Introduction
Coastal ecosystems are affected by changes in environmental conditions (both natural and anthropogenic), and their ongoing continuous evolution is well documented. Unusual natural phenomena, such as typhoons/hurricanes or extreme temperatures, can have severe effects on coastal ecosystems (e.g., UN Atlas of the Oceans 2005: www.ocean satlas.org/results-search/en/?querystring=2005). In particular, a tsunami caused by an earthquake, submarine landslide, or other disturbance could cause an unusual set of conditions. The tsunami that followed the Great East Japan Earthquake in 2011 had a notable effect on the environment, mainly on coastal topography and ecosystems in coastal wetlands, on the shallow sea floor, and on the shoreline (Ecosystem Sub-Working Group 2011;Hara and Higuchi 2013). The tsunami that followed the Indian Ocean Sumatran Earthquake in 2004 reportedly affected phytoplankton populations recorded by IRS P4-Ocean Colour Monito data (Tang et al. 2009) and benthic ecosystems which showed large declines in abundance and diversity (Whanpetch et al. 2010). After the Great East Japan Earth Quake, several studies related with dinoflagellate cyst ecology were performed by Nishitani et al. (2012), Kamiyama et al. (2014), Natsuike et al. (2014), Ishikawa et al. (2015), and Matsuoka et al. (2018). These studies focused the unusual big blooms of Alexandrium tamarense (Lebour) Balech accompanied with paralytic shellfish poisoning (PSP) along the Sanriku Coast of East Japan.
Intensive A. tamarense blooms were first observed in Osaka Bay in 2002. This dinoflagellate species had not previously formed such dense blooms in Osaka Bay or caused any paralytic shellfish poisoning (PSP) incidents in the area (Yamamoto 2004). Several studies of the horizontal distributions of ellipsoidal dinoflagellate cysts identical to A. tamarense in Osaka Bay have been undertaken recently, and it was found that these ellipsoidal cysts are abundant in surface sediments (Yamamoto et al. 2009, 2011;Matsuoka and Ishii 2018). A high cyst density may play an important role in the formation of blooms by acting as a seed population. A. tamarense blooms now occur almost every year in the inner part of Osaka Bay and off its east coast. However, the accumulation of large numbers of A. tamarense cysts in coastal ecosystems can pose other important risks. For example, the coastal ecosystem and fishing industry of the Sanriku Coasts were badly affected by strong A. tamarense blooms caused by the germination of large numbers of cysts after the Great East Japan Earthquake (Ishikawa et al. 2015;Matsuoka et al. 2018). It has been predicted that a huge earthquake (referred to as the Tonankai Earthquake) will affect the Pacific coast of Western Japan in the near future. This earthquake and the tsunami likely to follow it will cause serious damage to the environment and to society as a whole (Osaka Prefecture 2013).
Here, we investigate the possibility of the occurrence of harmful algal blooms after the Tonankai Earthquake using data on the spatiotemporal distribution of ellipsoidal Alexandrium cysts in surface sediments in Osaka Bay. We also investigate the possibility of similar effects on the south and east coasts of the Korean Peninsula, where ellipsoidal Alexandrium cysts are also preserved in surface sediments.
Materials and Methods
Three 10 cm long core samples, labelled OS 2, OS 19, and OS 23, were collected in 2017. The sampling sites are marked on Figure 1. Core OS 23 was collected from the inner part of Osaka Bay, and core OS 19 was collected from the southeast coast of the bay. The samples were collected using a KK-type gravity corer (Kimata et al. 1960). Core OS 2 was collected from the bed at the mouth of the Yodo-Gawa River. The sediment in each core was composed of medium to fine silt. The OS 19 and OS 23 cores were cut into 1 cm slices, and each slice was stored in a black plastic bottle to exclude light. Only the top and bottom 1 cm sections of the OS 2 core were analyzed due to the particular location of the sample. The samples were processed using a method described by Matsuoka and Fukuyo (2000). Each sample was treated with HCl and HF at room temperature to remove calcium carbonate and silicate particles, respectively. Each sample was then passed through 125 and 20 μm stainless steel sieves. The material that passed through the 125 μm sieve but retained by the 20 μm sieve was analyzed. Dinoflagellate cysts (including ellipsoidal Alexandrium cysts) in each sample were counted by examination using an optical microscope with a magnification of 400-600. Dinoflagellate cysts were identified following the procedure described by Matsuoka and Fukuyo (2000) and Matsuoka and Ishi (2017) (Table 1).
