Introduction

Marine dinoflagellates are a major component in plankton communities and play an important role as primary producers in marine ecosystems. They are composed of autotrophic, heterotrophic, and mixotrophic species, and some of which produce during their life cycle resting cysts that can be preserved in sediments (e.g., Head 1996). The encystment of dinoflagellates is known to be affected by water temperature, nutrient availability, day length and endogenous encystment rhythms (Taylor 1987), whereas cyst distribution in sediments has been related to environmental factors such as water temperature, salinity, turbulence and the availability of nutrients (Dale 1996; Marret and Zonneveld 2003). This suggests that dinoflagellate cysts recorded from sediments can be utilized to reconstruct environmental conditions in a given study area. Many ecologists and paleontologists have tried to reconstruct the past environmental characteristics using dinoflagellate cysts, as understanding the past environmental conditions help to develop information for interpreting on environmental changes of present and future. However, in Korean coastal areas previous studies concerning dinoflagellate cysts have concentrated on only understanding their distributional characteristics (Lee and Yoo 1991; Lee and Matsuoka 1994), and there is no information available yet on the relationship between dinoflagellate cyst distribution and environmental conditions.

The relationship between dinoflagellate cysts and environmental conditions has been well documented around the world (e.g., Zonneveld 1997; Zonneveld et al. 2000; Marret and Zonneveld 2003; Radi and de Vernal 2004; Radi et al. 2007; Shin et al. 2007; Pospelova et al. 2008). In particular, paleo-environmental studies and ecological studies of eutrophied or industrial areas are increasingly being conducted with the use of dinoflagellate cysts (e.g., Sætre et al. 1997; Dale et al. 1999; Matsuoka 1999; Pospelova et al. 2002, 2005; Matsuoka et al. 2003; Dale 2009). However, the relationship between specific dinoflagellate cysts and environmental conditions has yet to be fully understood. For instance, Lingulodinium machaerophorum (Deflandre and Cookson) Wall (the cyst name for Lingulodinium polyedrum), and Selenopemphix quanta (Bradford) Matsuoka (the cyst name for Protoperidinium conicum) were regarded as species responding to cultural eutrophication in the Oslofjord of Norway (Dale et al. 1999; Dale 2001). However, these species do not occur abundantly in Tokyo Bay, Japan, which is also a heavily eutrophied area (e.g., Matsuoka 1999, 2001; Matsuoka et al. 2003). This may be due to differences in habitat preferences between geographically separated populations of the same species and/or competition with a different set of taxa (Smayda and Reynolds 2003). In other words, dinoflagellate cysts as indicators of environmental conditions may also vary in different ecosystems.

Coastal areas, including estuaries, are influenced by domestic and industrial waste waters, urban sewage, and agriculture effluents caused by anthropogenic activities, as well as freshwater runoff from rivers, and because of these, their biology and physical dynamics are often altered. Pospelova et al. (2002) reported on the relationship between dinoflagellate cyst abundance and human activity in estuarine area of Massachusetts, USA and suggested that more studies are needed to better understand such relationships as indication of eutrophication by dinoflagellate cysts varied in different areas and different eutrophication levels. Although the coastal areas of Korea, particularly the southern areas, are exposed to the open ocean, adjacent human activities cause heavily polluted surface waters. However, in Korean coastal areas, use of dinoflagellate cysts as indicators of environmental changes caused by the human activities is not well known.

In this study, we document the records of dinoflagellate cyst assemblages and report their relationships with environmental characteristics in southern coastal area of Korea. We also suggest a number of dinoflagellate cyst taxa that can be considered as indicators of eutrophication in Korean coastal areas.

