Mitigating Off-Flavour in Tilapia using PE-lined Pond RAS Technology with Shading

Earthen ponds are most commonly used for grow-out of Nile tilapia in the tropics, as it is often the most cost-effective method for rearing this species.  The downside is that yields are fairly low, at between 4-7 tonnes per hectare, since there is a limit to the rate at which phytoplankton, the main utilizers of waste in this system, can remove nitrogen.  Increasing levels of phytoplankton leads to reduce sunlight penetration of the water column.  Since they rely on sunlight for photosynthesis, they inhibit their own growth as they become too dense. Furthermore, muddy or musty off flavors are common due to geosmin and 2-MIB metabolites produced by actynomycetes and cyanobacteria (blue-green algae), the latter of which are very common phytoplankton in tilapia ponds.  This reduces the market value of the fish produced and makes export of tilapia to countries that eat non-spicy food highly problematic.  Water exchange can be used as a solution to both these problems, but pumping costs can be high and discharge of waste water into waterways is not an acceptable practice in many countries.

BIOFLOC VERSUS RAS

Nam Sai Farms, established in Prachinburi Province, Thailand in 1994, is a large 100 ha tilapia hatchery. Nam Sai has conducted various trials over the years, some in collaboration with the Institute of Aquaculture, University of Stirling, looking at increasing density of tilapia and removing wastes by use of biofloc or simple recirculating aquaculture systems (RAS). From earlier research it was found that muddy off-flavour was found to be very common in tilapia produced in earthen ponds in central Thailand, particularly in the dry season. The acid clay soil and use of green water grow-out systems, in which blue-green algae are very common, are likely contributory factors to the off-flavour problem.

Two alternative grow-out systems, biofloc technology (BFT) and simple recirculating aquaculture systems (RAS) were tested in a collaborative trial in 2006. The results of the study showed that high suspended solids loading when using biofloc methods to remove nitrogenous waste was detrimental to fish growth and health of the fish.  Tilapia growth and food conversion performance was found to be much better in RAS systems which involved removal of suspended solids by settling and removal of dissolved ammonia by nitrifying bacteria in a biofilter based on plastic biomedia.  The downside of the latter was the high pumping costs.

RAS POND DESIGN

Based on these experiences, Nam Sai came up with an RAS pond system in which fish waste products could be removed in the pond itself, without the necessity of pumping water through a filtration system and using only blowers for aeration and water movement.  Two shading levels were employed using a single or double layer 60% shade netting to control or eliminate phytoplankton to hopefully eliminate off-flavor problems. This is used as a strategy to mitigate off-flavour; 68.6% of sunlight was reduced using a single layer of shade netting, and 99.4% with a double layer. The ponds were lined with 1 mm polyethylene to prevent soil water turbidity, improve water quality and avoid any off-flavors arising via the earthen pond bottom.

Plastic biomedia (12 mm diameter 4 spoked wheel), with a reported surface area of 900 m² per m³, was chosen as a substrate for nitrifying bacteria that convert toxic dissolved ammonia to nitrite and then to non-toxic nitrate.  Enclosed in a two meter circular cage in the middle of each pond, the biomedia was constantly circulated by rising air bubbles emanating from aerotube attached to the bottom of the cage.  The cage itself was raised on legs one inch above the pond bottom. This allowed settled solid wastes, consisting mainly of fish faeces, to move towards a central 8” PVC pipe where they could be sucked out with a submersible pump every few days. 

Aeration and some circular flow in the pond was created using a cheap and simple air lift design consisting of a pillow-shaped diffuser at the bottom of a square column (60cm wide x 30 cm deep, height depending on water depth) formed using stainless steel rod and covered in an air-proof material.  A gap at the bottom on one side allowed water intake and a curved top and exit hole on the opposite side allowed directional water flow.  The whole thing was held down on slab of concrete to stop it floating away, but could be detached to enable netting of the pond. 

TRIAL DESIGN

A total of six ponds were constructed.  Three ponds were covered a single layer of 60% black shade netting and three were covered with two layers to remove the majority of light.  There were 8 air lift pumps and a single biofilter installed in each pond.  This allowed three replicates of each shade treatment and was compared to a control treatment of three earthen ponds that were unshaded, stocked at low density and un-aerated. 

Red tilapia (F1 hybrid Taiwanese x Thai strain) were stocked in all PE-lined ponds at a relatively high density of 2.35 fish/m³ and in earthen ponds at a regular density of 0.85 fish/m³.   A single 3 x 4 x 1.2 m deep cage was stocked with 310 fish as a second control for taste testing purposes. 

The fish were raised from 107g up to 600g over a period of 4 months from 4^th November 2013 to 29^th March 2014 during the cold season in Thailand.  Feeding was carried out by feeding to demand for 15 minutes and was given three times per day with floating commercial feed containing 30% crude protein and 6% fat content.  All ponds were run under zero-exchange conditions and water was only added to replace that lost during removal of settled solids and evaporation.  Sampling of fish for individual size and length was initially carried out every 2 weeks, but was reduced to 4 weeks after a month, as the sampling caused stress to the fish and reduced feed consumption. 

The table below summarizes the trial design/pond parameters:

• 

At the end of the trial all fish were weighed and sold. An organoleptic assessment was made at Nam Sai by two taste testers sent by Grobest Corporation Ltd. (Thailand).   The trial sample fish were compared with cage-reared red tilapia from Nam Sai reservoir and the nearby Ban Pakong River and also fish from Tesco Lotus in Prachinburi town.  Fish were prepared by filleting and dividing into approximately 12g samples before microwaving in a food grade, microwave-proof HDPE plastic container.

