Aquaculture in Bangladesh: prospect of high density mixed culture of fish with low cost diets. (byM.A. Hossain)

MFTS, BFRI. Motel Road, Cox's Bazar, Bangladesh

ABSTRACT

To find out the maximum production potential of pond based aquaculture using low-cost diets on mono and mixed culture of three selected species experiments were conducted for a period of 18 months. Nine miniponds were used for monoculture of catfish (Pangasius sutchi), silver barb, (Barbodes gonionotus) and Genetically Improved Farmed Tilapia (GIFT), with three replications. Three other miniponds were used for mixed culture of the three above-mentioned fish. Initially, specific numbers of fish were stocked in miniponds and daily a mixture of low-cost supplemental feeds were applied to each group of ponds. Production of natural food in the investigated ponds was promoted to minimize supplemental feeding. Partial harvests of both stocked fish and newly produced ones were done periodically. Low-cost feed included pellets, mustard oil cake, wheat bran and rice bran with the same proportions (25%) for each trial. For the whole period of culture 1.5 m depth and deep green water colour were maintained by occasional intake or discharge of water and periodic fertilization. Two groups of fish, silver barb and GIFT, produced numerous new individuals both in mono and mixed culture. Cumulative total fish biomass production in each trial varied from 16.1 to 31.6 tons/ha with a feed conversion rate of 20.1 to 24.4% within the investigated period of 18 months. However, the projected feed conversion rate for a 12-month period was 31.0 to 50.4%. A huge cumulative amount of supplementary feed utilization (57.1 - 134.4 tons/ha/18 months) and high level of both plankton and fish biomass occasionally lowered the dissolved oxygen level to 0.7 mg/L but caused no fish mortality. So-called black soil formation at pond bottom was also minimal. The high level of fish biomass production resulted from the round-the-year equal water depth, presence of copious green algae, no fish mortality, negligible deoxygenation of the sediment resulting from the continuous stirring up of sediments by fish, and the ability of all three fish species to gulp air at the surface when the dissolved oxygen concentration declined.

1. INTRODUCTION

Tilapia (Oreochromis mossambicus) was introduced to Bangladesh from Thailand in 1954 (Rahman, 1985). It was expected at that time that tilapia would act as a miracle fish in aquaculture. While tilapias breed several times a year and depend mostly on vegetative food, their distribution and culture among rural farmers has not been as popular as expected. Gradually, nilotica (O. niloticus) and red tilapia (a mutant of O. niloticus x O. mossambicus) were imported to Bangladesh from Thailand (Gupta et al., 1992). Genetically Improved Farmed Tilapia strain known as GIFT (Eknath et al., 1993) was introduced to Bangladesh by ICLARM and BFRI during 1994. GIFT has now become a popular fish among farmers. Likewise, Thai silver barb (Barbodes gonionotus) and catfish (Pangasius sutchi) entered Bangladesh during 1985 and 1988, respectively, and soon became popular aquaculture species.

Seed production of these three species is easy. GIFT and silver barb naturally breed in closed water bodies. Seasonal water fluctuation in culture ponds due to seepage, and evaporation and turbidity during culture time is a problem in pond-based aquaculture. In view of the 21st century perspective, a minipond complex was constructed at the Marine Fisheries and Technology Station (MFTS), Cox's Bazar on sandy soils using concrete walls and polythene bottom laying (Hossain et al., 1994). The structural design made it possible to maintain water depth to a desirable level the year round and effectively stopped seepage and turbidity problems. Once water level and turbidity level stabilized natural productivity of the ponds dramatically improved and it is now possible to maintain rich phytoplankton for a long time with little initial fertilization.

Experimentally, mass production of both tilapia and GIFT as live food for sea bass showed very high rate of biomass production in miniponds with low-cost diets. Similarly, silver barb and Thai catfish production in those miniponds also showed a fairly high rate of biomass production (Hossain et al., 1999).

