Biofilm composed of mucus plays an important role in the settlement processes of invertebrate larvae such as Haliotis discus hannai (Pawlik, 1992; Keough and Raimondi, 1996; Roberts and Watts, 2010; de Vicose et al., 2010). In the ar-tificial seed production of H. discus hannai, it is crucial for microalgae of sufficient quantity and quality to attach to a sub-strate such as acrylic plate.

Larval growth differs depending on the microalgal charac-teristics, and food resources for larvae may become insuffi-cient when large amounts of microalgae fall off the acrylic plate due to their weight. When the microalgal membrane is flat, larval settlement onto the biofilm tends to be successful. However, if the membrane is three-dimensional, larvae show low settlement rates (Kawamura and Kikuchi, 1992; Searcy-Bernal, 1996; Roberts et al., 2007a). Furthermore, some dia-toms with hard silica cell walls tend to have low digestion rates by the larvae. Therefore, it is necessary to obtain high-quality microalgae attached to acrylic plates to achieve the successful artificial seed production of H. discus hannai.

Studies on the cultivation of early and immediate post-set-tlement larval stages are extremely limited compared to those of later larval stages. Moreover, studies on the effectiveness of feed for the larvae of H. discus hannai have been limited to diatoms (Han and Hur, 2000; Parker et al., 2007; Roberts et al., 2007b). The aim of this study was to examine the growth and attachment rates of microalgae including the green algae Tetraselmis and blue-green algae. The settlement, metamor-phosis, and survival rates of H. discus hannai larvae fed the aforementioned microalgae were also studied.

The growth and attachment rates of 31 kinds of microalgae selected from the Korea Marine Microalgae Culture Center (KMMCC), including 24 diatoms, three green algae Tetrasel-mis, and four blue green algae were examined (Table 1). To examine the growth and attachment rates of microalgae, five acrylic plates (19 × 16 × 0.1 cm) were placed in a 14 L circu-lar tank filled with 10 L of f/2 medium (Guillard and Ryther, 1962). Then, full-grown microalgae at the end of the log phase were inoculated at the 10% level. The total substratum area for microalgae attachment was 5,109 cm², with the areas of the walls and floor of the tank and the plates being 1,583 cm², 451 cm², and 3,075 cm², respectively. Microalgae were cultured for two weeks, with aeration at 23℃ and 40 ㎛ol m^-2s^-1, with

continuous lighting.

Two weeks later, floating microalgae were collected us-ing GF/C filters. Attached microalgae were detached from surfaces of the substratum using a soft brush and filtered by GF/C filters. The collected microalgae from each tank were separately dried at 60℃ for 2 h and weighed to the nearest milligram. The following formulae were used for calculating the attachment and growth rates of the microalgae:

Attachment rate (%) = (Weight of attached microalgae/Total weight of microalgae) × 100

Growth rate (%) = [(n1 - n0)/n0 × 100 (n1: final collection amount (mg), n0: initial inoculation amount (mg)]

Larvae of the veliger stage were examined 56 h after fer-tilization at 21℃ and were used in the present study. Nine species among the 31 microalgae showed higher growth and adhesion rates. To culture these species, 10 mL of f/2 medium was placed inside a 6-well tissue culture chamber, and the full-grown microalgae were inoculated at the 10% level. They were cultured without aeration at 20℃ and 40 ㎛ol m^-2s^-1 of constant light for 10 days.

Each chamber in which the microalgae had been cul-tured was used for larval culture. Before larval culture, the chambers were cleaned several times with filtered seawater to remove the f/2 medium. Then, 30 larvae were placed into the chamber filled with 10 mL of filtered seawater. The cell density of the microalgae attached to the chamber was ap-proximately 5-10 × 10⁴ cells/mL. The larvae were reared at 20℃ and 40 ㎛ol m^-2s^-1 under a photoperiod of 8:16 h LD (light:dark) for 15 days. Upon the completion of larval settle-ment onto the microalgae, the culture water in the chamber was exchanged with 5 mL of filtered seawater daily. The con-trol group contained no microalgae and was filled only with filtered seawater. All experiments were repeated six times for each treatment. The settlement and metamorphosis of the lar-vae were observed through a dissecting microscope. Larvae attaching to the microalgal film and having a foot were consid-ered to be settled larvae and larvae with a shell molding were considered to be metamorphosed larvae. Larvae that showed no metamorphosis for 6 days were eliminated. The settlement rate measurement was repeated seven times. Measurements of metamorphosis rates and survival rates after metamorphosis were repeated twice. Dead larvae, immobile for more than 10 seconds with no heartbeat, were removed daily. Survival rates after metamorphosis were measured.

Settlement rate (%) = (No. of settled individuals/No. of in-oculated individuals) × 100

Metamorphosis rate (%) = (No. of metamorphosed individ-uals /No. of settled individuals) × 100

Survival rate (%) = (No. of surviving individuals/No. of metamorphosed individuals) × 100

Data were analyzed by one-way analysis of variance (ANOVA), and Duncan’s multiple range test (Duncan, 1955) was applied for the significance level (P < 0.05). SPSS version 17 (SPSS Inc., Chicago, IL, USA) was used for all statistical analyses.