Study Area

The study area consists of two semi-enclosed bays, namely Gamak Bay and Yeoja Bay, and the offshore area of Yeoja Bay located on the south coast of Korea (Fig. 1). Annual precipitation in this region amounts to 150 cm, with most of this falling in summer (from July to September). The amount of freshwater entering the bay through streams and ditches ranges from 2.7 × 104 to 18.0 × 104 m3 day−1 (average, ∼7.0 × 104 m3 day−1), with a large seasonal variation. Gamak Bay is oval in shape and approximately 9 km in width and 15 km in length, shallow with a mean depth of about 9 m and surrounded by the Yeosu Peninsula and Dolsan Island to the east. Surface sediments consist mainly of silty clay and clayey silt, however, slightly gravelly mud is also distributed near the channels at the bay mouth (Kim et al. 2000). Waters in this bay are exchanged mainly through the southern entrance of the bay and additionally through a narrow northeastern channel (Koo et al. 2004), and the direction of tidal flood current is shown in Fig. 1b. The bay is separated into two sub-basins (northwestern and southern basins) by a ridge that runs from east to west. The numerous oyster farms, which are densely positioned above the ridge, account for about 30% of total annual oyster production in Korea (Lee and Cho 1990), while mussel farms and fish cultures are also located throughout the bay (Fig. 1b). Filter feeders, including oysters, generally feed on diatoms, ciliates, and flagellates and their filtration activity may have an influence on local phytoplankton dynamics (Vaquer et al. 1996). Nevertheless, seasonal succession of phytoplankton assemblages in Gamak Bay has been clearly linked with nutrient dynamics (Yoon 2000). Human activity on the ridge and the dense presence of farms reduce water circulation between the northwestern and southern basins, and also between surface and bottom waters (Lee and Cho 1990; Kim et al. 2006). In particular, aquaculture wastewaters and the increase of urban sewage discharged from Yeosu City have led to eutrophic conditions since 1990 (Lee and Cho 1990; Koo et al. 2004). Anoxic conditions frequently appear in bottom waters of the northwestern area of Gamak Bay, with less than 2.0 mg/L of dissolved oxygen in the summer season (Kim et al. 2006). In addition, the frequency of harmful algal blooms has increased since 1990 (Park and Yoon 2003). Based on these facts, this bay can be considered as a typical eutrophied area.

Fig. 1
figure 1

Study area in the southern coast of Korea. a Circulation pattern of Tsushima Warm Current (TWC), Yellow Sea Warm Current (YSWC), Cheju Warm Current (CWC), and Chinese Coastal Current (CCC) suggested by Teague et al. (2003). b Sediment sampling stations (dots), observation sites of NFRDI (stars), and direction of tidal flood current in Gamak Bay (Lee 1992) and Yeoja Bay (Lee et al. 2000)

Full size image

Yeoja Bay is surrounded by Yeosu Peninsula to the east and the Goheung Peninsula to the west, and is connected to the South Sea of Korea through narrow inlets at the southern entrance of the bay. The bay is shallow with 5.4 m in average water depth, and extensive mud flats occur along the coast. Surface sediments consist mainly of silty clay with 6.0–8.7 Md ϕ, coarse sediments are dominant in the north and mouth of the bay (Hwang et al. 2005), and its distribution is related to a clockwise circulation of seawater during flood tides (Fig. 1b) (Lee et al. 2000). The bay is also characterized by an outstanding semidiurnal tide producing periodical and strong ebb tidal movements (Lee et al. 2000). Freshwater is loaded into the bay by the Beolgyo, Isa, and Dong river that flow through Suncheon City, discharging fluvial water of 0.07, 0.19, and 0.14 km3/year, respectively. However, despite the intrusion of fresh waters with nutrients, the environmental conditions of this bay are less polluted, as indicated by chemical oxygen demand, total sulfide, and ignition loss of sediments (Hue et al. 2000; Lee and Youn 2000).

The offshore area of Yeoja Bay is strongly affected by intrusion from open seawaters such as the Tsushima Warm Current and Yangtze-River-diluted water derived from the Chinese coast (Beardsely et al. 1985; Lee 2006; Kim et al. 2007) (Fig. 1a). Although the intrusion of the Yangtze-River-diluted water, which is characterized by high temperature (>22°C), low salinity (<29 psu), and highly dissolved inorganic nitrogen (DIN) concentration, is related to the occurrence of red tides in summer, the area is relatively less eutrophicated (Kim et al. 2001). Water depth is about 20 m, while mean water temperature and salinity recorded in August were 26.2°C and 31.7 psu, respectively (Choi 2001). Salinity is similar to that of Gamak Bay (Kim 2002).

