Sustainable Shrimp Farming: High Density, Biofloc-Dominated, No-Water-Exchange Systems
Wild-caught seafood supply from the ocean has reached capacity. Aquaculture thus must meet future demand. Like all farming activities, aquaculture impacts the environment. Informed consumers increasingly drive adoption of sustainable practices that reduce discharge of waste, feed ingredients derived from over-harvested fish stocks, antibiotics, excessive use of water, and escape of cultured stock to wild gene pools. This requires a shift from traditional flow-through to Recirculating Aquaculture Systems (RAS).26.08.2019
This article describes a sustainable alternative: high-density, indoor, biofloc-dominated shrimp production with no water exchange based on in situ microbial floc that removes harmful metabolites and provide supplement nutrition.
The demand for protein from increasing world population - together with decreased fishery landings - has resulted in rapid growth of aquaculture. Rapid expansion of the world shrimp farming industry has stimulated the intensification of production systems, which has resulted in the release of nutrients and organic waste, and sometimes the spread of diseases.
Increasing concerns over negative environmental impacts from shrimp farm effluent and diseases outbreaks have led to the development of culture systems with minimal or zero water exchange (Hopkins et al., 1993). These Recirculating Aquaculture Systems (RAS), if properly managed, can reduce or eliminate nutrient and pathogen release into receiving streams. Biofloc systems are a unique type of RAS that maintain a community of suspended microalgae, autotrophic and heterotrophic bacteria which develop in limited-exchange systems (Ray et al., 2010). Operating high-density shrimp culture systems with limited or no water exchange results in production of large volumes of suspended flocculated organic particles referred to as biofloc. The biofloc in these systems have been reported to have beneficial effects on shrimp culture to include: (1) improved water quality (WQ) through removal of toxic nitrogen species; (2) improved feed utilization and shrimp performance from natural productivity (Xu et al., 2012); and (3) enhanced shrimp health through possible probiotic effect (Kim et al., 2014).
Biofloc Composition, Structure and Development
"Biofloc" is a general term that describes an assemblage of living (bacteria, cyanobacteria, algae, fungi, protozoans) and non-living (detritus, uneaten feed, waste products) components that form suspended aggregates also in aquaculture systems. The aggregates (Figure 1) vary in size from the microscopic to > 1 mm with wet-weight density of slightly > 1 g/ml (Sears et al., 2006).
Bacteria typically dominate biofloc aquaculture systems, being abundant (up to 100 million bacteria/mL) and exhibiting high diversity. Floc composition is determined by many factors including temperature, salinity, pH, photoperiod, vertical mixing intensity, and the type of organic carbon available for bacterial metabolism (Emerenciano et al., 2013).
Biofloc develops in newly filled systems soon after organic matter - uneaten feed, shrimp waste, or an added organic compound - has accumulated to a sufficiently high level. The floc development rate can be advanced by ‘boosting’ or adding an organic carbon source to stimulate floc formation (De Schryver et al., 2008).
The nutritional quality of biofloc is related to the carbon-to-nitrogen ratio of culture water, dietary protein level, and light intensity. These and other factors are discussed in detail by Crab et al. (2012), Ekasari et al. (2014), and Martins et al. (2014). The quantity and quality of organic matter stored by bacteria ultimately determine the nutritional value of floc.
The stored organic matter depends on the amount and type of organic carbon available for bacterial metabolism. If the proper organic substrates are provided, then biofloc will store high-quality organic compounds that contribute to the nutritional needs of the shrimp.
On a dry-weight basis, the proximate analysis of biofloc is: protein: 12-50%, lipids: 0.5-41%, carbohydrates: 14-59%, and ash: 3-61.4%. The wide variation in these figures is due to differences in composition between young and mature floc aggregates and the culture conditions.
Marine biofloc typically is rich in the amino acids valine, lysine, leucine, phenylalanine, and threonine, but can be deficient in arginine, methionine, and cysteine, and vitamin C (Crab et al., 2012, Taw, 2012, Ekasari et al., 2014).
Biofloc alone, therefore, is insufficient to guarantee the level of growth and survival required by high-density shrimp culture. Thus, formulated feed is required to satisfy the nutritional requirements of the shrimp in these systems.
Biofloc and Water Quality
Beyond its nutritional value, biofloc also can be managed to improve water quality in culture tanks.
