World Aquaculture March 2019

40 MARCH 2019 • WORLD AQUACULTURE • WWW.WA S.ORG 1986). However, if that limitation could be resolved, additional opportunities also developed, especially if the price of the milkfish fry could be reduced (Lee and Gordon 1986, Shang 1986). In 1984, the Oceanic Institute was the recipient of a multiyear grant (US$ 1.2 million/yr) from the United States Agency for International Development. Milkfish that had been stocked in ponds in Kona and grew on natural pond productivity were collected, transported and placed into various research facilities located at the Oceanic Institute campus on Oahu (Kelley and Lee 1986). From that point on, milkfish were provided with a continuous source of seawater, aeration and fed a commercial diet of 40 percent crude protein and 14 percent crude fat. The beginnings of the first project was summarized in the proceedings of an international workshop held in Taiwan (Lee and Liao 1985). That project dealt with developing the technology for artificial propagation of milkfish. At that time natural spawning and commercial production of milkfish fry was already occurring in Taiwan (Lin 1985). The reproductive characteristics of female milkfish were determined and the technology for intensive larval rearing of both striped mullet and milkfish was developed. Additional milkfish broodstock were obtained fromMolii fishpond on Oahu and these were stocked into earthen ponds on the Oceanic Institute campus and allowed to spawn naturally (Kelley and Lee 1986). A summary of the major findings of the first USAID project included defining the state of maturity most receptive for hormonally inducing a female milkfish to spawn (Lee et al . 1986). Oocytes were obtained from anaesthetized milkfish broodstock using a polyethylene cannula and then fixed and diameters measured. There was a unimodal size frequency distribution of oocytes and an average oocyte diameter > 750 µmwas necessary for successful hormonal induction of spawning. The team then utilized a more aggressive sampling procedure in which oocytes were measured at approximately 7-d intervals and a statistical model of ovarian maturation rate was determined (Fig. 3). Collaboration with the University of Tokyo resulted in characterization of the serum steroid profiles that accompany oocyte maturation (Fig. 4). The serum level of E 2 reaches its lowest level when oocyte diameter averaged 850 µm, a physiological indication that maturation was complete (Tamaru et al. 1988). Natural spawns were detected in outdoor ponds with an egg collector installed in the drainage system and checked daily. When eggs were present, the percent fertilization, fecundity and a small sample were preserved to determine egg diameters, all part of the process of monitoring natural spawnings. Fatty acid and amino acid profiles of eggs from natural and hormonally induced spawns revealed that naturally spawned eggs had higher concentrations of the majority of the biochemical constituents investigated (Ako et al . 1992). In particular, arachidonic acid (C20:4n6), the precursor of a host of pheromones that are released when a female ovulates, was elevated in naturally spawned eggs. Those events are important in attracting males to participate in spawning and a different indicator of when the ovary is fully mature. The most practical means of producing fertilized milkfish eggs of the highest quality was to allow the broodstock to spawn naturally (Fig. 5). It is not surprising that this became the means of commercially producing fertilized milkfish eggs used for hatchery production of milkfish fry. Efforts to control of maturation and spawning of milkfish used a combination of ancient and contemporary knowledge, science and technology that led to understanding how and why maturation of female milkfish occurs. Japan has always been considered the leader in the development of hatchery technologies of marine organisms and the program recruited three Japanese scientists: Dr. Atsushi Hagiwara, Dr. Yoshioki Oozeki and a larval rearing expert, Hiroki Eda. Dr. Hagiwara brought the L-type ( Brachionus plicatilis ) and SS-type ( B. rotundiformis ) rotifers that formed the basis for food selectivity experiments. Hiroki Eda was the source of a phytoplankton ( Nannochloropsis oculata ) that was key to resolving a number of challenges in larval rearing of marine species, including the striped mullet. Dr. Oozeki brought thirty 20-L polycarbonate tanks used for small-scale larval rearing experiments. The experimental design essentially broke up the rearing protocol into discrete time units that could be investigated separately, systematically and quantitatively. The results of these small-scale trials were then combined in larger-scale larval rearing trials that were done in 5,000-L tanks. Each tank was a round fiberglass tank with the inside painted black and equipped for continuous flow of seawater, temperature control, aeration, and drainage through a center drain and a so-called Banjo filter. Trials ended with a stress test in which larvae were suspended in air in a dip net for thirty seconds and survival of the larvae after recovery was determined. These additions were a quantum leap forward compared to LEFT, FIGURE 5. Milkfish broodstock hormonally induced to spawn, Oceanic Institute, circa 1986. RIGHT, FIGURE 6. Backyard milkfish hatchery, Gondol, Bali, Indonesia, 1993.