2. FIELD TRIALS AND RESULTS

2.1 NURSERY MANAGEMENT, REARING OF MILKFISH FRY, SHRIMP CULTURE RESEARCH CENTRE, 1974–75

Four trials in the rearing of milkfish fry to early fingerling size of from 0.2 to 1.0 were made in the D-series ponds from June 1974 to September 1975. These ponds are small (400 to 500 m²), shallow, with clay soil. The first crop was raised in pond 1 from June to July 1974. Ponds 1, 2 and 6 were used for the second test from December 1974 to January 1975. The third series was conducted in ponds 1 and 6 in April and May 1975, while the last was done in ponds 2, 3 and 6 in September 1975. The first and last were dry season runs, made at the start of and later period of the two seasons of 1974 and 1975, respectively. The second and third covered the same portions of the rainy period of the year 1974–75.

Preparation of the ponds for the first, second and third series was invariably the same. Leaks were eliminated, dikes repaired and plastered with soft soil to fill up holes and minimize seepage. Fine nylon screens were installed in the sluices. To kill all fish inside the ponds, derris root was crushed, pounded and the extract sprayed on the water.

In the fourth series, preliminary work, aside from the above, included draining of the ponds by means of pumps, drying of the pond bed for up to two days and application of cow dung while the pond bottom was still exposed.

Manure was applied to the water in two stages, the first, a heavy fertilization with rice bran, urea or ammonium sulfate, and triple-superphosphate before stocking. This was to stimulate and produce a good crop of natural food, either plankton, periphyton, or lab-lab ¹ or a combination of the three. After the fry were released, one to three dressings with chemical fertilizers alone or with rice bran were applied. Computation for the quantities of fertilizers were based on pond area for the initial application, and water volume for the dressings. Tables 4 and 5 compare the amounts used in the trials.

Stocking procedures for the two seasons differed. The high salinity natural to the ponds in the dry period and the practice of fry dealers to keep the fry in containers under almost freshwater condition called for measures providing gradual acclimatization lasting up to 48 hours. This was accomplished with the use of an accliwation tank, temporarily constructed in a very accessible area inside the nursery pond. The fry were released here, in water previously toned down to about 20 ppt salinity by addition of fresh water. Salinity was raised gradually by diffusion through a small cut on the dike, made about 12 to 24 hours after the fry were released. The stock was not allowed to swim out until the fry were seen crowding against the fine screen set inside the cut.

During the rainy season, acclimatization was done only in the big tubs, taking at most an hour before the fry were planted directly in the pond.

¹ Lab-lab is the term now commonly used for the microbenthic community of plants and animals forming a mat on the bottom of brackishwater milkfish ponds. It is considered, so far, the choice food of milkfish under cultivation.

No artificial feeds were given. However, in the last series, pond 6 had to be provided with lab-lab from other sources when its own supply finished much earlier due to the higher stocking rate and subsequent infestation by chironomid larva.

The narrow tidal range and the difficulty of draining the ponds limited water exchange. Heavy rains reduced salinity to almost fresh water condition. On the other hand, salinity in the lateral and supply canals during the September 1975 series went beyond 40 ppt when the main water source was closed in the course of major renovation and improvement work.

Results of the trials were poor to fair (Tables 6 and 7). Highest recoveries, with fingerlings of almost uniform sizes were attained during the dry period in spite of water salinity reaching over 40 ppt. The food generated during the June-July 1974 trial was a combination of thin lab-lab, plankton and periphyton; in the September 1975 series, mostly lab-lab. On the other hand, with the exception of pond 6 where lab-lab managed to grow during the second experiments, low survival, poor growth and a wide disparity in sizes were noted in the wet season, more particularly in ponds where only plankton blooms resulted from fertilizers application. During this period, heavy rains reduced salinity drastically and caused pond waters to become turbid. Palaemonid shrimps were abundant also.