Materials and Methods

Sampling and Dinoflagellate Cyst Analysis

Surface sediment samples for dinoflagellate cyst analysis were collected from 11 stations with a gravity corer in Gamak and Yeoja Bays in May 2006 (GMB1, GMB2, and GMB3 and YJB1, YJB2, YJB3, and YJB4) and in the offshore area of Yeoja Bay in January 2007 (YJB5, YJB6, YJB7, and YJB8). The top 2 cm of core samples were sliced and all of them stored in dark and cool conditions at 4°C prior to further analysis. Sedimentation rates in Gamak Bay and Yeoja Bay were estimated using 210Pb dating method, which were averaged at 2.5 and 1.1 cm/year, respectively (Shin et al. 2010a, b). Based on these results, we assumed that the top 2 cm of the sediments from the offshore area of Yeoja Bay represent less than 3 or 4 years of deposition.

The samples were processed according to the palynological method of Matsuoka and Fukuyo (2000) as follows: Approximately 2 g of each sample was placed into an acid-resistant 100-ml beaker and then treated to remove calcium carbonate and silicate materials with 10% HCl and 47% HF at room temperature, respectively. The residues were rinsed with distilled water, sonicated for about 30 s, and sieved with stainless steel screens of 125- and 20-μm opening mesh size. The residues on the 20-μm mesh were made up to a 10-ml aliquots by adding distilled water. Identification of dinoflagellate cysts was carried out on a 1-ml subsample of the 10-ml aliquot under an Olympus IX 70 interference optical microscope. Alexandrium-type cysts were identified with the Primuline-staining direct count method of Yamaguchi et al. (1995) because primuline stains cysts after the treatment with acids (Shin et al. 2010b).

For measurement of water content, a part of each sample was weighed wet and then dried in an oven under 70°C for 24 h. In this study, cyst concentrations are shown as the number of cysts per gram of dry weight sediment (cysts g−1 dry weight).

Statistical Treatments and Environmental Data

Cluster analysis was conducted with the PRIMER software, which contains the Bray-Curtis similarity coefficient, using the components of cyst concentration and the number of species in order to clarifying the characteristics of dinoflagellate cyst assemblages.

The mud content was measured with a laser particle size analyzer (SALD-3100, Shimadzu), and defined less than 63 μm as the proportion of sediments. Mean surface water temperature, salinity, DIN and chlorophyll a concentration in February and August from 2005 to 2006 were obtained from five sites of National Fisheries Research and Development Institute (Table 1; Fig. 1b). According to de Vernal et al. (1994) and Radi et al. (2001), the environmental data for the coldest and warmest months provide information on the annual hydrographical cycle, which seems to be more of a determinant for the dinoflagellate cysts distribution that seasonal or annual means. Consequently we selected the environmental data obtained made in February and August. In order to identify the environmental characteristics of our study area, principal component analysis (PCA) was performed on the obtained environmental parameters using SPSS version 10.0 for windows. Results are summarized in Table 2.

Table 1 Mean surface water temperature, salinity, dissolved inorganic nitrogen (DIN) and chlorophyll a concentration of 2005 and 2006 obtained from NFRDI (http://www.nfrdi.re.kr)
Full size table
Table 2 Principal component analysis results (eigenvalues, eigenvectors, and scores) from environmental parameters
Full size table