In recent studies, new species names and combinations have been proposed for the species previously known as A. tamarense, Alexandrium fundyensei Balech, and Alexandrium catenella (Whedon and Kofoid) Balech ( John et al. 2014;Fraga et al. 2015). However, use of the new species name A. catenella, instead of the old name A. tamarense and use of the new species name Alexandrium pacificum R.W. Litaker instead of the old name A. catenella, causes confusion when attempting to interpret data published before 2014. We therefore use the old species names in this study.
Results
Vertical distributions of dinoflagellate cysts in the three cores samples
Dinoflagellate cysts of >45 species in 21 genera were found in core OS 19. The cysts belonged to five genera and nine species in the Gonyaulacales, three genera and four species in the Gymnodiniales, and 13 genera and 39 species in the Peridiniales. The dinoflagellate cyst densities ranged from 1,109 cysts g-1 at 5-6 cm deep to 2,016 cysts g-1 at 8- 9 cm deep, and the mean cyst density was 1,563 cysts g-1. (Fig. 2, Table 2). Heterotrophic dinoflagellate cysts predominated, and contributed >70% of the cysts in each sample. The heterotrophic dinoflagellate cyst contribution increased from the bottom to the top of the sample. As shown in Figures 2 and 3, ellipsoidal Alexandrium cysts were most abundant at a depth of 1-2 cm, where the cyst density was 935 cysts g-1, and least abundant at a depth of 9-10 cm, where the cyst density was 11 cysts g-1 (Fig. 2).
Dinoflagellate cysts from >40 species in 19 genera were identified in core OS 23. These cysts belonged to five genera and nine species in the Gonyaulacales, three genera and four species in the Gymnodiniales, and 13 genera and 39 species in the Peridiniales. The dinoflagellate cyst densities ranged from 1,684 cysts g-1 at 7-8 cm to 3,375 cysts g-1 at 6-7 cm, and the mean cyst density was 2,395 cysts g-1. Heterotrophic dinoflagellate cysts were extremely abundant, and contributed >65% of the cysts throughout the core (Fig. 2, Table 2). Ellipsoidal Alexandrium cysts were most abundant at 0-1 cm, where the cyst density was 616 cysts g-1, and least abundant at 4-5 cm, where the cyst density was 20 cysts g-1 (Fig. 2).
Dinoflagellate cysts from 12 genera and 20 species were abundant in the top 1 cm of core OS 2, but few ellipsoidal Alexandrium cysts, Brigantedinium simplex (Wall) Reid, and Votadinium calvum Reid were found (Table 2). Ellipsoidal Alexandrium cysts were most abundant at the top of the core, where the cyst density was 1,080 cysts g-1, and least abundant at the bottom of the core, where the cyst density was 190 cysts g-1.
Horizontal distribution of dinoflagellate cysts in surface sediment in Osaka Bay
The results of our study and Matsuoka and Ishii (2018) are combined in Figure 5. The dinoflagellate cyst densities were higher in the northern part of Osaka Bay than elsewhere. Ellipsoidal Alexandrium cysts were abundant in the northern part of Osaka Bay, at the mouth of the Yodo-Gawa River, and along the east coast of Osaka Bay (Fig. 4). The ellipsoidal Alexandrium cyst density was highest (3,610 cysts g-1) at Station 7 in Matsuoka and Ishii (2018) and lowest (4 cysts g-1) at Station 3 in the same work. Ellipsoidal Alexandrium cysts were abundant in the inner part of Osaka Bay, but were also found in the southern part of the bay. Other dinoflagellate cysts were less abundant in the southern than in the northern part of Osaka Bay.
Discussion
Concentrations of ellipsoidal Alexandrium cysts in sediment
Hamano et al. (2002) stated that the first recorded PSP events caused by bivalves Ruditapes philippinarum (Adams & Reeve 1850), Mytilus galloprovincialis Lamarck, 1819, and Crassostrea gigas (Thunberg 1793) occurred in Osaka Bay on 2002. Yamamoto (2004) reported that PSP in Osaka Bay could be caused by the dinoflagellates Gymnodinium catenatum H.W. Graham, A. tamarense, A. catenella, and Alexandrium tamiyavanichii Balech. The year 2007 saw the first red tide (discoloration) in Osaka Bay caused by A. tamarense (Yamamoto et al. 2009).
