Rotifers are some of the most important zooplankton, and are the focus of much attention from aquaculture scientists. Rotifers have been used as a primary live food for the seedling production of many economically important marine animals, because rotifers can be easily cultured at high densities. Stable mass culture of rotifers is needed for successful fish larval rearing. A variety of mass culture methods should be devel-oped to facilitate successful aquatic animal culture through stable live food organism production. However, the instability or sudden crashes of rotifer mass cultures remain problematic. The causes of such culture failures are not fully understood, although it is predicted that one major cause is contamina-tion by other organisms such as protozoa (Takayama, 1979; Reguera, 1984; Chen et al., 1997), copepods (Fukusho et al., 1976) and bacteria (Yu et al., 1990).

Coexing populations of different taxonomic groups depend on the organisms that occur together in space and time, and interact with each other through the processes of mutualism, parasitism, predation and competition (Begon et al., 1990). The relationships between Brachionus rotundiformis and other zooplankton species have been well examined (Gilbert and Stemberger, 1985; Gilbert, 1988; Hagiwara et al., 1995a; Jung et al., 1997), and competitive interference has been re-ported between rotifers and cladocera in fresh water. Under marine culture conditions, there is considerable evidence for interspecific relationships between a rotifer, B. rotundiformis, and two species of copepods, Tigriopus japonicus and Acartia sp. (Jung et al., 1997).

Copepods and rotifers coexist in estuaries as well as in brackish water fish culture ponds in North Sulawesi, Indone-sia. Of several copepod species collected, the only one that could be adapted to laboratory culture was Apocyclops bor-neoensis. It is unknown which species (B. rotundiformis or A. borneoensis) dominates earlier, or which is more susceptible to the rigorous conditions of the ponds. Therefore, it is of in-terest to examine the nature of the relationship between these two species. Such information is useful both for assessing their ecological significance in a brackish water ecosystem, as well as for establishing techniques for mono- and mixed-spe-cies cultures for aquaculture purposes as live food organisms.

Here, I observed, the survival strategies involved in the in-terspecific relationship between B. rotundiformis and A. bor-neoensis under laboratory experimental conditions. I focused on the interspecific interactions of the rotifer and the copepod from a microcosm viewpoint of aquaculture ecology.

Copepods and rotifers were collected from a milkfish pond in Bitung, 30 km east of Manado, North Sulawesi, Indonesia. The pond is separated from the adjacent sea by mangroves, but is connected through an inlet during high tide. Throughout the year, salinity of the pond varies from 12 to 25 psu and temperature ranges from 29℃ to 35℃.

The specimens were kept in darkness during a three-day acclimation culture to laboratory conditions before isolation. Various copepods were included in the sample, but only a cyclopoid copepod survived. The species was identified as Apocyclops borneoensis by Dr. H-S. Kim (Research Institute for Basic Science, Cheju National University of Korea). The rotifers were morphologically analyzed by Fu et al. (1991), and evidently belonged to an ultra minute strain of Brachionus rotundiformis (Hagiwara et al., 1995a). We refer this rotifer B. rotundiformis as a Bitung strain.

Experimental design and conditions used were the same as those described by Hagiwara et al. (1995b). Salinity, tempera-ture, and culture volume were 22 psu, 25℃ and 40 mL, re-spectively. The organisms were cultured in total darkness. The initial number of animals in mixed cultures was 20 females of the Bitung rotifer strain of B. rotundiformis and 3 ovisac-bearing females of A. borneoensis. In the monocultures, the numbers of rotifers and copepods were the same as in mixed cultures, but were cultured separately. Mono- and mixed-spe-cies cultures were conducted with three replicates for 16 days. Stereo-microscopic observation was carried out on fresh cul-ture medium including Tetraselmis suecica (7 × 10⁵ cells/mL) every two days. Total number of test animals was counted and remaining algal food density was also counted by a haemacy-tometer (Kayagaki Irika Kogyo Co. Ltd., Tokyo, Japan). The algal food T. suecica, was grown in modified Erd-Schreiber medium (Hagiwara et al., 1994) and centrifuged. Density of food added was 7 × 10⁵ cells/mL, and was readjusted every two days after observation.

For the observation and calculation of the rotifer mixis rate (%), all individual non-egg bearing fe-males, amictic females, unfertilized and fertilized mictic females, males and resting eggs were counted, and the mixis rate was calculated (Hagiwara et al., 1988). The numbers of all individual nauplii, copepodites and egg-bearing females of the copepod A. bor-neoensis were recorded. Relative population growths between monocultures of each species and mixed culture conditions were compared by student's t-test.

Predator-prey interactions were not observed during this experiment between the experimental rotifer B. rotundifor-mis and copepod A. borneoensis (Fig. 1A and 1B). However,

distinct correlation interactions between the two experimental organisms were observed (Fig. 1C).

The remnant food (T. suecica) volumes under rotifer culture conditions significantly decreased with the lapse of experi-mental time compared to those in copepod monoculture con-dition (P < 0.05) (Fig. 2A and 2C), and almost all microalgae (T. suecica) cells were consumed. A remnant food volume of approximately 7 × 10⁵ cells/mL was maintained under cope-pod monoculture conditions (Fig. 2B).

Population growth of rotifers in the mixed culture was sig-nificantly decreased (P < 0.05) (Fig. 3). Growth of rotifers was suppressed by the presence of copepods in the culture container. However, the population growth of rotifers in the monospecific culture continually increased (Fig. 3).