Results

Dinoflagellate Cyst Assemblages

A total of 35 dinoflagellate cyst taxa were identified from the 11 studied stations (Tables 3 and 4), some of which are shown in Fig. 2. Concentrations of the most abundant taxa at each station are shown in Fig. 3. Cysts produced by Protoperidinium (Brigantedinium spp.) and Gonyaulax (Spiniferites spp.) were dominant, while Brigantedinium spp. were particularly abundant in the offshore area of Yeoja Bay and Gamak Bay. Spiniferites spp. mainly composed of Spiniferites bulloideus (Deflandre and Cookson) Sarjeant (the cyst name for Gonyaulax scrippsiae) were only abundant at YJB6 and YJB8. Cysts of Polykrikos kofoidii Chatton and Polykrikos schwartzii Bütschli were abundant at YJB6, YJB8, and GMB3. Cysts of Alexandrium affine (Inoue and Fukuyo) Balech type and Alexandrium catenella (Whedon and Kofoid) Balech/tamarense (Lebour) Balech type were abundant in the offshore area of Yeoja Bay and Gamak Bay. L. machaerophorum (Deflandre and Cookson) Wall and S. quanta (Bradford) Matsuoka appeared to be relatively low in abundance.

Table 3 List of identified dinoflagellate cysts from surface sediments
Full size table
Table 4 Species composition and abundances (cysts g−1 dry weight) of dinoflagellate cysts collected from the surface sediments and mud contents in each station
Full size table
Fig. 2
figure 2

Photomicrographs of some dinoflagellate cysts collected in surface sediments from the southern coastal waters of Korea: a Spiniferites bulloideus, b Spiniferites bentorii, c Lingulodinium machaerophorum, d Cyst of Alexandrium affine type, e Tuberculodinium vancampoae, f Cochlodinium spp., g Cyst of Gymnodinium catenatum, h Oblea acanthocysta, i Cyst of Protoperidinium americanum, j Brigantedinium irregulare, k Cyst of Protoperidinium cf. minutum, l Quinquescuspis concreta, m Selenopemphix quanta, n Trinovantedinium applanatum, o Cyst of Polykrikos kofoidii, p Cyst of Polykrikos schwartzii, q Stelladinium reidii, r Votadinium spinosum, s Selenopemphix nephroides, t Votadinium calvum, u Trinovantedinium pallidifulvum, v Protoperidinium latissimum, w Cyst of Alexandrium catenella/tamarense type under normal light after staining with primuline, and x Cyst of Alexandrium catenella/tamarense type under UV light after staining with primuline. Scale bars: 10 μm

Full size image
Fig. 3
figure 3

Concentrations of total number of cysts (A), and the most dominant cyst taxa: Spinìferites spp. (B), Brigantedinium spp. (C), cysts of Polykrikos kofoidii/schwartzii (D), cysts of Alexandrium tamarense/catenalla type (E), Spiniferites bulloideus (F), Selenopemphix quanta (G), Selenopemphix nephroides (H), cysts of Protoperidinium americanum (I), Lingulodinium machaerophorum (J), and relative abundances (%) of autotrophic and heterotrophic dinoflagellate cysts (K) in surface sediments from the southern coast of Korea

Full size image

Dinoflagellate Cyst Distribution

Cyst concentration ranged from 173 to 1,276 cysts g−1 (Fig. 3), and the highest cyst concentration was recorded at YJB6 and the lowest at YJB5. Cyst concentration in the offshore area of Yeoja Bay was relatively higher than in Gamak and Yeoja Bays. Heterotrophic cysts accounted for more than 50% of total assemblages in all samples (Fig. 3). The highest proportion of heterotrophic cysts was recorded at GMB1 (74.2%) and the lowest at YJB4 (51.6%). The proportion in Gamak Bay was generally higher than in Yeoja Bay and the offshore area of Yeoja Bay.

Based on cluster analysis, dinoflagellate cyst assemblages of the study area were divided into two main groups at a level of 71 Bray-Curtis similarity coefficient; Group I consisted of the assemblages in Yeoja Bay (YJB1, YJB2, YJB3, YJB4, and YJB5) characterized by low cyst concentration with relatively low proportion of heterotrophic cysts and group II consisted of the assemblages in Gamak Bay and the offshore area of Yeoja Bay (YJB6, YJB7, and YJB8 and GMB1, GMB2, and GMB3) characterized by relatively high cyst concentration, with high proportion of heterotrophic cysts (Fig. 4). Protoperidinium cysts were abundant in stations of group II. However, Selenopemphix nephroides Benedeck (the cyst name for Protoperidinium subnerme) and cysts of Protoperidinium americanum (Gran and Braarud) Balech were abundant in group I, and P. americanum cysts are also observed in inner part of Gamak Bay (Fig. 3).

