Previous Page Table of Contents Next Page





Milkfish is one of the most important aquaculture species and is a successful industry in Asian countries like Philippines, Taiwan and Indonesia. More than 200,000 metric tonnes are produced annually in Philippines and most of it is for food. The culture industry is dependent on the wild fry collected along specific shorelines, mangrove areas and tidal flats of the islands, from the north to the south of the country. Fry collection is an economically important industry, particularly in the rural fishing villages. Over 1 billion wild fry are collected annually and are grown in commercial brackish and freshwater ponds.

Collection techniques for the past 50 years have been refined through research and on-hand experiments by fry collectors in industry to better improve catching efficiency. In general, collection depends on the natural availability and the prevailing weather conditions. The natural environment and weather conditions greatly affect the occurrence of fry and there is differentiation in each place of collection as to the peak season and time of collection. Spawning of female milkfish in the sea is still an unstudied activity, although assumptions have been made by scientists regarding the season and time of the fish maturity and consequent hatching. Occurrence of fry is dependent on many factors such as time and season, which are also related to the moon phase and tidal fluctuations.

Previous studies show that milkfish is present in the Indo-Pacific region, including Fiji. Previously surveyed countries in the South Pacific are Nauru, Kiribati and Tuvalu where milkfish collections are already established. This survey will confirm the availability of milkfish fry around Fiji waters.

Rural communities can benefit from milkfish farming, as the fish provide a cheap alternate source of protein to remedy the rampant malnutrition that occurs due to the lack of protein in rural diets. In addition to this, the farming of milkfish will contribute greatly to resource management by decreasing the pressure on already depleting fish stocks. It will also provide coastal people with a means to generate income, which in turn will decrease poaching activities.


2.1 Wild Fry

Captive milkfish broodstocks have been reported to mature and spawn spontaneously in circular floating net cages. However, mass production of milkfish seed in countries such as Philippines and Indonesia who produce large quantities of milkfish, has not been realised to date due to ineffective methods for collecting eggs (Garcia et al. 1998). Thus, the collection of wild fry is encouraged in order for farming to be successful. Surveys were carried out to assess the fry availability in Fiji.

1 Resource Specialist, Aquafarming Development Foundation, Incorporated, Philippines

2.2 Objectives of Fry Survey

The government of Fiji had planned to start and develop this milkfish culture industry primarily for tuna bait production. Tuna bait requirements based on their Fisheries Business Plan for 1997–1999 was estimated to be about 28 million fish for long-line fishing boats. However, milkfish can also be used as a food supply, and should be looked at as an alternative to dwindling natural fish resources brought about by pressures on natural stock by the increasing fishing activity around the islands.

Schuster (1960) and Rabanal (1970) reported an availability of milkfish fry in Fiji, but no relevant data was available on the extent of this resource. Therefore, it was necessary that this survey be made in order to establish data on fry occurrence in an effort to complement the establishment of milkfish culture industry for tuna bait and food supply. Trial importation of milkfish fry from Kiribati proved to be very expensive with the high cost of freight, and there were problems with unreliable flight schedules, limited freight space, and unstable supply and lack of materials for packing (Appendix 2).

To develop a milkfish culture industry is to establish a source for the fry. The extent of the fry resources will be the basis of the program for nursery and grow-out pond operations. As milkfish fry and fingerlings can be “stunted” for up to a year, collection during the peak season can be maximised. I was given the opportunity by the South Pacific Aquaculture Development Project (Phase II) of the Food and Agriculture Organization of the United Nations to assist the Fisheries Division of the Ministry of Agriculture Fisheries and Forest of the Government of Fiji, to conduct this milkfish fry resources survey. The survey was conducted for six months, ending in June 1997 because of the cold season during which no milkfish fry were caught. Recommendations were made to the Fisheries Division to continue the survey once the weather improved, estimated to be by September or October, and to continue collections until January, in order to complete the one-year record.

The objectives of the fry resources survey are summarised as follows:

  1. to confirm availability of milkfish fry resources in Fiji waters;
  2. to conduct trial collections of milkfish fry using available materials;
  3. to transfer the techniques of site identification, gear construction, catching, collection, sorting, acclimatisation, storage, maintenance, feeding, packing and transport; and
  4. to recommend adaptable fry collecting gear to specific fry collection sites.