The September 1975 series, although giving satisfactory results in terms of survival and uniformity of size brought to the foreground the problem of chironomid infestation that could seriously affect early fingerling production and limit stocking densities. In this series, ponds 2 and 3 started with a fair lab-lab cover. A week after they were stocked, chironomid larvae became established, competing with the fry for food. Pond 2, besides suffering the higher mortality, which may be attributed to imporper handling which weakened the fry at stocking time, gave the poorer rate of growth, performing only slightly better than pond 6 despite its lower stocking rate. The incidence of chironomid larvae infestation, although only between 1.05 and 1.43 per 100 cm² was serious enough to warrant early transfer of the fish, while still quite small to other areas.

Tables 6 and 7 give the results of all the four tests:

Tables 6 and 7

2.2 REARING OF MILKFISH FINGERLINGS

There was need for multi-size groups of milkfish fingerlings for the stock manipulation technique as envisaged in the programme of work for raising marketable size fish in the production ponds. For the plan to be properly executed, a minimum of three size groups was required, the first composed of fingerlings weighing over 50 g, the second between 10 to 50 g and the third, below 10 g. To produce these different sizes, use was made of the B-series ponds and two ponds of the F-series. The B ponds are between 1 600 and 2 000 m². Ponds F-9 and 10 are over one hectare.

Although in a few cases early fingerlings from the nursery ponds were directly transferred to production ponds without passing through the intermediate compartments, those that were produced were generally allowed to attain medium size of between 3 to 10 g in the B ponds and to post fingerlings of more than 50 g in the F ponds.

The condition of the B ponds was far from ideal, precluding employment of efficient techniques of pond management. Like the D-series ponds, total drainage by natural means was almost impossible. To prepare them for stocking, the ponds had to be poisoned with derris root extract, then fertilized with organic fertilizer, usually rice bran, cow dung or chicken droppings, combined with urea or ammonium sulfate and triple-superphosphate in quantities sufficient to grow a good plankton bloom. Fertilizer dressings, also a mixture of both organic and inorganic fertilizers were applied after stocking to maintain the natural food for the fish population. Fish withdrawal was by seining the whole pond.

In contrast, the F ponds had slightly higher elevation. During low tide, provided no other ponds from upstream are being dewatered, these ponds could almost be totally drained except for the peripheral canals. These ponds were originally intended for shrimp culture but with the need for bigger areas to hold and produce larger fingerlings, they were temporarily converted into fingerling ponds, with a catching pond constructed in front of the sluices to affect easier impoundment of the stock.

Withdrawal of fish was by herding the school into the catching compartment by taking advantage of their instinct to go against the current, and then seining them out from this small area.

Preparation of the F ponds included draining, poisoning with derris root extract, and when possible, fertilization with cow dung or rice bran and urea and triplosuperphosphate. Snails (Conithidae) were killed with Brestan 60 (60 percent tri-phenyltin-acetate) at a concentration of 0.3 ppm, a.i. (active ingredient). Lab-lab grew easily in these two areas.

The B and F ponds used for this purpose also became holding areas where stock for the production ponds were drawn from time to time. No accurate estimates on production and survival rates could be made, however. The lack of other ponds which could be more easily manipulated and where fish could be readily and safely taken out, curtailed regular transfers of the stock from one pond to another both for inventory purposes and to sort out the carnivorous fishes. Consequently the ponds remained with fish for an indefinite period.

Subsequent stocking eventually produced as many as 8 groups of fingerlings, each batch from an individual pond being classified as a size group. A 9th group, averaging 160 g came from other production ponds and were used for possible brood stock rearing for milkfish maturation experiments.

2.3 PRODUCTION OF MARKETABLE MILKFISH

2.3.1 Testing of Improved Methods

Improved methods of milkfish culture were tried in a series of production experiments conducted in the E ponds of the Centre from July 1974 to March 1976. Five crops were raised in ponds 2 and 3; four in ponds 1 and 4.

The ponds are from 1.19 to 1.39 ha, shallow, yet undrainable unless pumps are used. They are located about 1.5 km from the main sea water source, getting the supply through a lateral canal which, like the main canal was also silted and narrow in some sections.