Next, an attempt was made to find out maximum production of these species individually and collectively using low-cost diets under pond-based aquaculture. Attention was focused on maximum utilization of natural productivity, and a low level of supplemental feeds was applied. The experiment was conducted for a period of 18 months and arrangements were made to discard or add water when necessary.

Twelve miniponds, each having identical size (200 m²), depth (1.6 m) and standard bottom conditions (30 cm sand layer on polythene) were used. Three fish species were reared on a monoculture basis in different groups and a mixed culture using all three species in another group. To simplify management, a certain amount of daily supplemental feeds was allotted for each group for a 2-month period. Cummulative harvesting followed and at the end of the experiment, ponds were drained and harvested.

2. MATERIAL AND METHODS

Miniponds

12 miniponds out of 42 at MFTS were used. As mentioned earlier these ponds were identical in size, and constructed on totally sandy soil. Pond bottoms were level, underlaid by polythene. On polythene, a 30 cm layer of sand was deposited. Length, width and height of the miniponds were 25.0, 8.0 and 1.6 m, respectively. Water holding capacity of each minipond was roughly 320 tons. Each minipond was exactly 1/50th of a hectare. The 12 ponds were divided into 4 groups of 3 ponds each.

Pond preparation

Before starting the experiment all ponds were drained, the bottom dried and sand agitated. Calcium carbonate at a rate of 5 kg/pond (250 kg/ha) was mixed with bottom soil. Then the ponds were filled with ground water, and the water was fertilized with urea, TSP and potash at a rate of 1.0, 1.5 and 0.5 kg/pond, respectively (50, 75 and 25 kg/ha). Initially, some green water was pumped from the nearby ponds to each of the experimental ponds with the help of a submergible pump. When water turned deep green, fish were stocked.

Fish

Three species were used. GIFT juveniles were produced at the Freshwater Station of the BFRI with joint collaboration of ICLARM, who delivered them to the MFTS in 1996. Several successive generations were produced at MFTS. Adult fish of 80-100 g were used in the experiment. Each of three ponds were stocked with 100 GIFT. Silver barb juveniles raised to adult stage at MFTS were also used. 100 silver barb with a body weight range of 100-150g were stocked in each pond of a group. Similarly, catfish raised to 350-400 g using high density culture, were used in another group at a rate of 100 fish/pond. The remaining of 3 ponds were each stocked with a mixture of the 3 species (a total of 100 fish in each pond). All fish were stocked at the beginning of April, 1999. Stocking particulars are shown in Table 1.

Table 1

Details on stocking each pond of a group

Feed

Four types of feeds were mixed together for each pond. The mixture consisted of commercial pellets (Saudi-Bangla Fish Feed Ltd.), mustard oil cake, rice bran and wheat bran. The feeds were mixed together for a couple of days, and divided into 12 parts and stored in medium plastic containers for each pond. In the container the feeds were dry, but before application to ponds a portion of water was mixed to give it a semi-solid form. Fish in ponds were fed twice daily.

Feed ration

To minimize the use of supplemental feeding and to encourage fish to depend on natural foods produced in the ponds, a strict feed ration was followed as shown in Table 2. The feed ration was modified for each two-month period and the daily ration was strictly maintained to avoid management problems.

Rearing

The culture continued for an 18-month period. Care was taken to maintain identical water level in each pond through inspections. This was ensured by pumping ground water through a pipeline network when the water level went down. During rains, excess water was automatically discarded through piped outlets connected to a drain. Outlets were guarded by nets to prevent fish from escaping. After stocking of fish, a current was occasionally produced in the ponds having silver barb to induce them to breed in captivity. For this purpose water was pumped and a part of it was then discarded from the ponds through outlets. Twice a day, a fixed amount of supplemental feed was provided to each group as shown in Table 2. Identical amount of feed was supplied in each pond of a group. Water quality parameters, such as temperature, transparency, pH, and dissolved oxygen levels in ponds, were recorded from time to time. Whenever transparency of water rose above 25 cm, water was fertilized by urea and TSP (50% of the initial dose given during the pond preparation).