3.1 Fry Identification

With the Fisheries staff, a trial collection was made to help staff recognise and become familiar with the fry and collecting methods. It requires good eyesight and practice to differentiate and identify milkfish fry (Figure 1 on Page 3). A good description was made by Villaluz et al. (1983):

Milkfish fry are late postlarvae 10–17 mm in total length (average about 14 mm) which are caught from shore waters when they are about 2–3 weeks old from the time of spawning. Upon capture from the shore waters, milkfish fry have transparent, elongated bodies like those of larval anchovy and sardines and some gobies. In the collector's basin, milkfish fry can be readily picked out by their energetic movements and their conspicuous eyes. They swim together and circle continuously in the same direction. They are able to stay alive where the fry of most other fish species in the same catch have died. Under the microscope, milkfish fry can be seen to have a straight gut without transverse foldings of the intestine. A single line of pigments runs along the lower edge of the abdomen from the throat almost to the anus. The liver is large and sometimes looks like a yolk, which is why the fry have been mistaken to be newly hatched when they are in fact 2–3 weeks old.

Figure 1

Figure 1. Milkfish Fry
(Source: Baguarinao et al.)

3.2 Fry Occurrence

Milkfish spawn in the sea not far from the shore. The resulting eggs or larvae are dispersed in water and drift over the coastal areas. Movements of larvae depend on ocean currents and wind or wave movements. After 2 to 3 weeks, resulting fry, (Figure 1) migrate to the shore and enter coastal wetlands, river mouths, tidal flats and beaches. This migration pattern allows fish farmers to collect them in large quantities. The milkfish fry seasons, however, occur at different times of the year, and the peak season varies in different places. The monsoon or tradewinds greatly affect the peak season as much as pronounced dry and wet seasons. In general, the dry season is usually the peak season in most Asian countries. Seasonal peaks are predictable, but may still vary from year to year.

For the period that the survey team conducted fry collection (Appendix 1), the peak months were December and January. The surveys made in these months showed that using a skimming net with a 1-meter opening, an average of 150 fry were collected per hour, although at the western side of Viti Levu, in the Raviravi area and Sigatoka River, collection could be as high as 500 to 1,000 fry or more per hour. Fry collection was greatly affected by prevailing weather conditions and the type of collecting gear used. In such places as Nasese, Deuba and Tokotoko in Navua, the fry bulldozer (Figure 6, Page 18) or the fry sweeper (Figure 7, Page 19) can be the most efficient, while in the Sigatoka river mouth and at Raviravi, the tidal set net could be the best collecting gear. We found it difficult to construct these types of gear because netting materials were not available in the local market.

In the last month of the survey period, we managed to prepare a fry barrier net, which was set up in the Nasese area. However, the fry occurrence seemed to be off-season by this time and only four pieces were collected. The temperature of 22 to 24°C was too cold owing to the prevailing winds, which according to weather forecasters would last for the next 3 to 4 months, i.e. May to August. We therefore concluded that these months would be part of the off-season, while the dry months of September to February could be the season for fry collection. This, of course, must be verified by continuous collection until the end of the year, in the same sites that we surveyed. Table 1 shows a list of sites and the recommended collecting gear.

As in many Asian countries with developed fry grounds, collection surveys were done early in the morning, and the period of collection is usually during the new and full moons. It was already known that the best time to collect is 3 to 4 days before new and full moons; however, in villages where there is more time to spend, collection may continue 3 to 4 days after these phases, with decreasing numbers of fry collected. Fry occurrence is also affected by the prevailing wind direction: wind direction towards the shore showed an increase, but strong winds and waves may cause higher turbidity, making it difficult to operate the gear and putting more stress on the fry. Collection during rough weather usually results in higher fry mortality.

Table 1. List of sites and recommended collecting gears

PlaceEstimated areaGear type*Estimated catch/person/day (4 hours/day)
Nasese1 km area beachFry barrier net, skimming net, fry bulldozer and push net500 to 1,000 fry
Bau/Gau beach4 km river mouth and beachesFry bulldozer and push net500 to 1,000 fry
Kototonga Village, Sigatoka2 km beach/coveFry barrier net, push net and fry bulldozer500 to 1,000 fry
Sigatoka River4 km river mouth (both sides)Tidal set net, fry barrier net and push net2,000 to 5,000 fry
Raviravi area, Bafronting three floodgatesFry barrier net and push net2,000 to 5,000 fry
Nabila village, Momi bay1 sq. km mangrove areaSkimming net and fry barrier net200 to 500 fry
Tavua bay0.5 km beachSkimming net and pushing net200 to 500 fry
Navua/Deuba2 km beach/river mouthFry barrier net, push net and tidal set net500 to 1,000 fry 
Tokotoko Beach fronting floodgate1 km beach and mangrove areaFry bulldozer, push net and fry barrier net500 to 1,000 fry
Nakalou Village, Dreketi500 m sandy beachPush net and fry bulldozer500 to 1,000 fry
Lekutu Village1 km sandy beachPush net and fry bulldozer500 to 1,000 fry

* Estimated number only if collection is to be made during the peak season (Est. January/February) of the year.