The growth and attachment rates of each microalgal group

were examined (Table 2). There was no significant differ-ence in diatom biomass within group A, except for Amphiprora gigantea var. sulcata (18.1 mg) and Amphora sp. (KMMCC-823, 12.1 mg) (P < 0.05). However, the attachment rate of Amphora sp. (KMMCC-773) was the highest (97%) and that of Achnanthes sp. was the lowest (14%). The growth rates of Amphora sp. (KMMCC-823) and Amphora veneta var. coffeaeformis, 642% and 593%, were highest and that of Achnanthes sp., 150%, was lowest (P < 0.05).

The microalgal biomass of Navicula was 21.3-26.1 mg, and there was no significant difference in the biomasses of spe-cies in group B. However, the attachment rates of N. incerta, N. viridis, and Navicula sp. (KMMCC-841), 71-72%, were significantly higher than those of N. cancellata and N. ele-gans (P < 0.05). The growth rate of 1,826% for Navicula sp. (KMMCC-957) was significantly higher than those of other species (P < 0.05).

The biomasses of the diatoms Nitzschia sp. and Caloneis schroder in group C were significantly higher than those of other species, at 15-16 mg. Phaeodactylum tricornutum and Pleurosigma angulatum had the lowest biomasses at 6 mg (P < 0.05). The attachment rates of Rhaphoneis sp. and P. an-gulatum were highest at 77-80%. Cylindrotheca closterium, which had relatively high biomass, had the lowest attachment rate at 24%. However, the growth rate of C. closterium was the highest, at 2,302% (P < 0.05).

The biomasses of three kinds of Tetraselmis in group D were 12-15 mg, and were not significantly different from each other. In comparison with Tetraselmis sp. (KMMCC-109, 13%), the attachment and growth rates of Tetraselmis sp. (KMMCC-40), were significantly high, at 87% and 593%, respectively.

The biomasses of Oscillatoria splendida (14.5 mg) and Phormidium luridum (7.7 mg) in group E were the highest and lowest, respectively. The attachment rates of O. splendida (90%) and Lyngbya taylorii (83%) were significantly high, and that of Trichodesmium erytheaeum (6%) was significantly low. The growth rates of the four species of cyanophyceae, however, showed no significant differences, ranging from 676-1,378% (P < 0.05).

Nine microalgal species showing high growth and attach-ment rates were examined for their effectiveness as feed for H. discus hannai larvae. The larvae reacted to the substrate an hour after their introduction and began settling. The settlement rates of all experimental groups began to rise after 12 h and

reached their maxima after another 12 h. After 24 h, the larvae showed a tendency to constantly perish (Table 3).

After 24 h, the settlement rates in all experimental groups except for P. tricornutum were higher than that of the control group. Rhaphoneis sp. and Nitzschia sp. showed the highest rates of 98% and 91%, respectively (P < 0.05). etraselmis hazeni also indicated a high settlement rate of 80%. The settle-ent rates after 96 hours in T. hazeni and Rhaphoneis sp. were 69% and 65%, respectively, which were significantly high. However, P. tricornutum and C. closterium showed lower rates than for the control group (P < 0.05). The experimental group of Nitzschia sp. at 24 h showed a settlement rate of 91%, which was as high as that of Rhaphoneis sp. The rate, how-ever, rapidly decreased to 50% by 96 h.

The metamorphosis rates of the larvae that fed on Rhapho-neis sp. in the experimental group were 39% on the fourth day and 57% on the sixth day. This was significantly the highest rate, with metamorphosis rates in the O. splendida and T. ha-zeni groups being second highest (P < 0.05). The C. closte-rium group showed a significantly lower metamorphosis rate than did the control group, of 0.5% and 1.5% on the fourth and sixth days, respectively (Fig. 1).

The experimental group fed on Rhaphoneis sp. had a sig-nificantly higher larval survival rate on day 15 after meta-morphosis than did other experimental groups, at 67% (P < 0.05) (Fig. 2). All experimental groups except the Rhaphoneis sp., T. hazeni (42%), and O. splendida (35%) groups showed lower survival rates than the control group (P < 0.05). On the second and fourth days, all larvae in the C. closterium and P. tricornutum groups had perished. Such low survival rates meant that it was not possible to conduct an analysis of growth difference according to microalgal species.

The larvae of H. discus hannai settle to substrate within 96 h after hatching and begin metamorphosis (Morse, 1985; Searcy-Bernal, 1999; Roberts and Lapworth, 2001). Meta-morphosed larvae begin feeding on attached diatoms (Seki and Kanno, 1981; Kawamura and Takami, 1995; Roberts et al., 2007b). As prospective feed for H. discus hannai larvae, microalgae showing fast growth and high attachment rates were examined in this study.