Table 2

Feed ration given to each pond of a group (kg/day)

Water quality analysis

Temperature was recorded by ordinary thermometer, pH by digital pH meter, water transparency by Secchi disc and DO by Hach kit. After 10-11 months of rearing, when the biomass of fish increased rapidly, DO was measured more frequently in the early mornings.

Non-target fish

Ponds designated for silver barb or catfish were invaded by GIFT, which were periodically harvested using seine net. Total weight of non-target fish harvested from each pond was recorded separately. Harvested fingerlings were weighed and their numbers were estimated. To get rid of non-target fish from ponds, dry feed particles were scattered to lure the fish to the surface and the fish were harvested into a fine mesh seine net.

Harvesting

Cumulative harvesting was done in 3 groups of ponds where numerous fry were found. Fingerlings and fry, specially those of GIFT and silver barb, were harvested 3-12 months after stocking. After 12 months there was no selective harvesting. The final harvest was done at the end of the 18-month period. Catfish were partially harvested at the end of 12 months both in mono and mixed culture groups. In mixed culture groups 25% of catfish were harvested before the final harvest. At the end of the experiment, all ponds were drained and all fish harvested. Fish harvested from each pond during the culture period or at the end of the experiment were recorded separately after weighing. To estimate the projected fish biomass in each pond of a group within a 12-month period, fish harvested from July 1999 to March 2000 were recorded and no further harvest during 13-18 months was done (Table 3). The projected figure includes actual harvest till the 12th month and an estimated quantity of fish remaining in each pond thereafter. Projected production/year is presented in Table 6.

Table 3

Cumulative harvest of fish from each pond of a group (average)

* Included significant number of non-targeted species, mostly GIFT.

Figure in parenthesis indicates final harvest, t/ha

Apparent conversion

Conversion rate based on supplemental feed applied to each pond was calculated on basis of net weight gained by fish at the end of experiment and the total amount of feed utilized. Conversion rate of feed was calculated from the ratio of total feed supplied and net weight gained by fish as percentage. Projected conversion rate for a 12-month period was also calculated.

3. RESULTS

3.1 Natural fish food production in ponds

Liming of bottom soil with calcium carbonate (CaCO₃) and initial fertilization of water by urea (TSP) and potash created a plankton boom within 15 days in all ponds. As shown in Table 1, during stocking of ponds with fish, transparency levels were 20-23 cm only. On the other hand, when ponds were filled with ground water, the bottom of each pond was almost visible. During the whole period of culture, the colour of pond water remained deep green. During the initial 2-4 months, ponds were occasionally fertilized, especially when the water transparency exceeded 25 cm. After 4 months no fertilization was needed. The color of water indicated a high level of phytoplankton in all ponds during the whole period of investigations.

During partial harvesting and when bottom soil was examined to find out whether any so-called black soil formed at the bottom, surprisingly, no black soil was detected. Black soil formation at the end of the 18-month period was negligible in all ponds except those used for catfish monoculture, where a small amount of black soil was visible after the final harvest. A cumulative high level of supplemental feeds 1142 - 2688 kg/pond (57.1-134.4 tons/ha) as shown in Table 4, and absence of black soil indicated the energy budget in the pond ecosystem was also enriched with detritus-based food. Feed conversion rate or feed efficiency shown in Table 4 does not include natural food produced in the pond and ultimately consumed by fish.

