed in 1.0 mL of 100 mM sodium acetate buffer, pH 4.8) at 50°C for 30 min. The reducing sugars liberated were determined by the 3,5-dinitrosalicylicacid (DNS) method. The enzyme reaction was terminated by adding 3.0 mL of DNS reagent and then boiled for 5 min. The solution was completely cooled and the optical density of the reaction mixture was measured at 540 nm. In order to evaluate the hydrolytic activity of cellulase on SB (SBase), only lter paper was replaced by SB powder as the substrate, and the other operations were the same. One unit of FPase or SBase was the quantity of enzymes which produced 1 µmol of glucose or reducing sugar per minute under the speci ed conditions. All experiments of cellulolytic activity assay were performed in three replications and the results were reported as mean ± SD.

Safety of cellulolytic marine bacterial strains to L. vannamei
Before experiments, juvenile L. vannamei were acclimatized for two months with continuous water exchange and constant aeration. The temperature and salinity were maintained at 28°C and 25‰. The shrimps were fed with commercial pellets. Aeration was stopped for about 45 min during feeding, then feces and feed residues at the bottom were removed. Only 20% of the rearing water in each tank was replaced daily, and unhealthy or injured shrimps were removed. After the acclimation period, shrimps were randomly divided into 87 tanks (5-L), each containing 10 shrimps. The shrimps in each tank were not fed for one day, and unhealthy or injured shrimps were removed. Finally, ensure that the number of shrimps in each tank was 10 by constantly replenishing healthy shrimp. During experiments, the temperature and salinity remained constant without replacing water.
The safety of the cellulolytic marine bacterial strains was tested by subjecting healthy shrimps to immersion challenge. The safety experiment for each strain was divided into two experimental groups, each having three replicates. The experimental group was immersed in suspension of nal concentration of about 10 7 CFU mL -1 of rearing water, while the control group was immersed with rearing water without bacteria. All of the shrimps were fed basal diet and the mortality was monitored daily for one week.

Antibiotic susceptibility test
Antibiotic susceptibility tests were conducted by the disc diffusion method. The tests were done by spreading 0.1 mL of 18-h-old broth culture of the test strains on marine 2216E agar plates and then placing antibiotics susceptibility discs (Wenzhou Kangtai Biotechnology Co., Ltd., Wenzhou, China) on the plates. Growth-inhibition zones around the discs were measured after incubation at 30°C for 19 h. In this experiment, 30 kinds of antibiotics belonging to 10 classi es of antibiotics susceptibility discs were selected for antibiotic sensitivity test with three replications (Table S1).
Multiple antibiotic resistance (MAR) index of present isolates against the tested antibiotics was calculated based on the following formula [68]: MAR Index = X/(Y×Z); where: X = Total cases of antibiotic resistance; Y = Total number of antibiotics used in the study; Z = Total number of isolates. A MAR Index value of equal or less than 0.2 was de ned as antibiotics that were seldom or never used.

SB decomposition of cellulolytic marine bacterial strains
The cellulolytic marine bacterial strains were incubated in SB broth (g/L) containing 10 NaCl, 4 SB powder, 0.24 MgSO 4 , 0.011 CaCl 2 , 6.78 Na 2 HPO 4 , 3 KH 2 PO 4 , and 1 NH 4 C at 30°C for 15 days. The culture was ltered by sterilized neutral detergent (g/L) containing 37.2 EDTA, 13.6 sodium perborate, 30 SDS, 10 glycol ether (mL/L), and 23 Na 2 HPO 4 , and then washed by ethanol. The remaining SB (dry weight) in the culture was used to measure the rate of SB decomposition.      Figure 1 Diversity of 97 cellulolytic marine bacteria in the study sites.

Supplementary Files
This is a list of supplementary les associated with this preprint. Click to download. SupportingInformation.docx