A PSP-causing dinoflagellate species found around Harina- Nada in the Seto Inland Sea and in Osaka Bay has been identified as Protogonyaulax catenella (Whedon & Kofoid) Taylor (=A. catenella) from the results of cyst germination experiments and plankton observations (Fukuyo 1983). Interestingly, no A. tamarense was detected at that time. Yamaguchi et al. (1996) investigated the distributions and abundances of A. tamarense and A. catenella cysts in surface sediment in the eastern Seto Inland Sea, including Osaka Bay, and found high cyst densities between the inner part and middle of the east coast of Osaka Bay. Yamamoto (2004) and Nagai et al. (2007) suggested that the cyst-producing dinoflagellate species might be A. tamarense because the cyst distribution found by Yamaguchi et al. (1986) was similar to those found after the A. tamarense blooms in 2002.
Seven PSP events and red tides caused by A. tamarense have been reported since 2002 (Yamamoto 2018). The ellipsoidal cysts with colourless walls found were identical to the cysts of Alexandrium acathenella (Whedon & Kofoid) Balech, A. tamarense, or A. catenella. However, these cysts could not be differentiated from their external morphologies, so Yamamoto et al. (2011) performed cyst incubation experiments and concluded that most of the ellipsoidal cysts preserved in the surface sediment were identical to A. tamarense cysts. A. acathenella was found extremely rarely, and only small amounts of A. catenella have ever been found in Osaka Bay. The ellipsoidal cysts related to PSP events and red tide were therefore concluded to be A. tamarense/catenella (mainly A. tamarense) cysts.
Ellipsoidal cysts identical to A. catenella/tamarense cysts were found in all three cores, but the cyst densities varied. The cyst densities were lowest in the lower parts of the cores, and increased from the middle parts of the cores in an upward direction. The highest cyst densities were found at 0- 1 cm in core OS 23 and 1-2 cm in core OS 19. A similar trend was found for core OS 2. This suggests that the production of A. catenella/tamarense ellipsoidal cysts has increased strongly in recent years.
The A. tamarense cyst densities in surface sediments around the world have been investigated, as shown in Table 3 in Matsuoka et al. (2018). The maximum cyst densities (915 cysts g-1 in core OS 19 and 616 cysts g-1 in core OS 23) were not high compared with those found in other areas. Matsuoka and Ishii (2018) found an ellipsoidal Alexandrium cyst density of 3,610 cysts g-1 at station OS 5 in Osaka Bay, which was ~3 km southwest of core OS 23. Yamamoto et al. (2009) found a maximum cyst density of 5,683±631 cysts g-1 in wet sediment on the east coast of Osaka Bay (near core OS 19) in 2007 after an A. tamarense bloom. The highest cyst density recorded after that was 90,672±11,269 cysts g-1 in 2017 (unpublished data). Mizushima and Matsuoka (2004) found that empty A. tamarense and A. catenella cysts representing after germination are not preserved well compared with the empty cysts of other dinoflagellates, because A. tamarense and A. catenella cysts are extremely weak and have thin autophragm. More resting cysts than were counted were therefore probably produced and deposited in the surface sediment. These data suggest that ellipsoidal Alexandrium cysts are widely distributed at high densities in the inner to southeastern shallow parts of Osaka Bay (Fig. 4B).