Table 4

Gross and net production of fish biomass and feed efficiency

Figure in parentheses shows amount of feed used, t/ha

3.2 Fish production in ponds

The average gross production of fish biomass in the 4 groups ranged from 288.0 - 632.4 kg/pond (Table 3). Highest average production of 632.4 kg/pond was recorded in the mixed culture group which corresponds to 31.6 tons/ha. Out of the total fish biomass produced in the mixed group silver barb, GIFT and catfish represented 24.4, 40.6 and 35.0 %, respectively. In this group no fish were listed as non-target species (Table 5) as both GIFT and silver barb were stocked in this group, whereas GIFT found in the silver barb group and vice-versa were listed as non-target species. The biomass of non-target species found in ponds was also included in the respective groups as they grew and took nutrition from those ponds. Size variations of GIFT and silver barb were remarkable, some GIFT were more than a kilogram in weight, some were tiny. The largest silver barb was 1369 g. Size variation in catfish was not so much although some individuals were more than 3 kg and some were slightly over 1 kg in body weight. The second highest production of fish biomass was recorded in monoculture of GIFT. An average 442.3 kg/pond (22.1 tons/ha) were harvested from this group within the 18-month period (Table 3). Final harvest for this group was 203.4 kg/pond, which represented GIFT of various sizes, some fish as big as 1125g and some smaller than 5g. The ponds designated for monoculture of catfish or silver-barb were heavily invaded by GIFT (Table 5), but presence of non-target species in GIFT ponds was minimal. Only an average of 10.6 kg of silver barb were present per pond, out of an average biomass production of 442.3 kg/pond during the 18-month period (Table 5). Monoculture of catfish yielded a cumulative biomass production of 374.7 kg/pond (18.74 tons/ha). However, some non-target species, especially GIFT, were found in the group which occupied 23.3% of the total biomass production. Biomass production from catfish alone was 287.4 kg/pond (14.4 tons/ha) as shown in Table 5.

Table 5

Fish biomass production based on target and non-target species (kg)

Figure in parenthesis shows calculated production, t/ha

The lowest biomass of fish among the 4 groups was obtained in monoculture of the silver barb group. Gross production of the fish biomass was 288.0 kg/pond (14.4 t/ha). As already mentioned GIFT somehow entered all ponds of this group and average biomass from GIFT was recorded as 53.7 kg/pond. Therefore, production of silver barb alone was 11.72 tons/ha (2 34.3 kg/pond).

3.3 Feed efficiency

A cumulative total of 2 688 kg of feed was used in each pond of the mixed species group (Table 4). The average production of fish was 632.0 kg/pond (gross weight). The initial stocking biomass of all 3 species in this group was 56.12 kg on average. Therefore, net biomass production stood at 575.88 kg/pond/18 month. The resultant feed efficiency was 21.4%, which means 4.7 kg of supplemental feeding were necessary to raise 1 kg of fish in this group. Feed efficiency was 20.8% for catfish, 24.4% for GIFT and 24.2% for silver barb. The apparent feed efficiency may be considered low when natural food production in ponds is taken in consideration. Here, it may be mentioned that initial stocking weight, especially that of catfish, was high (350 g or more), and the stocked biomass was also high, which reduced the apparent feed efficiency.

Projected production of fish and feed efficiencies 12 months from the beginning of experiment are summarized in Table 6. Projected feed efficiency varied from 31.0 to 50.4% when net fish biomass produced and the feed consumed during that period are considered. This feed efficiency is also apparent as natural foods produced in ponds were also utilized by fish.

3.4 Water quality parameters

Temperature throughout the investigation period ranged between 18.0 and 32.5°C (higher than in the nearby earthen ponds), pH was 7.9-9.0, dissolved oxygen (DO) level varied from 0.7 to 8.0 mg/L. Transparency in pond water varied from 25-45 cm throughout the investigation, after stocking fish.

Table 6

Projected production of fish and feed efficiency after 12 months

3.5 Seed production in pond

Initial stocking numbers of GIFT and silver barb were deliberately kept low (5 000 fish/ha) in mixed and monoculture as these species breed in confined water. Initially little flow of water in ponds stocked with silver barb in mono or mixed culture induced silver barb to produce numerous fry. GIFT reproduced throughout the rearing period. Even an accidental entry of GIFT in the pond with monoculture of silver barb and catfish resulted in a high number of fry of GIFT. The fry of GIFT and silver barb were routinely harvested using fine meshed nets. Total biomass of fry and juveniles and their estimated numbers were recorded whenever those were harvested. This is shown in Table 3. The total number of fry harvested from monoculture of silver barb and GIFT were 27 500 and 58 000 on average, respectively. 40 400 fry of silver barb and GIFT were collected from the mixed group.