In both experimental conditions, inductions of mixis rates (%) were observed. However, the mixis rate (%) of rotifers was notably changed by presence or absence of copepods. The difference between rotifer monoculture and mixed culture

with copepods was particularly notable on the 4th day (Fig. 4). The presence of copepods influenced the mixis rate (%) of co-existing rotifers (Fig. 4), as well as the production numbers of rotifer resting eggs. On day 4 of the observations, the highest values of mixis rates (%) was observed in both experimental conditions (Fig. 4), and the production of rotifer resting eggs was observed from the 6th day. However, there were differ-ences in the numbers of rotifer resting eggs produced. More resting eggs were formed in the copepod mixed culture than in the rotifer monoculture (P < 0.05) (Fig. 5). A comparison of the production of rotifer resting eggs between rotifer mono-culture and mixed culture with copepods is shown in Table 1.

The production of rotifer resting eggs was 1.56 times higher in the mixed culture than in the rotifer monoculture (P < 0.05) (Table 1), with 1,327 ± 213 (mean ± SD) resting eggs in mixed cultures, and 853 ± 171 (mean ± SD) in monocultures.

Fig. 6 shows that copepod population growth did not differ between copepod monoculture and mixed culture with roti-fers during the 16 day experimental period. This trend was not changed with the separation of the developmental stages (nauplius, copepodid and egg-carrying females) of the cope-pods (Fig. 7).

Rotifer mass culture tanks comprise intricate relationships among the constituents of the small ecosystem. These small ecosystems are mainly composed of rotifers and contaminant micro-organisms, such as bacteria, microalgae, protozoa and copepods. Many of these contaminant micro-organisms affect rotifer population growth through interactions such as exploit-ative competition for food, conmensalism, ammensalism and physical interference competition (Hagiwara et al., 1995b; Jung et al., 1997).

No predator-prey relationship was observed between the two zooplankton species during the experimental period. However, interference between the two species was observed. It was noted that the copepod used a filter feeding mechanism in which it filtered the slowly swimming rotifers together with algae (food), but immediately rejected the rotifers, which started to swim. Rejected rotifers appeared to be unharmed, but may have suffered some physical damage or injury. The physical damage by copepods may have brought about the de-crease in rotifer population growth.

Microalgae, including Nannochloropsis, Dunaliella, Tetra-selmis and enriched freshwater Chlorella are commonly used foods in the mass culture of rotifers (Witt et al., 1981). Of these, Tetraselmis is well known among aquaculturists as a good food for rotifer and copepod cultures. In this study, in both mono and mixed cultures of rotifers, the density of T. suecica continuously decreased. However, in the monoculture of the copepod A. borneoensis, algal density did not decrease, but maintained a constant level after small amounts were used as food by the copepods, and the added food volume appeared sufficient to maintain algal density. This indicated that A. bor-neoensis may not eat much and/or did not actively feed on T. suecica in monoculture. This suggests that there was not competition for food, and that sufficient amounts of food (T. suecica) were supplied for the copepod A. borneoensis under these experimental conditions.

The presence of the rotifer did not influence the popula-tion growth of the copepods, likely because of the large size differential between B. rotundiformis (c. 02 mm) and A. bor-neoensis (c. 0.8 mm). However, the copepod suppressed ro-tifer population growth after 10 days in mixed culture. The population growth of B. rotundiformis in mixed culture was associated with a reduction in the numbers of both non-egg bearing females and egg-bearing females.

An analysis of the population structures revealed that the rotifer population growth decreased due to a 30% reduction in females without eggs and amictic females. In contrast, mictic females in mixed cultures were more abundant than those in monocultures. Thus, the presence of copepods stimulated the sexual reproduction of the rotifer (mixis rate or resting egg formation), which in turn caused decreased rotifer population growth. It is of interest to conduct further research to clarify this mechanism.

The dormant stage of the rotifer (resting egg or cyst) is very resistant to harsh environmental conditions and may be dis-persed over wide areas by wind, water or migrating animals. Sometimes, scientists induce the production of mixis repro-duction (cyst) as an easy method of storing and transport-ing for marine fish larvae culture or aquaculture study. The hatching of rotifer resting eggs is caused by stimulation from light, temperature, salinity and some chemicals (reviewed by Hagiwara, 1996), but there is no known method to artificially produce rotifer cysts.

The high mixis rate (70% or more) in this study could be related to the algal species used as food. Tetraselmis tetrathele as a food source stimulated a high mixis rate (81%), which led to the successful mass production of resting eggs (101 x 10⁶ Ind.) of the Hawaiian strain of B. rotundiformis (Hagiwara and Lee, 1991). Another report indicated that bacteria populations in mass cultures regulated the sexual reproduction of rotifers (Hagiwara et al., 1994). Two marine copepod species were successfully cultured when consuming bacteria alone (Rieper, 1978). This mechanism has also been discussed with the ef-fects of bacteria on interspecific relationships between B. ro-tundiformis and the two copepod species, Tigriopus japonicus and Acartia (Hagiwara et al., 1995b; Jung et al., 1997). The present study also highlights the significance of bacterial flora on the population growth of A. borneoensis.

Population growth of A. borneoensis in mixed culture did not differ compared with that in monoculture. However, the number of nauplii decreased to 17% and the number of egg bearing females was 1.62 times higher in mixed culture. The reason for this was not clear, but may be associated with the feeding mechanisms of copepods such as Apocyclops, which may utilize feces more efficiently than they do bacteria in the water column. Various copepod species used as live food or-ganisms utilized bacteria as their food. The common marine copepod species Tigriopus japonicus utilized bacteria (Jung et al., 1998) and rotifer feces (Jung et al., 2000) as food in the early nauplius developmental stages. The individuals in each developmental stage may also have different feeding habitats, due to changes in their mandible structures (Itoh, 1970). The experimental copepod Apocyclops appears to utilize egg sac conservation for the preservation of the species. This may explain the low survival rate of Apocyclops nauplii that was observed in mixed cultures.