3.3 Site Identification

The Fiji Fisheries staff assigned to this survey team were familiar with the localities, and in other ways, the situations in different parts of the island. This helped the survey team greatly in identifying possible sites. There were four types of sites used as the basis in the selection:

  1. beaches fronting tidal flats or shallow continental shelf with slight wave actions;
  2. river mouth directly connected to the sea with good tidal flats or sandy banks;
  3. mangrove areas with tidal flats or wetlands, as seen in Plate 6 on Page 12; and
  4. areas in front of floodgates directly facing the sea, as seen in Plate 4 on Page 11.

3.4 Types of Collecting Gear

There was no fry net available in the local market suitable for making collecting gear. The net used came from one of the Research Project's holding cages, and was suitable for only the smallest type of collecting gear: the skimming net. Eight units were made so that each team member had one (Plates 1 and 2, Page 10). During the last month of the survey, one roll of fry net about 50 meters long became available. This was made into a fry barrier net, and was installed in the Nasese area (Plates 12 and 13, Pages 15 and 16).

Other collecting gear could have been used if net material had been available locally. However, the design and operation of other types of collecting gear were explained through a videotape shown to the Fisheries staff. The figures on pages 17 to 21 show some of the most efficient fry collecting gear now used in most collecting grounds in Asia. The design and layout may be adjusted according to the prevailing situations, such as available materials and site specifications, as long as the design follows the principle of the system.

3.5 Mode of Operation

The skimming net (Figure 3, Page 16 and Plates 1 and 2, Page 10), was made of 0.5 mm fine mesh nylon net normally used for tilapia breeding hapa. It has a triangular frame with an opening width of about 1 meter or less. The principle is that the collector filters milkfish fry from the water by pushing the net forward. Since milkfish fry congregate mainly at or near the water surface, the driving effect of the skimming net increases the chances of catching more. The skimming net is more effective in places like mangrove areas and tidal flats where other mobile gears are difficult to use. However, other gear with wider openings would enclose bigger areas, and so would result in increased catches (Plates 7 and 8, Page 13).

Each collector has one skimming net (Plate 6, Page 12), a small bucket, and a small white bowl. After pushing the skimming net for 5 to 10 meters, the collectors slowly lifts the net while observing the fry that usually jump on the side of the net. The collectors have to be quick in scooping the fry with some water into the bowl, and then into the bucket. The bucket (Plate 3, Page 11), is usually sitting on top of a floater or a Styrofoam box cover and towed to permit the easy movements of the collectors while they push the nets. Debris and other fish should be avoided as much as possible, and a final sorting has to be done after the collection is ashore.

The operation of the fry barrier net as seen in Figures 2 and 4 (Pages 6 and 17), is a complement to the skimming net. The principle is that the fry barrier concentrates the fry by using favourable wind or tidal currents to move them towards the end of the barrier nearest the shore, then the collector gathers the fry with a skimming net.

Figure 2

Figure 2. Fry Barrier Net

3.6 Sorting

Collected fry are transferred to a white basin and clean water is added. By gently swirling the water, debris and dead fry will concentrate in the center (Plate 10, Page 14). The collector discards these by scooping them with a bowl while the fry swim around the basin. The fry are then counted and sorted, and other species of fish fry are discarded. Generally, the milkfish's prominent black eyes, opaque glass-like body and particular swimming movements distinguish it. The head count is done by visual estimation if the fry have to be transported long distances, e.g. to Nadruloulou Fisheries Research Station. The final count is carried out later when fry have recovered from transport stress.

3.7 Transporting, Acclimatisation and Storage

If there are only a few hundred collected fry, it is not necessary to keep them in the oxygenated plastic bags shown in Plate 11, Page 15. Fry were packed in plastic bags without oxygen, at a maximum of one thousand per bag of 10 litres of seawater. The water was diluted with 2 litres of fresh water, bringing the salinity down to an average 15 to 20 ppt. In the storage area, fry were kept in basins and covered to minimise activity and to prevent entrance of dirt. About 50% of the storage water was changed every day. Fry were fed with either hard-boiled egg yolks or dried wheat flour daily after the water was changed. The storage basins were inspected every morning and afternoon, and excess food, dead fry and debris were removed. The amount of feed was adjusted according to the consumption, as indicated by the amount of food left each day.