FISH PERFORMANCE AND WATER QUALITY

Red tilapia grown in low and high shaded PE-lined ponds exhibited very similar performance in terms of growth, survival, FCR and overall production.  Growth in earthen ponds was slower, despite being stocked at much lower density, and survival poor.  FCR as a result was very high.  Interestingly, for the first 3 months, growth of fish in the earthen ponds was similar to that in the PE ponds, but then slowed down.  This could have been attributed to increasingly high total suspended solids, particularly in pond E8, due to suspended clay particles. Chlorophyll A levels were very low in both pond E4 and E8 and inorganic fertilizer (15-15-15) was added to all earthen ponds from February onwards to try and improve phytoplankton levels and general water quality.  Unfortunately, the high clay turbidity was not favorable for light penetration and was only successful in pond E6.

• 

It can be seen from the graph of feed consumption against mean daily water temperature that between mid-December and the end of January mean daily temperature was only 21-26^oC.  Feed consumption was depressed and this no doubt reduce growth rate and increased FCR. 

Severe mortalities occurred in the earthen ponds after the onset of cold weather, most likely due to bacterial disease.  All fish in both low and high shaded PE ponds were unaffected despite water temperature being lower. The single mortality spike in the high shaded treatment was due to blower failure.  Similar blower failure in low shaded ponds surprisingly had less impact as DO levels were higher due to photosynthesis by phytoplankton.

Better performance would have been expected if the trial had been carried out when temperatures were more ideal.  Daily temperature fluctuation and mean daily temperature reduced with increasing level of shading.  This would have big advantages in controlling Streptococcus agalactiae disease which typically causes significant tilapia mortality during the hot season from April to July in Thailand when water temperatures regularly exceed 35^oC in unshaded ponds. 

PAR (photosynthetically active radiation) readings indicated that 68.6% of the sunlight was removed in the low shaded ponds.  This was still sufficient to develop a dense phytoplankton bloom with mean chlorophyll a levels of 299 µg/litre.  The double layer of shade netting in the high shaded ponds was sufficient to remove 99.4% of the light and mean chlorophyll a levels were only 12 µg/litre. 

Ammonia levels were unexpectedly higher in the earthen ponds and it was evident that the biomedia in the PE-lined ponds was found to be colonized by nitrifying bacteria very quickly and was very efficient at converting ammonia to nitrate.  No spikes of ammonia were experienced in any of the PE-lined ponds and only a few short-lived small spikes of nitrite.  Nitrate levels were highest in the high shaded ponds, as nitrogen was removed by phytoplankton in the low shaded and earthen ponds.    

Dissolved oxygen levels were maintained above 4.5 mg/liter in both high and low shaded PE-lined ponds, whilst in earthen ponds the DO dropped low in the early morning since no aeration was provided.

TASTE TEST RESULTS

Note: Total scores >85 = Very good, 75-84 = Good, 61-74 = Acceptable, ≤60 = Reject

The odour, taste and texture results carried out by Grobest Corporation Ltd. staff indicated that fish were very good from all ponds, including the earthen ponds.  This shows that tilapia grown in earthen ponds can taste very good.  One of the reasons in this case could have been the low levels of phytoplankton due to the high soil turbidity.  Interestingly, fish from the low shading ponds, which had very high phytoplankton levels, scored very high.  This was unexpected since high phytoplankton (cyanobacteria) has been related to episodes of off-flavor.  It is not the only reason however, and in recirculating systems (RAS), Streptomyces and Nocardia growing as bio-films on tanks, in filters and other components, have been reported to be sources off-flavor.

CONCLUSIONS

This trial indicated that the RAS PE-lined pond technology certainly has great potential as a grow-out or fattening system to produce tilapia with good taste suitable for export.  Final density of fish at harvest was approximately 1.4 kg of fish per m3, which is three times the density possible in earthen ponds without water exchange.  The downside is the cost of construction and electricity needed for aeration.  Savings were made in terms of the reduced amount of land required for rearing, reduced losses due to disease and easier management, since more fish are raised in a smaller space.

FUTURE DEVELOPMENT OF THE SYSTEM

Nam Sai now has a plan to modify and upscale the system to enable efficient removal of solids, reduce costs of shading, improve efficiency of aeration, provide good water flow for fish exercise purposes and allow easy management of the system.  Alarm systems and better blower back-up systems would be incorporated to reduce the risk of aeration system failure.

Based on a 6,000 m3 design, the new system could produce anything between 30-120 metric tons of fish per pond based on a production level of 5-20 kg/m3.  Nam Sai and Stirling University are now looking for a private partner who would be willing to invest in 6 of these ponds so that research work can go ahead.

Acknowledgements

Mr. Warren A. Turner, the managing director of Nam Sai Farms Co. Ltd., was responsible for the system design with assistance from Mr. Renatus Remmerswaal, Aquaculture and Engineering Consultancy. Email: warrennamsai@gmail.com

Dr. David C. Little from the University of Stirling was involved in planning, monitoring and analysis of data.

Mr. Nathan Atkinson was responsible for day-to-day running

Mr. Michael Kreis organised the taste-testing as part of his MSc. in Aquaculture at the university.

Arlene Nietes Satapornvanit and Dr Kriengkrai Satapornvanit provided support as part of the Sustaining Ethical Aquaculture Trade (SEAT) project 466, which is cofunded by the European Commission within the Seventh Framework 467 Programme, Sustainable Development
Global Change and Ecosystem (project 468 no. 222889).
http://www.seatglobal.eu