rosion, resuspension, and redeposition of surface sediments containing ellipsoidal Alexandrium cysts as caused by a tsunami
Case study of Osaka Bay
The high densities of ellipsoidal Alexandrium cysts preserved in surface sediments appear to indicate a high potential risk of PSP incidents in future. A. tamarense blooms have occurred regularly along the Osaka Bay coast ever since the first bloom was recorded (Yamamoto 2004, 2018). This means that cysts preserved in surface sediments may play a role in initiating new blooms. The risks posed by a tsunami after the expected Tonankai Earthquake should also be considered. According to the Osaka Prefecture (2013), a tsunami 3.2-5.1 m high will reach the Osaka Bay coast after the expected Tonankai Earthquake off central Japan. The tsunami is expected to be 5.1 m high in the innermost part of Osaka Bay (Fig. 5). The Showa Nankai Earthquake of 1946 caused a tsunami that was ~3 m high when it reached the inner part of Osaka Bay. Unfortunately, no detailed data on sediment disturbance in Osaka Bay were recorded at the time. However, we can predict the effects of a large tsunami on a shallow bay by examining data recorded after the Great East Japan Earthquake. After the tsunami that followed the Great East Japan Earthquake, A. tamarense blooms were observed at Sendai Bay (Kamiyama et al. 2014), Kesennuma Bay (Nishitani et al. 2012;Ishikawa et al. 2015), Ofunato Bay (Kaga et al. 2012;Matsuoka et al. 2018), and Funka Bay (Natsuike et al. 2014). These blooms were initially caused by the germination of living cysts in sediment suspended by the tsunami. In Ofunato Bay, the maximum A. tamarense cell density reached 100,600 cells L-1 on 26 May 2011 and 179,500-676,500 cells L-1 in early June 2011 (Kaga et al. 2012). These extremely dense blooms appeared to be related to the environmental conditions. The tsunami on 11 March 2011 eroded the Ofunato Bay sea floor by more than 25 cm, and sediment containing A. tamarense cysts was resuspended and redeposited on the bottom of the sea. Ellipsoidal Alexandrium cysts are more abundant in the inner part of Osaka Bay, where the water is <20 m deep, than in other parts of the bay. This means that the sea floor could easily be eroded by a tsunami. The disruption caused by a tsunami will cause A. tamarense cysts to reach unusually high concentrations near the sea bed. A water temperature of around 10°C is suitable for the germination of A. tamarense cysts, so living cysts that resettle on the sea floor will start to germinate when the temperature reaches ~10°C, and dense blooms toxic to shellfish will form. A further environmental factor was also important to A. tamarense blooms in Ofunato Bay. Zooplankton that are predators of A. tamarense and other phytoplankton are scarce in Ofunato Bay (Yamada 2012). Strong A. tamarense blooms were therefore caused by resuspension and redeposition of ellipsoidal Alexandrium cysts and sediments by the tsunami, a water temperature of ~10°C (because it was early spring), and the absence of predation by zooplankton.
A high ellipsoidal Alexandrium cyst density in sediment can therefore act as a seed population for subsequent blooms and also cause extremely large blooms after a tsunami caused by an earthquake.
Case study of the Korean Peninsula
Ellipsoidal cysts identical to A. tamarense and/or A. catenella cysts are also found in surface sediments along the southern coast of the Korean Peninsula. Ellipsoidal Alexandrium cysts were first found in Jinhae Bay by Kim (1994) and Shin et al. (2017, 2018) and have since been found at several locations, including Masan Bay (Kim et al. 2002), Gwangyang Bay (Kim et al. 2003, 2009), Gamak Bay (Shin et al. 2008, 2010), and Yeoja Bay (Shin et al. 2009, 2010). However, the Alexandrium cyst densities in these locations were <200 cysts g-1, which is low compared with the cyst densities found in Osaka Bay. The east coast of the Korean Peninsula has been affected by tsunamis several times (Cho 2018a). An earthquake called the Middle Japan Sea (Nihonkai- Chubu) Earthquake (magnitude 7.7) off Akita Prefecture in North Japan in 1983 caused a tsunami that reached the east coast of the Korean Peninsula and was 3.9 m high at Imwon, 1.56 m high at Sokucho, and 0.62 m high at Pohang ( Japan Marine Science and Technology Center 1997;Cho 2018a, b) (Fig. 5). It can be seen that the Korean Peninsula experiences tsunamis of considerable height, and may be affected by tsunamis from either the Sea of Japan (East Sea) or from near Tsushima Island. Tsunami risk for the area around Tsushima Island provided by Nagasaki Prefecture is shown in Figure 5. The highest risk of a tsunami around Tsushima Island is from the possibility of an earthquake caused by the active submarine Nishiyama Fault and its northward extensions. Such an earthquake could cause a tsunami that could be 5 m high at the Tushima coast (Nagasaki Prefecture 2016). It can be seen that strong tsunamis could affect the east and south coasts of the Korean Peninsula. Such a tsunami would resuspend and redeposit sediment containing ellipsoidal Alexandrium cysts, and the cysts may then cause strong blooms and PSP. Such a scenario should be kept in mind and plans made to minimize damage that may be caused.