If the fry are to be stocked in freshwater ponds, they have to be acclimatised slowly by changing 25% of the storage water with freshwater each day. It may take 4 to 7 days to reduce the salinity to 0 ppt in the basins, but this is an essential step because an abrupt water change may cause salinity shock to the fry. As well, the temperature of both the storage water and the freshwater should be the same. Fry should not be stored for more than 15 days unless aeration is provided and proper water exchange can be made. Stunted growth is expected especially if the container is too crowded; however, releasing the fry into a nursery pond would ensure normal growth.

Fry are usually stocked in the nursery pond when enough have been collected. The number of fry required depends on the type of culture and the size of the grow-out ponds available for operation. The total number of fry collected for the survey period was 4,600 pieces, and about 3,500 of them were stocked at Nadruloulou Fisheries Research Station where a nursery pond was assigned. Unfortunately, the area was flooded after two months causing the fry to be lost, which should have by then been of fingerling size. It was very difficult to retrieve the fish, which were believed to have escaped to the adjacent ponds. The flood took about two weeks to subside. Recovery and inventory of the stocks in the station began soon after the tide receded, but another cyclone flooded the area again even before the ponds had been drained. Recovered milkfish fingerlings could not be kept alive due to their stress during the seining and draining of the ponds, although some were able to find their way to some grow-out ponds nearby.

About 800 pieces were stocked in Montfort Boystown, Wainadoi in a 600 m2 pond prepared for brackishwater pond grow-out culture. These fish were part of the last collection after the first cyclone, and were stored in plastic containers at Nadruloulou Station. The fry were about 1 inch long at the time of stocking, and grew to an average of 60 grams after 60 days of culture, using the benthic algae or “lab-lab” method.


  1. There was limited time to carry out a complete one-year survey.

  2. It should be noted that there were more collections in the Nasese area compared to the others. This was because of the irregular availability of transport to go to the sites targeted, and in some cases due to inability of the Fisheries Division to provide rations or allowances for their staff assigned to the survey team. In general, logistical arrangement hindered some of the scheduled trips. The Nasese area is just walking distance from the area of the city where I stayed, and those collections were made to make the best use of collection opportunities, as the schedule is limited by being based on the phases of the moon.

  3. Fry net materials were unavailable in the local market, and even getting quotations from companies abroad for these materials was difficult. Other collecting gear could have been tested as shown in Figures 5, 7, and 8 to 12 (Pages 18, 19–21), if net materials had been available.

  4. Since there was no clear documentation of the last survey made by Schuster in 1960 and Villaluz in 1970s, no other assessment can be made on the sites, time, season and extent of the fry resources. More time was spent roaming around looking for the best place than was spent collecting fry.


The possible fry collecting grounds recommended for commercial operation in Viti Levu are Nasese area, Sigatoka River mouth, Raviravi, front of Coral Reef Prawn Ltd., Deuba river mouth, and Tokotoko. In Vanua Levu, Nakalou village, Lekutu area and Dreketi area are potential sites for fry and also fingerling collection. These areas were visited at least twice at different periods of the moon, and showed possibilities of commercial quantity of fry especially with optimal weather conditions and the recommended collecting gear adapted to the site. The quantity may vary with the type of collecting gear used and the skills of the fry collector, but practice and persistence would help greatly in increasing the catch.

This fry survey should be done for a full year in order for us to glean a better understanding of the situation, including all the environmental changes, from the dry through to the wet season. To this end, it is necessary that the Fisheries staff involved in the survey work as a team to continue collecting fry even after the completion of this report for a full year's documentation. Although the data involved in this report shows only a part of the real situation for a complete assessment of the fry resources of Fiji, the results of the survey already show a good indication that milkfish fry also occur in Fiji waters and could be in commercial quantities if fry collections could be developed at village level. Since this survey is to establish the occurrence and the extent of milkfish fry resources, the mechanism and strategies of organising the village people to do the fry collection themselves lies with the government agencies, specifically the Fisheries Department, which is the main component of this survey program. The law, giving the villagers their customary fishing rights in their respective areas, will play a role in organising a group of collectors for the specific sites mentioned here. The Fisheries staff involved in the survey will have to demonstrate the fry collection techniques to villagers where commercial collections are recommended.