The soil is predominantly clay. Analysis of samples from two ponds as shown in Table 1 indicated conditions favorable for milkfish production. However, water retention was poor due to seepage and leaks along the dikes and ill-fitting flush boards in the concrete sluices. Maximum water depth that could be allowed during spring tides reached barely over 40 cm. This would fall to about 20 cm during the neap tide period when water cannot be taken in. As a remedial measure, trenches were dug along the main dikes where leaks were located, then filled and packed with new soil. Catching ponds were also constructed in front of the main concrete sluices and equipped with standardized wooden gates which could be easily scaled with soil (text plates 5, 6, 7 and 8).

The techniques used are briefly described as follows:

2.3.1.1 Pest Control

Before Stocking

The ponds were poisoned with rotenone extracted from soaked and crushed derris root. Good quality derris root was applied at a rate of 4 kg/1 000 m³ of water. Thiodan (35 percent endosulfan) was occasionally applied at a concentration of 0.1 ppm of the product.

Snails (Conithidae) were killed with Brestan 60 (60 percent tri-phenyl-tin-acetate) at a concentration of 0.3 ppm, a.i. Ponds treated were completely drained by pumping after the snails were dead.

After Stocking

For checking chironomid larva, gamma BHC granules, diazinon, either the granulated (Basudin 10) or emulsified (Basudin 60) form, and liquid Sumithion were used at concentration of 0.08, 0.08 and 0.1 ppm, a.i. respectively. These were applied whenever heavy infestation, particularly in ponds with lab-lab, occured.

2.3.1.2 Fertilization

Before Stocking

In ponds that could not be drained, an initial heavy dose of both organic and inorganic fertilizers was applied to the waters immediately after poisoning. Cow dung and/or rice bran up to a maximum of 2 000 and 1 500 kg/ha, respectively, plus urea or ammonium sulfate and triple-superphosphate in quantities to supply at least 20 kg N/ha and up to 30 kg P₂O₅/ha were used.

When total drainage and substantial drying of the pond was accomplished as in the 4th series organic fertilizers were broadcast evenly on the dry bed, shortly after which water was allowed and chemical fertilizers applied.

After Stocking

Dressings with chemical fertilizers alone or in combination with organic manures were made at almost weekly intervals up to about 2 weeks before total harvest. Applied on the basis of water volume, urea at 3.33 ppm (1.5 ppm N) and triple superphosphate at up to 4.17 ppm (1.87 ppm P₂O₅) were used. On the basis of pond area, resorted to when water depths exceed 30 cm, application was usually 11 kg urea (5.0 kg N) and up to 16.6 kg triplesuperphosphate (7.5 kg P₂O₅)/ha. Ammonium sulfate was sometimes substituted for urea. Rice bran or cow dung never exceeded, regardless of depth, 300 kg/ha, air dry basis. Fertilizers were applied during fair weather, preferably in the morning.

2.3.1.3 Stock Manipulation

Stocking with up to 9 size groups of fingerlings was tried. Releases were made within 2–3 days time or distributed over a period of up to 7 weeks. Harvesting was started as soon as the fish reached a marketable size of 135 g and upwards using gillnets dragged throughout the pond, or effectively herding them into the catching ponds by allowing then to go against the current. The small individuals coming in with the schools were made to swim back through a sorting screen fixed in the sluice by reversing the water current.

2.3.1.4 Supplemental Feeding

To a very limited, extent artificial feed was given to supplement natural feed. Rice bran was broadcast on the water or soaked and placed on feeding platforms set in the deeper areas about 15 cm below the water surface. Daily food fed was 3 percent of the estimated body weight of the stock. A two-day ration was supplied every other day.

2.3.1.5 Water Management

New tide water was taken in every spring tide during the dry season. Freshening was performed whenever water conditions became bad, when salinity and H₂S levels were high, oxygen content low, tide conditions permitting. During neap tide periods, when this is not possible, and more so when the fish come to the surface during the early morning hours, hydrated lime was applied at a concentration of 100 to 200 ppm.