There are two main types of microorganisms in these systems: autotrophs and heterotrophs. Autotrophs are organisms that produce organic compounds from inorganic compounds. They obtain carbon from inorganic carbon sources such as carbon dioxide and bicarbonate. They are further classified as photoautotrophs, that derive energy from sunlight, and chemoautotrophs that derive energy from inorganic chemical compounds. In aquatic environments, the former are the algae, and the latter include bacteria such as the nitrifiers that obtain energy from oxidizing ammonia to nitrate. Heterotrophs on the other hand obtain carbon from organic carbon sources.
Both autotrophs and heterotrophs found in biofloc improve WQ by assimilating dissolved inorganic nitrogen compounds (ammonia, nitrite, and nitrate) that are harmful to shrimp. To this end, a biofloc-dominated system can be managed to favor autotrophic bacteria, heterotrophic bacteria, or some combination of the two.
Biofloc and Immune Response
Shrimp have a non-specific, labile immune system, meaning they have no specific immune mechanism to respond to new pathogens entering production units (Roch, 1999). However, the dense microbial population in biofloc systems may play a role in activating the non-specific shrimp immune system, resulting in a type of defense that may permit a quick response against bacterial diseases (Kim et al., 2014). Biofloc also has a probiotic effect, in which short-chain fatty acids (lipopolysaccharides, peptidoglycans, and β -1, 3-glucans) in bacterial and fungal cell walls play a role (Crab et al., 2012). Biofloc microorganisms also suppress pathogen growth by competing for space, substrate, and nutrients, and by excreting inhibiting compounds (Emerenciano et al., 2013).
Autotrophic Versus Heterotrophic Systems
Biofloc consists of a mix of chemoautotrophic and heterotrophic bacteria and photoautotrophic algae. Among the chemoautotrophs, one group of nitrifiers oxidizes ammonium (NH4) to nitrite (NO2) and another oxidizes nitrite to nitrate (NO3). The NO3 end-product is less harmful to shrimp than either NH4 or NO2, so it can accumulate to higher levels before requiring removal. In no exchange systems, continued use of the same culture water leads to accumulation of total nitrogen, usually as nitrate. Nitrate can be removed by water exchange or through a denitrification process in which denitrifying bacteria reduce NO3 to N2 (Van Rijn et al., 2006).
Biofloc systems favor development of heterotrophic bacteria when the C:N ratio is high.
These bacteria can rapidly remove ammonia from culture water. The biomass production per unit nitrogen of heterotrophs is about 40 times greater, with greater O2 consumption and CO2 production per unit nitrogen than nitrifiers. Feeding shrimp high-protein feed without supplemental organic carbon results in low C:N ratio which favors development of chemoautotrophic bacteria, including nitrifying bacteria, which oxidize ammonia to nitrate and reduce alkalinity.
Under no water exchange, sufficient organic carbon is available from the feed and shrimp waste for heterotrophic bacteria to metabolize approximately 1/3 of the ammonia. The remaining 2/3 is available for nitrification by chemoautotrophs (Ebeling et al., 2006).
Compared to a purely heterotrophic biofloc system, the mixotrophic system demands less oxygen, requires fewer carbohydrate additions, generates less CO2, and produces lower microbial biomass. If supplemental organic carbon is added to the culture the system would shift toward a more heterotrophic regime. Biofloc-dominated systems, such as the one developed at the Texas A&M AgriLife Research Mariculture Lab (ARML), rely on floc aggregates that contain heterotrophic bacteria and nitrifying bacteria. This “mixotrophic” system provides an efficient way to improve WQ and supplement shrimp diet.
Evolution of Recirculating Aquaculture Systems
The expansion of the shrimp farming industry has stimulated the intensification of the production systems. Recently, these systems started to incorporate Biofloc Technology (BFT), which can reduce facility footprint, water use, nutrient releases, escape of exotic species, and spread of pathogens to the environment (Samocha et al., 2010 and 2012).
Advantages and Disadvantages of Indoor Biofloc Systems
Some advantages of these systems are: water conservation, stable WQ, reduced fertilizer use, small footprint, year-round production, faster growth, lower disease susceptibility with greater biosecurity, more efficient use of protein in feed, lower feed requirements, higher yields, and sustainability. Some disadvantages of these systems are: high capital investment per unit area, high energy input, power failure is critical, technical operating complexity, potential exposure to toxins, and disease risk.
Key Factors to Consider in the Design and Operation of Biofloc Production System
Biofloc production systems must have tools in place to monitor and maintain WQ within the optimum range as sub-optimal conditions can reduce shrimp performance. The WQ parameters that should be monitored and controlled include dissolved oxygen (DO), temperature, nitrogen species (TAN, NO2, and NO3), pH, alkalinity, salinity, solids, ionic composition, and heavy metal concentrations.
