The variation of volume transport during the period from 1965 to 2000 through the western channel of the Korea Strait was estimated by obtaining an relation function between the ADCP volume transport and the geostrophic volume transport estimated by the sea level difference between Pusan and Izuhara. The estimated climatological mean volume transport during past 36 years has seasonal variation with a minimum of 1.15 Sv in February and a maximum of 1.88 Sv in October. The mean volume transport for 36 years is 1.51 Sv. The annual mean volume transport has an interannual variation with a minimum of 1.26 Sv in 1968 and maximum of 1.90 Sv in 1973, with three dominant periods of variations of 14.96 years, 4.96 years and 2.99 years.

Sequences of the large subunit ribosomal (LSD) rDNA D1-D2 region of Alexandrium catenella(=A. sp. cf. catenella) and A. tamarense isolates, which were collected along the Korea coasts, were analyzed to understand their phylogenetic relationships and geographical distributions. All A. catenella and A. tamarense isolates belonged to the A. tamarense/catenella/fundyense complex and were grouped with the North American and temperate Asian ribotypes, respectively, regardless of the presence or absence of a ventral pore in the first apical plate. A consistent and peculiar characteristic that differentiated the Alexandrium isolates was amplification of a second PCR product with a lower molecular weight in addition to the predicted one; ten A. catenella isolates belonging to the temperate Asian ribotype yielded this additional PCR product. Sequence alignment revealed that the shorter PCR product resulted from an unusual large deletion of 87 bp in the LSD rDNA D1 domain. The North American and temperate Asian ribotypes were prevalent along the Korean coasts without geographical separation. Given the high genetic homogeneity among widely distributed Alexandrium populations, each ribotype appeared to be pandemic rather than to constitute a distinct regional population.

LSU rDNA D1-D2 and SSU rDNA genes of 23 strains in seven Alexandrium (Halim) species, A. tamarense (Lebour) Balech, A. catenella (Whedon et Kofoid), A. fraterculus (Balech) Balech, A. affine (Inoue et Fukuyo) Balech, A. insuetum Balech, A. pseudogonyaulax (Biecheler) Horiguchi ex Yuki et Fukuyo and A. tamiyavanichii Balech, were sequenced and the data were used for molecular phylogenetic analysis. The sequence data revealed 11 and 7 ribotypes in the LSU rDNA D1-D2 region and 4 and 17 ribotypes in the SSU rDNA region of A. catenella and A. tamarense, respectively. Other Alexandrium species had also 1 to 5 ribotypes in the two regions. With the exception of CMC2 and CMC3 of A. catenella, all A. tamarense and A. catenella strains had a common ribotype, a functionally expressed rRNA gene (here termed type A), in both gene regions. In addition to the functionally expressed gene, several pseudogenes were obtained that were found to be good tools to analyze the population designation of regional isolates by grouping them according to shared ribotypes. From the phylogenetic analysis of the sequence data determined in this study and retrieved from GenBank, the genus Alexandrium was divided into 14 groups: 1) A. tamarense, 2) A. excavatum, 3) A. catenella, 4) Tasmanian A. tamarense, 5) A. affine (and/or A. concavum), 6) Thai A. tamarense, 7) A. tamiyavanichii, 8) A. fraterculus, 9) A. margalefii, 10) A. andersonii, 11) A. ostenfeldii, 12) A. minutum (or A. lusitanicum), 13) A. insuetum, and 14) A. pseudogonyaulax. The SSU rDNA gene sequence of A. fundyense was so similar to those of A. tamarense used in this study that the two species were difficult to discriminate each other. A. tamiyavanichii was closest to the A. tamarense strain isolated in Thailand and close to the long chain-forming species of A. affine and A. fraterculus. The phylogenetic tree showed that A. margalefii, A. andersonii, A. ostenfeldii, A. minutum and A. insuetum constituted the basal relative complex, and that A. pseudogonyaulax is an ancestral taxon in the genus Alexandrium.

Since seagrasses are highly productive and provide a source of organic carbon for a wide variety of marine organisms in coastal and estuarine ecosystem, accurate assessment of seagrass production is critical to understand the functions and values of seagrasses in these ecosystems. Zieman's leaf marking technique has been mostly used to estimate seagrass leaf production rates. However, inherent problems on the traditional leaf marking technique have been discussed by the several researchers, and these problems can cause underestimation of seagrass production. To develop an accurate and reliable assessing method for seagrass production, production rates of eelgrass, Zostera marina in three bay systems on the south coast of the Korean peninsula were estimated using the conventional leaf marking technique and the plastochrone method. The plastochrone method has been recently suggested as an effective method for reliable assessments of seagrass production. In the present study, leaf production rates estimated by the plastochrone method were significantly higher than the rates derived from the traditional leaf marking technique. Annual eelgrass leaf production assessed using the leaf marking technique was about 65 to 89% of the estimated production using the plastochrone method. The differences in annual productions between assessment techniques imply that the conventional leaf marking technique significantly underestimated eelgrass leaf production. Total eelgrass productions estimated using the plastochrone method in the present study sites were about 600 to 806 g DW $m^{-2} y^{-l}$, and below-ground production accounted for about 20 to 23% of the total production. The plastochrone method was suggested to be an effective and accurate assessing method for eelgrass production.

Mussel, Mytilus galloprovincialis, was transplanted reciprocally between contaminated site (St. STP) and relatively less contaminated site (St. UB) in Onsan Bay, Korea in order to estimate heavy metal accumulation and release. Transplanted and indigenous mussels were collected 8 times over 108 days deployment at both sites and analyzed for Cd, Cu, Pb and Zn using ICP/MS. Cu and Zn concentration did not show any significant differences between transplanted and indigenous mussels throughout the experiment period, although dissolved Cu and Zn concentrations in seawater were significantly higher at contaminated site. Pb accumulated rapidly, while Cd did not show any accumulation in the mussels transplanted from St. UB to St. STP. These accumulation patterns might reflect the variation of dissolved metal concentration; dissolved Pb concentration was significantly higher in St. STP than St. UB, while dissolved Cd concentration was similar between both sites throughout this experiment. Release of Cd and Pb in the mussels transplanted from St. STP to St. UB was not significant during the transplantation period. The levels and variation of Pb and Cd concentrations in mussels transplanted from St. STP to St. UB were similar to those in the indigenous mussels at St. STP throughout the period. Therefore, Cd and Pb accumulated in indigenous mussels may indicate the integrated metals throughout their lives, not responding to the present status of seawater because the release of accumulated metals in mussels occurs very slowly when the seawater metal levels are lowered. On the other hand, since the transplanted mussels can respond immediately to the levels of metals in seawater, it may be suggested that the present status of heavy metals in seawater can be only acquired from the metal levels of mussels transplanted from clean environments.