2.3.2 Trials on Production of Marketable Milkfish

2.3.2.1 July to November 1974 Crop Season (Dry Period and Early Part of the Rainy Season

After the ponds were reclaimed with derris root extract and given the initial fertilizer dosage, fingerlings ranging from 0.80 to 66.6 g average weight were released, with stocking operations finished in a few days. In ponds 1 and 3, a thin layer of lab-lab developed after the manures were applied. In pond 4, lab-lab was slightly thicker in approximately 1/6th of the area. This pond had been previously partially exposed. It was temporarily enclosed with a small dike and sufficiently dried when the water was reduced and maintained to its lowest level with a small pump. In pond 2, filamentous green algae (Chaetomorpha spp.) already present before stocking, began to fill the pond and areas had to be cleared manually to provide swimming space for the fish. Weekly dressings with fertilizers stimulated both lab-lab and filamentous algae.

Water in all ponds was most often clear. Whitish waters which developed at one time in ponds 1 and 2 after a sudden heavy rain were treated with hydrated lime. Salinity fluctuated between 31 and 43 ppt.

Chironomid larva infested ponds 1, 3 and 4. Sumithion, diazinon and gamma BRC, the last two in granulated form effectively controlled them but had to be applied at 12-to 15-day intervals when reinfestations occurred. Shrimps were killed by the pesticides. Chironomid larva were hardly noticed in pond 2.

Partial harvests started first in pond 4 where the biggest size group became marketable in 44 days. Harvests in ponds 1 and 3 commenced 75 days after stocking; in pond 2 after 125 days. All ponds were fished out after a culture period of 102 to 136 days. The highest yield in terms of net milkfish production was 446.7 kg/ha, attained in pond 3; the lowest: 148.4 kg/ha in pond 2, although this pond yielded the highest crop of miscellaneous species.

The main reason for the poor yield in pond 2 was the small size of fingerlings used, and the food present. Another reason was the low recovery. The poor growth rate of the fingerlings could have been due to their inability to utilize young Chaetomorpha spp. The high mortality could have been caused by carnivorous fishes, birds, snakes, etc. The limited space as a result of algal coverage and the shallow waters where the fingerlings were confined for some time could have made them very vulnerable to predation.

Results of the test are shown in Table 8. Inputs for this series and all the succeeding trials are shown in the economic assessment of the operations, Tables 14–17.

2.3.2.2 December 1974 to March 1975 Crop (wet season)

The programme planned for this crop season was patterned after the deep water method, a technique that showed promising results in the Philippines. (This particular method calls for developing lab-lab, gradually raising water depth and shifting natural food production to plankton, yet maintaining a thin film of benthos under conditions of deep water and low salinity. The theory is to provide the fish with detrital plankton raining down on the pond floor from the water column above. The benthos serves as a food reserve when plankton production declines in between fertilizer applications.)

The dikes were repaired, low portions raised and berms constructed to reinforce and widen the base of the portions along the supply canal and the adjacent privately operated areas, hopefully, to be able to collect rain and maintain deeper waters inside the ponds. The ponds then were drained to their lowest levels, exposing portions of the bottom for a short period. Sea water was allowed, Thiodan applied and the following day, the manures were spread.

Six days later the first batch of 1 500 fingerlings was released. Two weeks after and spread over a period of five weeks, some additional 1 500 fingerlings were added. A total of nine size group composed the fish population.

Fertilization consisted of a heavy basal dose of wet cow dung, filter press cake (a by-product in the manufacture of sugar from sugar cane), rice bran, urea and triplesuperphosphate. Weekly dressings were combinations of cow dung, urea and triple-superphosphate. In pond 3, ammonium sulfate was used in place of urea.

The sluices were blocked with soil after the ponds had been filled to the maximum with sea water. Thereafter, all rain water was impounded. Depths reached up to 43 cm. Salinity went down to 3 ppt but lab-lab was still present on the shallow portions of pond 3 while examination of surface mud samples from ponds 1 and 2 showed blue green algae and diatoms.

In 60 days, the bigger fingerlings grew to marketable size and were harvested. Thirtysix days later, the ponds were totally fished out. Results of the trials are shown in Table 9.