Other components that significantly impact the economic viability of biofloc systems include: location, water availability and required pre-treatment, building type, size and type of culture tanks, availability of DO monitoring and control systems, methods used to maintain DO and to keep biofloc suspended, cost and availability of power and adequate power backup, availability and cost of temperature control, availability and type of feed storage, feed quality and feeding practices, theft and predator control, biosecurity protocols, solid control and waste disposal methods, water storage, post-harvest water treatment methods, harvest and product handling procedures, quality control, availability of cold storage, and marketing program.
Careful examination and selection of these components together with a well-trained workforce are critical for building and operating economically viable super-intensive, no exchange, biofloc-dominated shrimp production systems.
The Texas A&M-ARML Super-intensive, Biofloc-dominated, No Water Exchange, Shrimp Production System
The Texas A&M-ARML in Corpus Christi, Texas, USA has been a leading facility in developing super-intensive, biofloc-dominated, no-water exchange, shrimp production systems. Shrimp production trials at this facility were performed in two systems both located in greenhouses with no active water temperature control. Exhaust fans and roof shade cloths served to lower water temperatures during summer. Passive heat retention by the greenhouse enabled extension of the production trials from early spring to late fall in this temperate climate location.
The first experimental system had six raceways (RWs - Figure 2) lined with 1 mm ethylene propylene diene monomer membrane (EPDM, Firestone Specialty Products, Indianapolis, IN, US).
Each RW had a surface area of 103.7 m2 and working water volume of 40 m3 with 0.45 m average water depth and 0.5% slope. Every RW had five boardwalks and anti-jump netting around it. Water circulation, mixing and oxygenation were provided by one 2 hp pump, six 0.9 m long air diffusers, and eighteen 5 cm airlift pumps. A center partition was positioned over a 5 cm PVC pipe with spray nozzles for homogenous distribution of oxygenated water from a 5 cm pump-driven Venturi injector (Figure 4). All six RWs were equipped with an online DO monitoring system (5500D, YSI Inc., Yellow Springs, OH, USA, - Figure 3). Each RW had a settling tank (ST, Figure 5), a foam fractionator (FF, Figure 6), and a multi-cyclone filter (MCF, Figure 7) for solids control. Separation tanks helped dry the cropped-bioflocs to facilitate disposal and conserve water (Figure 8). Six spring-loaded belt feeders were used to deliver feed 24/7 in each RW.
Biofloc systems operated with air blowers alone are generally capable of producing shrimp yields between 2 to 4 kg/m3. The pump-driven Venturi injector in each RW enabled shrimp yield of > 9.5 kg/m3 when supplemented with pure oxygen.
The second system consisted of two 100 m3 RWs (100 m2 surface area) lined with the EPDM membrane used in the other system. Unlike the other system, these RWs were built half buried above the natural ground level and had a deeper average water depth (1 m). In addition, the RWs had a common concrete harvest basin outside the greenhouse which enabled the use of fish pump for harvest.
Similar to the other system, each RW had anti-jump netting, boardwalks, belt feeders, and a center partition (Figure 9). Unlike the other system, aeration and water mixing were generated solely by two 2 hp pumps per RW forcing high-pressure water (45 psi) at a flow rate of 28.4 Lpm through each of the 14 injectors (a3® All-Aqua Aeration, Orlando, FL, USA, Figure 10).
Biofloc concentrations were maintained with help of ST (Figure 11) and FF (Figure 12) using a similar setup as the other system for drying and disposing of biofloc. RWs were equipped with the same online DO monitoring system described earlier. The two 2 hp pumps and 14 a3 injectors in each RW supported marketable size shrimp yields >9 kg/m3 while using only atmospheric air to satisfy DO demand.
40 m3 Raceway System
All but one nursery trial were conducted in the 40 m3 RW system. Nursery tank stocking densities varied from a few hundred to several thousand post-larvae (PL) per m3. Trials initiated with virgin water, showed high TAN (>26 mg/L) and NO2-N (35 mg/L) concentrations did not negatively impact shrimp performance. The results of the nursery trials suggested the following operational WQ ranges: TSS (250-350 mg/L), SS (10-14 mL/L), and alkalinity (140-160 mg/L CaCO3). Continual DO monitoring in the nursery RWs using the online DO monitoring system helped optimize feed delivery. Other studies determined the preferable feed and feed particle size for different size PL. Additional trials helped define the role of different commercial feeds and feed management on shrimp performance.