Other Asian countries have developed milkfish fry collection into a big business, and perhaps their systems could be adopted in some ways to encourage more people in Fiji to get involved. A consignment system could be tried, which works by having a middleman who would be the exclusive buyer of the fry collected from a specific area, and who would pay for the fry after obtaining a license from the government or even the village chief. The Fisheries Division should devise an organisation whereby villagers and middlemen work together in a business partnership, while the government concentrates on developing the nursery or grow-out ponds for whatever fry are collected from the source.

Milkfish can be marketed as tuna bait or as food, and the government should be definite about where to concentrate production. Tuna bait consumption is enormous in terms of volume, but this bait must be cheap enough to compete with imported bait, and it must be available in large enough quantities to convince the tuna industry to install live bait wells. There really is a need for a demonstration pond to carry out a trial culture and use the produced bait for a tuna boat. As a food fish, milkfish will need to be promoted to the public. Because it is well known for its bony meat, currently other fish are preferred by buyers who do not know how to prepare this fish. Demonstrations of de-boning and soft-boned cooking could be made, along with other promotions, so the public could learn to accept milkfish as an alternative to other fish.


Annual Report. Fiji Fisheries Division 1995. Ministry of Agriculture Fisheries and Forests.

Baguarinao, T.U., Solis, N.B., Villaver, W.R., and Villaluz A.C. 1986. Important Fish and Shrimp fry in Philippine coastal waters: Identification, Collection and Handling. SEAFDEC Technical Report No. 10.pp.1–50.

Lee, C.S., Gordon, M.S., and Watanabe, W.O. 1986. Aquaculture of milkfish (Chanos chanos): State of the Art. Oceanic Institute of Hawaii, Makapuu Point, Waimanalo, Hawaii 96795, USA. pp. 153–178.

Lopez, N.A. 1994. The privatization process of the national bangus (milkfish) breeding program in the Philippines. In Y.C. Shang, P.S. Leung, C.S. Lee, M.S., and I.C. Liao (Eds.) Socioeconomics of Aquaculture. Tungkang Marine Laboratory Conference Proceedings 4: 103–113.

Villaluz, A.C., Villaver, W.R., and Salde, R.J. 1983. Milkfish fry and fingerling industry of the Philippines: methods and practices. SEAFDEC Technical Report No. 9. p. 171.

Photographs: Set A

Plate 1

Plate 1. Preparation of skimming net.

Plate 2

Plate 2. Other forms of skimming nets.

Plate 3

Plate 3. Preparing container to hold collected fry.

Plate 4

Plate 4. Floodgates and river mouths were likely to be the best collection sites.

Plate 5

Plate 5. Testing for water parameters such as salinity, temperature and turbidity.

Plate 6

Plate 6. The skimming net is the simplest and most practical gear in mangrove areas and shallow tidal zones.

Plate 7

Plate 7. Bucket and bowl in hand, the collector drives the skimming net to filter the water where fry are carried by water current, and usually float.

Plate 8

Plate 8. Milkfish can be found in groups not far from beaches.

Plate 9

Plate 9. Fry were found in tidal flats and mangrove areas.

Plate 10

Plate 10. Collected fry are sorted in a basin, and debris and other foreign matter was removed.

Plate 11

Plate 11. Fry is transported in plastic bags, and oxygen should be supplied for long journeys.

Plate 12

Plate 12. Fry barrier net installed along Nasese Beach.

Plate 13

Plate 13. Fry barrier net with bamboo posts to hold it upright. This is the setting during lowest low tide.


Figure 3

Figure 3. Skimming net operated in shallow mangrove areas.
Source: Baguarinao et al.)

Figure 4

Figure 4. Barrier net on shore perpendicular to currents.
(Source: Baguarinao et al.)

Figure 5

Figure 5. Scissors net or bigger skimming net.
(Source: Barguarinao et al.)

Figure 6

Figure 6. Fry bulldozer; fry filter net with raft and small hut.
(Source: Barguarinao et al.)

Figure 7

Figure 7. Push net or sweeper pushed along the shore, or fixed in place.
(Source: Barguarinao et al.)

Figure 8

Figure 8. Modifications of the push net: A) enlarged frame with one person dragging and another pushing it, and B) with long wings made of coconut leaves.
(Source: Barguarinao et al.)

Figure 9

Figure 9. Towed bag net with conical end secured with a string
(Source: Barguarinao et al.)

Figure 10

Figure 10. Tidal set-net.
(Source: Barguarinao et al.)

Figure 11

Figure 11. Tidal set-net set in river mouth.
(Source: Barguarinao et al.)

Figure 12

Figure 12. Operation of the double-stick net.
(Source: Barguarinao et al.)

Previous Page Top of Page Next Page