2.3.2.3 April to July 1975 Crop Season (Last Part of Wet Season and Beginning of Dry Season

The possibilities of rice bran applied as the main and only organic fertilizer under varied pond preparation and management operations was explored in this series. A budget of 1 000 kg/ha was earmarked, split in two to seven applications. Chemical fertilizers were also applied (See schedule, Table 10).

Pond 3 was totally drained by pumping. Waters in ponds 2 and 4 were reduced to their shallowest depths during two successive lowest tides, exposing almost ⅔ of their beds for short periods. Pond 1 was kept constantly under enough water to cover the whole area.

An 80 percent of the total rice bran input was evenly broadcast in pond 3 while the bottom was thus exposed. A few days later, tide water was allowed simultaneously in all ponds, Thiodan applied and the first heavy dose of a combination of rice bran, urea and triplesuperphosphate introduced. Eight days later, the ponds were stocked.

Weekly fertilizer dressings were applied during the next six weeks. It was programmed that by the end of the period all ponds will have received the same inputs at the same rates. During the first three dressings, urea and triple-superphosphate ratio was almost 1:1; in the last three, triple-superphosphate was increased by 25 percent, urea decreased by 25 percent. This was intended to induce vegetative growth of the natural plant food first, and later to hasten maturity and create, especially among the filamentous algal filaments an association of assorted plankters and epiphytes. In ponds where the filamentous green algae Enteromorpha spp. proliferated, the fertilizers were broadcast over the algal masses.

Unpredictable weather created many problems during this period. Rains fell immediately after the first application of rice bran in pond 3, causing filamentous algae (Enteromorpha spp.) to become established. Sporadic rainfall also caused lab-lab (which was dominant in pond 2) to deteriorate, starting from the sides and corners where they had floated and accumulated. This produced milky waters and hot spots of H₂S, causing the fish to surface during the early mornings. Hydrated lime was applied and later, new tide water was allowed.

Prolonged periods of calm and warm days also brought about a heavy infestation of chironomid larva in pond 4. This was checked with gamma BHC but the damage had gone too far and the lab-lab which was predominant at the start never recovered. Instead Enteromorpha spp. became evident. Gamma BHC also killed practically all the shrimps.

Fifty days after stocking, the fish had grown to marketable size and partial cropping commenced. Two weeks later, all the ponds were fished out. Harvesting operations centred in the catching ponds where the fish were confined by water manipulation. Table 11 gives the results of the trials.

1 Estimate, based on percent survivals in ponds 2, 3 and 4 and average weight of thefish taken out of the pond.

Loss of the fish occurred in pond 1 during the first partial withdrawal when an undercut developed below the bamboo gratings in front of the main sluice allowing the fish to escape to the outside canal. Fish harvested from this pond were also small. It was apparent that they were not getting ample nutrition. There was very little lab-lab in this pond at the start. Subsequent fertilizer dressings created plankton blooms.

The quantity of shrimps (Metapeneaus monoceros) taken out of pond 2 are also of interest. Of four and five g size, they were caught beginning one month after the ponds were poisoned, and barely three weeks after the milkfish fingerlings were stocked. Those taken out from ponds 1 and 3 were about the same size. Swarms of Brachionus spp. were observed in pond 2.

2.3.2.4 July to November 1975 Crop Season (Dry Season and Early Part of the Wet Season

During this period, a 20 cm low lift pump had become available, making possible testing of improved methods of milkfish culture hinged on the time tested procedure of growing lab-lab for the main food of the fish.

After the pond had been poisoned with Brestan 60 to eliminate all snails (Conithidae), they were drained simultaneously, then dried sufficiently until the ground cracked and became hard, short of dessication of the surface layer (text plates 10 and 11). Cow dung and rice bran were broadcast on the dry bottom. Immediately after, water was allowed to a depth of about 5 cm, urea and triple superphosphate applied to start lab-lab growth.

As it was impossible to dry the ponds again by natural evaporation (pumping had to be done continuously to evacuate all seepage water coming in and collecting in the peripheral canals during the process of drying), additional water was taken in gradually until depth reached about 20 cm. By this time, lab-lab had almost covered the entire pond bottom. Stocking commenced, using the catching ponds as acclimatization tanks, and was completed in two weeks. The time interval between draining and stocking of the first batch of fingerlings was 18 days.