Fig. 4
figure 4

Dendrogram and separated areas based on results from cluster analysis

Full size image

Environmental Analysis

PCA produced two important components, PCA 1 and PCA 2, which explains together 79% of total variation of the selected environmental parameters (Table 2). The eigenvectors for environmental parameters showed that PCA 1 is positively correlated with salinity, and has a negative correlation with water temperature and DIN in August. PCA 2 is positively correlated with DIN in winter and chlorophyll a concentration in August. Sample scores of the two components show distribution of the stations according to differences in salinity and seasonal variations of DIN (Fig. 5). Gb1, Gb2, and Yb3, which had a positive correlation with PCA 1, are located in the outer part of Gamak Bay and the offshore area of Yeoja Bay, where are strongly affected by open seawaters. However, Yb1 and Yb2, which had a negative correlation with PCA 1, are located in Yeoja Bay, where are easily affected by intrusion of freshwater. In addition, Gb1, Gb2, and Yb1, which has a positive correlation with PCA 2, are characterized by high DIN in February and chlorophyll a concentration in August. Salinity level in Gamak Bay (Gb1 and Gb2) and offshore area of Yeoja Bay (Yb3) was relatively higher than Yeoja Bay (Yb1 and Yb2).

Fig. 5
figure 5

PCA ordination of stations based on environmental parameters

Full size image

The mud content is more that 90% at all stations (Table 4), which suggests a very similar hydrodynamic system across sites.

Discussion

Characteristics of Dinoflagellate Cyst Assemblages

Marret and Zonneveld (2003) summarized the global distribution of extant dinoflagellate cysts from recent sediments and their relationship with environmental conditions based on integrative data from the literature. In comparison with their results, the species composition of dinoflagellate cysts in this study generally resembled that of surface sediments in other temperate regions (e.g., Cho and Matsuoka 2001; Pospelova et al. 2002; Matsuoka et al. 2003; Wang et al. 2004; Shin et al. 2007), and dominant species such as Brigantedinium spp., Spiniferites spp. and Polykrikos cysts were cosmopolitan. Of them, S. bulloideus has been reported to be abundant in Asian coastal areas such as the southern coastal areas of Korea (Lee and Matsuoka 1994; Shin et al. 2007), and southern Japanese waters (Matsuoka 1985) which is characterized by high salinity. Matsuoka (1985) reported that the abundant occurrence of S. bulloideus with Spiniferites spp. along coasts of western Japan may be due to the effect of Tsushima Warm Current. Ellipsoidal Alexandrium cysts are also abundant in those areas together with East China Sea and Chinese coasts (e.g., Cho and Matsuoka 2001; Wang et al. 2004). According to Marret and Zonneveld (2003), high relative abundances of S. bulloideus and ellipsoidal Alexandrium cysts are related to high salinity. These species are abundantly found along offshore areas of Yeoja Bay and Gamak Bay, which may be related to the effect of open seawaters.

Distribution of Dinoflagellate Cyst Assemblages and Environmental Characteristics

Tidal dynamics can control gradients of bottom sediment texture, and consequently distribution of dinoflagellate cysts (Marret and Scourse 2002). As such, the presence of strong tidal currents may lead to the transport of dinoflagellate cysts from places where they were originally produced. However, since mud content was very high in all stations of study area (>90%) and Gamak Bay and Yeoja Bay characterized by low water depth are enclosed by several small islands located in the southern and northeastern channels, there is no expectation of significant transport of produced organic particles, including dinoflagellate cysts.