Routine probiotic applications into the culture water suggested potential benefit through suppressing pathogenic Vibrio development in the culture medium. Other studies showed that high TAN and NO2 levels can be avoided by inoculating virgin water with nitrifying bacteria. Finally, the studies showed that the 40 m3 RW system could produce juvenile shrimp of about 2 g with high survival, low FCR, and high yields under no water exchange (Table 1).
|Treatment||Size at harvest (g)||Yield (kg/m3)||Survival (%)||FCR|
|FF-11||1.9 a||7.6 a||100 a||0.97 a|
|FF-2 1||2.0 a||6.9 a||92 a||1.08 a|
|WE-1 2||1.7 b||3.9 b||56 *||1.64 a|
|WE-2 2||1.4 b||4.7 b||82 a||1.36 a|
1 RW operated with FF and 3.35% daily water exchange.
2 RW operated with no FF and 9.37% daily water exchange.
Values with similar superscripts are not significantly different (α=0.05).
100 m3 Raceway System
The nursery trial conducted in the two 100 m3 RWs showed that a3 injectors can be used from stocking without damaging young PLs. Table 2 shows the excellent results obtained in a 62-day nursery trial in the two 100 m3 RWs stocked with L. vannamei at 540 PL/m3. FCR was low, with excellent survival, and yield >3.4 kg/m3. Only one of the two pumps per RW was required to maintain DO and mixing.
|RW 1||RW 2|
|Final weight (g)||6.5||6.4|
|Water use (L/kg)||420||447|
|Sugar added (kg/m3)||0.33||0.33|
|Bicarbonate added (kg/m3)||0.26||0.25|
40 m3 RW System
Numerous studies were conducted from 1999 to 2014 in this system re-using water used in previous nursery trials. The nitrifying-rich bacteria in this water kept TAN and NO2-N concentrations low (usually 1-6 mg/L) while maintaining high shrimp yields (>9.5 kg/m3) without organic carbon supplementation. Nitrate levels increased throughout the trials (from about 40 mg/L to about 450 mg/L NO3-N).
Although shrimp can tolerate NO3-N concentrations up to 450 mg/L at 30 ppt salinity, if the culture water is to be reused for subsequent production, this water must be treated to reduce nitrate levels to safe concentrations. Trials showed that NO3-N concentrations can be reduced to below 50 mg/L using a digester operated under anaerobic conditions. This process could be accelerated through organic carbon supplementation.
Nitrification reduces alkalinity and pH so these parameters were regularly adjusted to minimize stress on nitrifying bacteria and shrimp. Alkalinity was increased with sodium bicarbonate, sodium carbonate, or calcium carbonate. Although shrimp can tolerate alkalinity above 400 mg/L as CaCO3, efforts were made to keep alkalinity concentrations between 120 and 160 mg/L. pH in these systems tends to drop from about 8.2 to below 7 during the production cycle, but the alkalinity supplementations, in most cases helped maintain the pH at about 7.4. Nevertheless, when supplementations with the above mentioned carbonate sources could not increase pH, sodium hydroxide was added. Under high stocking densities (250-500 juveniles/m3) and no water exchange, when shrimp are fed dry feed 24/7, Total Suspended Solids (TSS) and Settleable Solids (SS) increase. High concentrations of these indicators can reduce shrimp performance. Trials evaluated and selected the most cost-effective tools to maintain optimal concentrations for shrimp (e.g., pressurized sand filters, FFs, STs, MCF’s, etc.). These trials showed that although shrimp grew well at relatively high particulate loads (e.g., >500 mg/L and 30 mL/L for TSS and SS, respectively), targeted levels were set between 250 and 350 mg/L and 10 mL/L and 14 mL/L, respectively. These lower concentrations provide system operators enough time to make any adjustments to bring the particulate load to the desired levels. Other trials with feed of different quality and price showed better shrimp performance with specially formulated high-quality feed compared to feed formulated for semi-intensive outdoor production ponds (Table 3).
1 HI-35: High-quality 35% CP feed for super-intensive shrimp systems (HI-35, Zeigler Bros., Gardners, PA).
2 SI-35: 35% CP feed for semi-intensive production systems (SI-35, Zeigler Bros., Gardners, PA).
Information from an online DO monitoring system showed significant DO reductions when feed was delivered a few times a day. This was corrected by switching to 24/7 feed delivery. This practice, together with the use of high quality feed and careful monitoring of feed consumption, enabled production of marketable size shrimp with good growth (1.8-2.0 g/wk), survival (79%-88%), yields (9.4-9.9 kg/m3), and low FCR (1.39-1.45).