Weekly dressings with chemical fertilizers were made for eight consecutive weeks. Chironomid larva infestations were checked with gamma BHC granules, later with Basudin 60 E.C. applied at almost 10-day intervals. Fine rice bran was used to supplement natural food but feeding had to be discontinued when the fish did not seem to take to it readily. Fertilizer and feed inputs were limited to the quantities used in the first production trials during the same period the previous year.

Harvesting started after 7 weeks. During the next seven weeks, all ponds were cleaned out. Harvesting operations were confined mostly to the catching pond.

The fish did not grow satisfactorily as expected. The biggest size group averaged only 287.6 g, an increase of only 127.6 g in 46 days. Deposition of fat on the abdominal wall, the best ocular indication of the condition of the fish, was very thin. Three possible causes were suspected:

1. Tri-phenyl-tin-acetate: Tri-phenyl-tin compounds (acetate, chloride and hydroxide) have been used in Philippine ponds for killing snails belonging to the family Conithidae. The effective concentration found to eliminate these pests was 0.3 ppm, a.i. In ponds where the maximum of 600 g Brestan 60 (360 g tri-phenyl-tin-acetate) was applied, the fish which were stocked after water depth was raised above 20 cm were observed to grow normally.

   In this series, tri-phenyl-tin-acetate applied was from 40 to 103 percent more, chiefly because the ponds were not level and needed deeper waters in order to keep the bottom completely submerged. The possibility that tri-phenyl-tin compounds in excessive dosage can profoundly curtail fish growth cannot be discounted.

2. Diazinon in emulsified concentrate: the diazinon that was applied during the previous year was in granulated form. This sank to the bottom immediately. In contrast, the emulsified type mixes first with the top layers of the water before diffusing downwards. It had been noticed that the fish were schooling and swimming lazily along the side of the pond farthest from the wind direction at noontime the day following application of diazinon. This seemed rather unusual for milkfish.

3. Prolonged high salinity: The trials were conducted when renovation work in the Centre's ponds was at its peak. Closure of the main canal left only one small auxiliary canal for the experiment ponds and the rest of the private areas in the vicinity. Salinity in the ponds reached 56 ppt and remained almost always above 40 ppt during the first 60 days of rearing.

These conditions notwithstanding, the average production of milkfish was 99.5 kg more than for the same period the previous year. Considering the manner with which the fish were harvested, the difference will have been more. Later observations showed that fish that had been confined in the catching pond for three hours lost 9.5 percent of their body weight, a decrease due to the emptying of the gut and energy spent in swimming against the current.

2.3.2.5 November 1975 to March 1976 Crop Season (Wet Season)

This was a replication of the field trials conducted about the same period the previous year (2.3.2.2). Ponds 2, 3 and 4 were totally drained with a 20 cm pump, then filled to maximum depth with tide water and the sluices sealed. Manures were immediately applied. Stocking was started one week after fertilization and was completed within three weeks. Fertilizer inputs were almost the same as those used in the first similar trial.

Maximum depth attained with added rains reached 46 cm. Salinity went down progressively from 12 ppt to about 4.5 ppt. Contrary to expectations, filamentous green algae (Chaetomorpha spp.) became established and dominant in all ponds. Ruppia spp. a higher aquatic weed, also started growing, finally covering almost all of pond 2 when the filamentous algae upon which the fish were feeding were almost completely gone.

Seven weeks after the initial stocking, the fish were thinned out. When the ponds were drained after the trials ended, only small patches of filamentous algae remained while Ruppia spp. in pond 2 still remained dense. Fish of the smallest size group did not reach marketable size but were harvested for tuna bait purposes.

Fish production during this season is shown in the following table:

1 Total production minus stocking weight.

2 Wet Prices as of close of 1975

1 Wet Prices as of close of 1975

1 Wet Prices as of close of 1975

1 Wet Prices as of close of 1975