Nevertheless, cyst concentrations within Yeoja Bay were relatively low in comparison with those of Gamak Bay and the offshore area of Yeoja Bay. High abundance of dinoflagellate cysts can be explained by high dinoflagellate production, and since high nitrate loading into the water column enhances occurrence and number of dinoflagellate species, dinoflagellate productivity is directly influenced by nitrate concentration (Lopes et al. 2007). Nitrogen loading into Yeoja bay is related to the intrusion of freshwaters derived from rivers located in the northern part of the bay. DIN in the inner area of Yeoja Bay (Yb1) is relatively higher than that in the outer part and offshore area of Yeoja Bay (Yb2 and Yb3) (Table 1). According to Lee and Kim (2007), relative proportion of dinoflagellates to diatoms increases from the low salinity and high DIN inner part to the high salinity and lower DIN offshore area of Yeoja Bay. As a result, this may lead to a lower cyst concentration within Yeoja Bay. In addition, Yeoja Bay has a large seasonal variation in salinity due to the intrusion of freshwaters, and precipitation that falls mostly in the summer season (from July to September). These specific conditions may also determine the low dinoflagellate cyst concentration of this bay as few species such as Spiniferites spp. tolerate low-salinity environments (Pospelova et al. 2004).

The result of cluster analysis shows distributional characteristics of dinoflagellate cyst assemblages in study area: groups I (Yeoja Bay) and II (offshore area of Yeoja Bay and Gamak Bay). Based on the PCA results, we identified that the differences in salinity levels are the main environmental factors affecting distributional characteristics of dinoflagellate cyst assemblages (groups I and II) in studied areas. Group I is affected by the intrusion of freshwaters, and has relatively low salinity levels. The assemblages related to group I are characterized by abundant occurrences of S. nephroides and cyst of P. americanum. Holzwarth et al. (2007) suggested that S. nephroides and cyst of P. americanum can be used as a proxy for fluvial influences. However, S. nephroides has also been observed in marine environments (Ellegaard 2000; Persson et al. 2000). This species possibly seems to have a tolerance to salinity. Marret and Zonneveld (2003) and Kawamura (2004) reported that the occurrence of P. americanum cyst is not restricted to river estuaries or even to coastal environments. In study area the abundant occurrence of P. americanum cyst was also observed in the inner part of Gamak Bay, which is affected by urban sewage (Fig. 3). Consequently, the occurrences of this species in study area may be related to low salinity levels.

In general, the distribution of heterotrophic cysts is related to availability of prey such as diatoms and small flagellates (Naustvoll 2000). According to Hwang et al. (2005), the freshwater input into Yeoja Bay brings with it high silicate concentration (mean dissolved inorganic silicate 36 μM in the summer season), and in summer the main primary producers are diatoms (Shim 1980; Lee and Youn 2000). This probably provides a favorable environmental condition for the growth of heterotrophic dinoflagellates. Nevertheless, the relatively low proportions of heterotrophic cysts in Yeoja Bay suggest that heterotrophic dinoflagellates in Yeoja Bay may be controlled by other environmental conditions such as salinity rather than the availability of prey, because heterotrophic dinoflagellates that can endure low-salinity stress can graze on the preys such as diatoms (Shin et al. 2010b).