3.6 Behavior of fish during culture

Throughout the investigation period of 18 months no fish were fed to satiation except during the initial 2 months when it was difficult to determine the satiation level as fish seldom surfaced to take feed. After 2-3 months of rearing fish in all ponds were observed to come to the surface whenever supplemental feed was administered. Even catfish used to take feed from hand after 12-13 months of regular feeding. Broadcasting of dried feed on the surface of a pond always lured fish, especially fry and juveniles, to the surface, which helped selective harvesting. No feed remained uneaten when occasionally the daily ration of feed was placed on a feed tray made of mosquito net and steel frame, and placed on the pond bottom. Experimentally, feed was mixed with sand and placed on a feed tray, and it was found that fish plough the sand to find feed particles when the tray is placed at the bottom of the pond. When DO went below 1.5 mg/L level, all species were seen to gulp at the surface. Gulping of GIFT and silver barb was frequent. Catfish, unlike the other two species, used to come to the surface and soon dived after a quick gulp.

4. DISCUSSION

The experiment was conducted for a long period to find maximum production of fish under pond-based aquaculture utilizing low cost feed. In this respect, the trial was highly successful. Over 30 tons/ha/18months was recorded using low-cost supplemental feeds with 3 popular culture species together when a cumulative harvesting process was followed. In the investigation standing crop in a given time was not that high. At the end of investigation the highest standing crop in mixed culture was 366.1 kg (18.3 tons/ha) as shown in Table 3 (mixed: 75.8 + 117.5 + 172.8 kg, yield from final harvest). Besides, catfish and GIFT individually registered 234.5 kg/pond (11.73 tons/ha) and 203 kg/pond (10.2 tons/ha) on a monoculture basis when final harvest was made at the end of the experiment. The results indicate 10-12 ton of GIFT or catfish may be produced per year using low cost diets if species are cultured separately. When fed rice bran as supplemental feeding with 5-6% of estimated biomass, GIFT yielded an average gross production of 4 411 kg/ha/6 months. Culture of catfish (P. sutchi) with a stocking density of 12 500/ha fed on 100% pellets (same brand as used in this investigation) gave 11.4 tons/ha/10 months which also included 9.8% of non-target fish mostly tilapia, in an abandoned coastal farm initially used for semi-intensive cultivation of tiger shrimps (AFL, personal communication, 1999). Mixed culture of silver barb, GIFT and catfish may yield 18 tons/ha/year as was found in this investigation based on final harvest as standing crop (Table 3) or more than 25.1 tons/ha (502.4 kg/pond) based on projected production per year (Table 6). As for silver barb, final harvest gave 139.1 kg/pond (7.0 tons/ha) which also included some GIFT. Hussain et al. (1989) registered a production of 1952 kg/ha/5 months of silver barb feeding on rice bran with a stocking density of 1 600/ha, where no fry was produced during the culture time.

Projected production within a year shows much higher rate of yield (Table 6). This is because a selective harvesting process was followed while rearing was in progress, and fish, especially silver barb and GIFT, renewed both their number and loss in biomass. Natarajan (1985) reported on the production of tilapia (O. niloticus) stocked at high density (70 000; 140 000; and 210 000/ha) in tanks manured with poultry manure. He found that total fish production (including fry) was highest at the lowest density. The maximum production was 39.3 kg/ha/day and minimum of 32.1kg/ha/day. Sariz and Arieli (1980) reported highest 25.0 tons/ha/200 days production of tilapia when oil-coated pellets were fed, with a stocking density of 100 000/ha. In case of standing crop the final harvest figure (Table 3) may be the limit under low-cost supplemental feeding in pond-based aquaculture.