Table 4 summarizes results from a 2011 trial in the 40 m3 RW system. Note that pure oxygen was supplied through the Venturi injector to support the high oxygen demand of the system.
|RW||Average Stocking||Weight (g)|
The trials showed that excellent results can be achieved when shrimp are exposed to optimal growing conditions. However, this system requires constant monitoring and adjustments to avoid stressing the shrimp. Exposing shrimp to unfavorable conditions for a few days can lead to pathogenic bacteria outbreaks and poor results.
100 m3 RW System
Monitoring and control of key WQ indicators in this system were similar to those used for the 40 m3 system. The grow-out trials in this system were designed to evaluate the effect of different stocking densities, different particulate matter control tools, and feed management on shrimp performance. These trials showed that homemade FFs and STs could control particulate matter concentrations.
As with the 40 m3 system, when nitrifying bacteria rich water was used for the grow-out trials under no water exchange and with adequate alkalinity and pH adjustment, TAN and NO2 concentrations stayed low while NO3 levels increased. Since aeration and mixing in these RWs was generated by the pump-driven a3 injectors, fine-tuning of the water flow was needed to improve shrimp performance (minimize shrimp energy expenditure on constant swimming against the current generated by the injectors). This together with 24/7 feed delivery and the delivery of the feed away from the pumps’ intakes helped reduce the FCR from as high as 2.56 to 1.43.
Other trials also showed that shrimp growth rates were greatly improved (increase from 1.38 to 2.31 g/wk) when RWs were stocked with juveniles produced by breeding populations of two genetic strains (Fast-Growth × Taura Resistant) compared with those produced by a Taura resistant strain. A significant finding from these trials was the high yields (> 9 kg/m3) achieved without using pure oxygen (e.g., the system oxygen demand was met by operating the a3 injectors using atmospheric air). Table 5 summarizes the results from a trial conducted in this system in 2012.
|RW||(Juveniles/m3) (g)||(g)||Harvest (g)||Growth (g/wk)||Survival (%)||Yield (kg/m3)||FCR||Water Use (L/kg)|
Economics of RAS and Biofloc Systems
Table 6 compares production costs in earthen ponds and RAS using data from the USDA-funded US Marine Shrimp Farming Program. Pond data are from an intensive farm in Arroyo City, Texas and RAS trials conducted at the Oceanic Institute, Hawaii over four years (2005, 2006, 2007, and 2009).
|System size (m2)||20,234||n/a||58.4||75||337||40|
|Harvest weight (g)||18.0||23.0||17.9||21.0||18.3||23.1|
Production costs per unit shrimp were less in RAS than in earthen ponds. Even at higher stocking densities, survival and growth in the RAS trials were better than in ponds. At harvest, shrimp produced in RAS were just as large and in some cases even larger than shrimp from ponds.
Closed, indoor super-intensive RAS can be operated for less than earthen ponds (Moss and Leung, 2006). Table 6 shows cost comparison between a farm and a closed, indoor super-intensive RAS system. The cumulative distribution of total cost for ponds and RAS indicates that RAS has a lower cost per unit weight than ponds.
Economic projections suggest that biofloc systems in the USA can be profitable when they target niche markets for live or fresh (never frozen) shrimp (Hanson and Posadas, 2004, Hanson et al., 2013). The Texas A&M-ARML has reduced indoor biofloc operating costs from $11.00/kg, the USA average for super-intensive systems, to about $4.53/kg. This work also suggests the feasibility of extending the number of crops from 3.5 to 5.5 per year.
Current Issues with Indoor Biofloc Shrimp Culture
Advances in indoor biofloc systems have been impressive, but current knowledge is not complete. For example, the failure of some indoor biofloc projects can be traced to the complex interrelationships that characterize the diverse and difficult-to-control microbial biofloc community. This assemblage can be unstable in small tanks stocked at high densities and driven by the large input of feed required for good shrimp growth. If the microbial community of the biofloc system is not balanced properly, inimical chemicals can accumulate, particularly TAN, NO2 and NO3. WQ changes are exacerbated when water is reused over multiple crop cycles.
Biofloc systems also are susceptible to outbreaks of noxious organisms, such as Fusarium solani (responsible for closure of a commercial facility in Kentucky) and Vibrio sp. (which caused a commercial operation in Texas to abandon biofloc). The Waddell Mariculture Center research facility has experienced outbreaks of the cyanobacterium Synechococcus sp. and the dinoflagellates Gymnodinium sp. and Pfiesteria piscicida, each with an unpredictable and decidedly negative impact on production. The Texas A&M-ARML indoor biofloc systems also have experienced crop-threatening outbreaks of Vibrio.
The systems and the results described in this paper suggest the need for the refinement of production management of super-intensive, biofloc-dominated operated with no water exchange to make these system more economically viable.