Diatoms generally have high attachment rates compared to other microalgae, but most microalgae can adhere to substrate to a certain degree. In terms of nutritional value, diatoms, green algae, and blue-green algae generally have high con-tents of lipid, protein, and minerals, respectively (Borowitzka and Borowitzka, 1988). Diatoms can be divided into eight types in terms of the existence of mucous, their attachment form, and mobility. Among them, the best types as feed for the larvae of H. discus hannai are phlegmatic with plane form and slow mobility (Kawamura, 1994; Roberts et al., 2007a). Considering these features, this study aimed to find the best kinds of microalgae to be used as feed among diatoms, green algae such as Tetraselmis, and blue-green algae.

In this study, diatoms generally showed higher attachment rates than did other microalgae. However, Tetraselmis sp. (KMMCC-40) and the blue-green algae O. splendida and L. taylorii also had high attachment rates in the range of 83-90%. Among the diatoms, Navicula and Amphora showed higher biomass and attachment rates. The results of this study on the growth and attachment rates of microalgae can be used as a foundational resource for the development of feed for adhe-sive larvae and the mass culture of adhesive microalgae.

With respect to the settlement and metamorphosis of the abalone larvae, Cocconeis scutellum (Roberts and Nichol-son, 1997; Parker et al., 2007) and Navicula ramosissima (Kawamura and Kikuchi, 1992; Roberts and Watts, 2010) are reported to induce high settlement and metamorphosis rates in H. virginea and H. discus hannai larvae. In the present study, larvae fed Rhaphoneis sp. showed the highest settlement rate of 98% at 24 h, and T. hazeni also showed a high settlement rate of 80%. At hour 96, these two microalgae showed no signi-ficant difference in their settlement rates of over 65%. Thus, it can be concluded that T. hazeni is also suitable as a settle-ment substrate for H. discus hannai larvae (P < 0.05). The metamorphosis of the larvae attached to Rhaphoneis sp. was the highest at 57%, and T. hazeni and O. splendida showed higher metamorphosis rates (20-30%) than did other diatoms. The growth rates of P. tricornutum and C. closterium were high, but the settlement and metamorphosis rates of the larvae fed these diatoms were low. The larvae fed P. tricornutum and C. closterium perished after growing to sizes of 344 ㎛ and 305 ㎛, respectively. Thus, these two species are considered to be unsuitable as feed for H. discus hannai larvae.

Within two days after metamorphosis, the larvae of H. discus hannai begin consuming their feed (Seki and Kanno, 1981; Martinez-Ponce and Searcy-Bernal, 1998; Roberts et al., 1999). They can consume bacteria or substances other than microalgae (Garland et al., 1985; Kawamua, 1996; Kitting and Morse, 1997); however, they tend to grow faster by consum-ing microalgae such as diatoms (Garland et al., 1985; Takami et al., 1997; Roberts et al., 1999). Furthermore, before grow-ing to 800 ㎛ in shell length, the larvae of H. discus hannai are known to utilize cell secretions as their source of nutrition (Kawamura and Takami, 1995; Roberts et al., 2007b).

However, immediately following metamorphosis, the lar-vae receive nutritional substances from yolk. Thus, it is dif-ficult to observe differences in their growth according to feed type during such early developmental stages (Kawamura et al., 1998). Moreover, the premature development of the diges-tive system of the larvae at early stages makes it hard to ex-amine the effectiveness of feed (Roberts et al., 1999). For the larvae of H. discus hannai, yolk is the main source of nutrition for 10 days after metamorphosis, and they grow up to 400 ㎛ without any external source of nutrition (Takami et al., 2000).

In this study, larvae which fed on microalgae except for Rhaphoneis sp. and the control group that fed on no microalgae showed no difference in growth immediately following metamorphosis. Such results are explained by the fact that, in the early developmental stage, larvae feed on yolk more than on micro-algae. The control group, which did not feed on microalgae, had 500 ㎛ larvae, and 36% of its survival rate could be explained by the utilization of bacteria and organic substances in the seawater as nutrition sources.

The diameter of the larval mouth after metamorphosis mea-sures only 10 ㎛; thus, the size of microalgae is also important as a condition of suitable feed for abalone larvae (Kawamura et al., 1998). The sizes of microalgae used in this study, except for Rhaphoneis sp., were over 10 ㎛. Rhaphoneis sp., being smaller than 10 ㎛, was a suitable feed for the larvae after the metamorphosis phase. Considering these results, the size, form, and nutrition and the attachment form, rate, and strength of each microalgae species should influence the settlement and growth of the larvae (Matthews and Cook, 1995; Roberts et al., 1999; de Vicose et al., 2010).

In this study, green algae and blue-green algae other than diatoms were also found to be prospective feeds for the larvae of H. discus hannai. The high content of protein in green al-gae and rich minerals in blue-green algae are inferred to have a positive influence on the formation of the larval shell of H. discus hannai. In conclusion, the most suitable microalgae for rearing H. discus hannai larvae was the diatom Rhaphoneis sp. The green alga T. hazeni and the blue-green alga O. splen-dida may also be useful when combined with Rhaphoneis sp.