The feed used in this investigation was a commercial pellet (Bangladesh Taka 19.00 - 20.00/kg), mustarded oil cake (Tk. 8.00 - 9.00/kg), rice bran (Tk. 4.00 - 5.00/kg) and wheat bran (Tk. 5.00 - 6.00/kg). Therefore, average cost of feed per kg was Tk. 9.00 - 10.00. The apparent conversion rate was 20.8 - 24.2%, which means roughly 4.1 to 4.8 kg of low-cost feed is necessary to raise one kilogram of fish. For a high density culture this cost may be viable if we consider the high price of the marketed fish. But if the culture period is shortened, feed efficiency will increase. This experiment was conducted to maximize the production to find out the potential of high biomass production under semi-controlled conditions. If projected production found in mixed culture within 12 months is considered, both feed efficiency and production rate tend to increase. Production rate of fish initially was very high compared to that of a later stage in terms of supplemental feed supplied. At the later stage of culture most of the feeds were utilized to sustain the standing biomass, not for its increase. Continued selective harvesting could have led to a better fish production during 12 - 18 months.

Most interesting in this experiment was that no reared fish died when the level of dissolved oxygen was low (0.7 mg/L) and that there was almost no accumulation of organic debris at the bottom of ponds in spite of the exceptionally high amount of cumulative supplemental feeding (134 t/ha/18 months) given during the 18 months of mixed culture. As all the test animals used in this experiment had gulping abilities, no mortality was observed. All three species used in this experiment had a habit of stirring up the bottom deposit.

The level of daily supplemental feeding in this experiment was kept deliberately low, which probably encouraged fish to continuously stir the bottom in search of food, thus agitating the bottom soil. This process helped to mineralize the accumulated organic material and encouraged plankton growth. That there was no need for fertilization to enhance plankton growth after 2-3 months of fish stocking was probably due to the fertility being recycled. The continuously high level of plankton in culture, which was indicated by richly green water during the investigation period, was due to the presence of the green alga Chlorella. Chlorella cannot be eaten by any fish employed in this investigation due to its size.

It may be concluded that the high fish biomass production in this investigation was possible due to the stability of the system, including undisturbed water volume, a continuously high level of natural food production, rational supplemental feeding, gulping and stirring abilities of the fish species used in this study.

References

AFL (Aquaculture Farms Ltd.), 1999. Catfish (Pangasius sutchi) production in abandoned shrimp farm. (Personal communication)

Eknath, A.E., M.M. Tayamen, M.S. Palada-de-Vera, J.C. Danting, R.A. Reves, E.E. Dionisio, J.B. Capili, H.L. Bilivar, A.T. Abella, A.V. Circa, H.B. Bentsen, T. Gjedrem and R.S.V. Pullin, 1993. Genetic improvement of farmed tilapia: the growth performances of eight strains of Oreochromis niloticus tested in different farm environments. Aquaculture 111: 171-188.

Gupta, M.V., M. Ahmed, M.P. Bimbao, and C. Lightfoot, 1992. Socio-economic impact and farmers assessment of nile tilapia (Oreochromis niloticus) culture in Bangladesh. International Centre for Living Aquatic Resource Management (ICLARM). Tech. Rep. 35. 50p.

Hossain, M.A,, K.A. Haq, M. Amjad Hossain and N. Sultana, 1994. Mixed culture of giant sea bass, Lates calcarifer with tilapia and other species. Final report. MFTS, BFRI. Mymensingh 2201. 37p.

Hossain M.A,, N. Sultana and M.N. Islam, 1999. Development of improved methods for the culture of some important marine and hyposaline organisms. Final Report. MFIS, BFRI. Mymensingh 2201. 40p.

Hussain, M.G, M. Akhteruzzaman, A. H. M. Kohinour and M.S. Shah, 1989. Use of rice bran in production of Thai Sharpunti, Puntius gonionotus under semi-intensive culture system. Bangladesh J. Fish. 12(2): 35-40.

Natarajan, M., 1985. Recycling of poultry droppings through tilapia (Oreochromis niloticus) fish culture. Effect of stocking density on production. Agricultural Waste 13: 305-313.

Rahman, A.A.K., 1985. Introduction of Exotic Fishes in Bangladesh. Fisheries Research and Training Centre, (booklet). DOF. Chandpur, Bangladesh. 15p.

Sariz, S and Y. Arieli, 1980. Growth capacity of tilapia (Oreochromis niloticus) in intensive culture. Bamidgeh 32(3): 57-65.