Group II is frequently affected by the intrusion of open seawaters, and is characterized by relatively low DIN in August. In particular, DIN in Gamak Bay (Gb1 and Gb2) exhibit a large differential gradient in between August and February, possibly due to nitrogen uptake by dinoflagellates in summer (Shim 1980) (Table 1). According to Kim (2002), the limiting nutrient responsible for the growth of dinoflagellates in Gamak Bay and the offshore area of Yeoja Bay is nitrogen, which suggests that the distribution of dinoflagellate cysts in group II may be closely related to DIN and that environmental conditions in group II are favorable for the growth of dinoflagellates compared with that of group I. The cyst assemblages related to group II were generally characterized by high abundance of heterotrophic dinoflagellate cysts such as Protoperidinium cysts (S. quanta and Brigantedinium spp.) and Polykrikos cysts. In particular, Brigantedinium spp. were highly abundant in this group. According to Holzwarth et al. (2007), high relative abundances of heterotrophic cysts are related to high chlorophyll a concentrations, which can be used as a productivity indicator (Zonneveld et al. 2001; Marret and Scourse 2002). PCA result showed that the inner part of Yeoja Bay (Yb1) and Gamak Bay (Gb1 and Gb2) has a positive relationship to chlorophyll a concentrations in August. As such, the abundance of heterotrophic cysts is likely to be used as productivity indicators within our study area. Brigantedinium spp. are also abundant in areas with nutrient-rich waters, such as upwelling areas (Marret 1994; Wall et al. 1977; Pospelova et al. 2008). According to Marret and Zonneveld (2003), their distribution is not restricted to any salinity or nitrate concentration ranges, whereas recently Shin et al. (2010b) reported that although the environmental conditions such as nutrients are favorable for the growth of dinoflagellates, low salinity levels in Yeoja Bay may hamper the growth. Consequently, their relatively higher concentration may be attributed to the intrusion of open seawaters characterized by high salinity.

For an accurate interpretation of the obtained data, it is essential to know the species-selective degradation of dinoflagellate cysts as a result of oxidation. According to Zonneveld (1997), cysts produced by Protoperidinium species, especially Brigantedinium spp., appear to be sensitive to decay in relation to oxygen availability, except for some species (e.g., Impagidinium species). This suggests that bottom oxygen concentration is an important factor that can affect dinoflagellate cyst assemblages (Zonneveld et al. 2007). In this study, Gamak Bay has been characterized by relatively low bottom oxygen concentrations (Kim et al. 2006), whereas bottom oxygen concentrations of Yeoja Bay and the offshore area of Yeoja Bay range from 2.73 to 6.09 ml/L (Shim 1980). According to Reichart and Brinkhuis (2003), the relative abundances of Protoperidinium species at oxygen concentrations of above 2.5 ml/L rapidly decreased. However, since sedimentation rates in our study area are very high, which indicates shorter exposure to bottom water conditions, good preservation of dinoflagellate cysts in sediment samples is expected. Consequently, we assume that there are no systematic errors in the interpretations of data based on dinoflagellate cyst assemblages from our study area.

Dinoflagellate Cysts as Indicator of Eutrophication in Gamak Bay

The utilization of dinoflagellate cysts as an indicator of eutrophication has been well documented (e.g., Dale et al. 1999; Matsuoka 1999; Matsuoka et al. 2003; Pospelova et al. 2002, 2005). For instance, total cyst concentration and the number of L. machaerophorum in Oslofjord increased with eutrophication (Dale et al. 1999), while high relative abundance of S. quanta with L. machaerophorum has been associated with eutrophication in estuaries of Massachusetts (Pospelova et al. 2002). Eutrophication in Tokyo Bay has also led to increased ratios of heterotrophic dinoflagellate cysts (Matsuoka 1999). Gamak Bay is an anthropogenically eutrophied area characterized by high nutrient levels. The abundance of L. machaerophorum in Gamak Bay ranged from 2 to 16 cysts g−1 and its relative abundance was only 1.9% of the total cyst assemblages. Pospelova et al. (2002) reported that L. machaerophorum occasionally tend to response negatively to eutrophication. However, vertical change in concentration of L. machaerophorum did not show a particular trend (Shin et al. 2010a). These occurrences are different from Oslofjord where relative abundances of L. machaerophorum varied from more than 5 to ca. 50% of the assemblages (Dale et al. 1999). In addition, the concentration of S. quanta cysts is not remarkable in Gamak Bay. In Korean coastal areas, L. polyedrum (the biological name of L. machaerophorum) is not abundant, and large bloom of this species has never been reported (Shim 1980; Kim et al. 1990; Lee et al. 1999). This suggests that L. machaerophorum do not respond to eutrophication in Gamak Bay, and as such cannot be used as an indicator species for eutrophication. More recently, Smayda and Reynolds (2003) formulated the ecological strategies of marine dinoflagellates, with emphasis on various bloom-forming species. According to their results, L. polyedrum is an upwelling relaxation taxa, and as such their cysts are reported as abundant in upwelling area (Sprangers et al. 2004). Upwelling and anthropogenic eutrophied areas generally have different mechanisms of nutrient enrichment. The major nutrients for the growth of dinoflagellates in Gamak Bay, especially nitrogen, are supplied by the inflow of sewage and bacterial decomposition of organic matters derived from aquaculture activities (Kim 2002; Noh et al. 2006). Consequently, although nutrients are satisfied, discharged organic matters do not allow the growth of L. polyedrum in Gamak Bay, possibly due to limited light penetration in the turbid waters.

Matsuoka (1999) suggested that an increase of heterotrophic dinoflagellate cysts is probably a good indicator for eutrophication based on the observation of dinoflagellate cyst assemblages in Tokyo Bay. Our results indicate that Gamak Bay is characterized by high abundance of heterotrophic dinoflagellate cysts, dominated by Polykrikos cysts and Brigantedinium spp. In general, increases of heterotrophic dinoflagellate cysts have been observed in both upwelling and eutrophied areas, characterized by high primary production and nutrient level (Marret 1994; Dale et al. 2002; Radi and de Vernal 2004). For this reason, a eutrotrophic indicator is difficult to distinguish from an upwelling indicator (Dale 2009). However, increase of Polykrikos cysts has been frequently observed in eutrophied areas such as Tokyo Bay, Japan (Matsuoka 1999; Matsuoka et al. 2003), and in Apponagansett Bay, Massachusetts, where nutrient enrichment has continued to increase since the beginning of the 20th century (Pospelova et al. 2002). Matsuoka et al. (2003) suggested that the high abundance of Polykrikos cysts observed from surface sediments in Tokyo Bay may reflect a hypertrophic nutrient condition developed from a nutrient rich condition. In addition, Shin et al. (2010a) reported that the changes in nutrient levels in Gamak Bay are coincided with the increase in the proportion of heterotrophic cysts. It is therefore possible that high abundance of heterotrophic dinoflagellate cysts dominated by cysts of Polykrikos cysts in Gamak Bay may have resulted from a response to anthropogenic eutrophication like Tokyo and Apponagansett Bays, although the relationship between Polykrikos cysts and eutrophication still requires verification.

According to Sætre et al. (1997), low concentration of dinoflagellate cysts dominated by heterotrophic dinoflagellate cysts is related to industrial pollution in Frierfjord, Norway. Cyst concentrations in Gamak Bay are low in comparison with those of the offshore area of Yeoja Bay, which is affected to a lesser extent by direct anthropogenic activities. This result is similar to observations from New Bedford Harbor, Massachusetts, USA (Pospelova et al. 2002, 2005) and Frierfjord, Norway, which is affected by both industrial pollution and eutrophication (Sætre et al. 1997). However, low cyst concentration dominated by high abundance of heterotrophic dinoflagellate cysts in Gamak Bay needs to be further investigated in relation to the effect of industrial pollution, as the concentrations of heavy metals in surface sediments are low (Kim 2002), although human activities such as aquaculture farms are permitted within the bay.

Conclusions

Although the relationship between dinoflagellate cysts and environmental conditions has been widely investigated in many coastal areas of the world, the knowledge-base in Korean coastal areas remains limited. Here, we present dinoflagellate cyst assemblages in relation to environmental conditions from southern coastal waters of Korea. Since the study area is affected by both open seawaters with relatively high salinity levels and fresh waters, the differences in salinity levels are related to the distribution of dinoflagellate cysts.

Dinoflagellate cysts as an indicator for eutrophication in Gamak Bay was not L. machaerophorum suggested in Oslo Fjord, however the proportions of heterotrophic cysts dominated by Polykrikos cysts reflected the eutrophicated condition in Gamak Bay. This indicates that dinoflagellate cysts as indicators of eutrophication may vary with environmental characteristics a given study area.