Population Dynamics of the ‘Golden Tides’ Seaweed, Sargassum horneri , on the Southwestern Coast of Korea: The Extent and Formation of Golden Tides

: Since 2015, troublesome masses of ﬂoating Sargassum horneri have been introduced via ocean currents and winds to the southwestern coastline of Korea, including Jeju Island. These massive mats have caused considerable damage to the aquaculture industry, tourism, and the marine ecosystem. Most previous studies of S. horneri have focused on cultivation, the development of gene markers, and photosynthetic activity, but few data on population dynamics are available. We investigated the population dynamics of native S. horneri o ﬀ the southwestern coast of Korea with the aim of predicting the formation of golden tides. Populations at two sites had obligate annual life cycles. Thalli were recruited during the period September–November, grew during the period December–April, and senesced by July. This pattern reﬂected seasonal trends in water temperature. Speciﬁc growth rates and heights of the thalli at Munseom were signiﬁcantly higher than those at Jindo. The greatest environmental di ﬀ erence between the two sites is probably the degree of exposure to wave action. Mortality density (thalli lost per unit area) in the Munseom population was highest during the period December–January (i.e., 2–3 months after recruitment) and in March. Most thalli in the Jindo population died o ﬀ in July when water temperatures increased. The maximum average biomass of S. horneri thalli detaching from the substrata reached 1.6 kg fresh weight m –2 during January and March. Thus, large-scale drifting mats were formed by S. horneri detachment from the substrata. Despite the di ﬀ erences in space and environment between China and Korea, our ﬁndings will enable quantitative assessments of the overall ﬂoating Sargassum biomass in the East China and Yellow Seas.


Introduction
Massive floating macroalgal blooms (so-called seaweed tides) of increasing frequency, magnitude, and severity have been reported from coastal regions around the world [1,2]. 'Green tides' formed by blooms of species of the genus Ulva, and 'golden tides' formed by blooms of Sargassum species This study should provide valuable information for a better understanding of the extent and formation of golden tides.

Study Area
This study was conducted at two sites, Jindo (34 • 22 37" N, 126 • 18 28" E) and Munseom (33 • 13 38" N, 126 • 34 04" E), which are on the southwestern coast of Korea ( Figure 1). The Jindo site was located in Supumhang Harbor; it is a sheltered area with a substratum comprising boulders and stones overlaid with muddy sand. The macroalgal community was dominated primarily by large brown algae (Sargassum horneri, S. thunbergii, S. fulvellum, and Undaria pinnatifida) and ephemeral green algae (Ulva spp.). In particular, S. horneri formed dense assemblages at 4-7 m water depth. In contrast, the Munseom site, located on Jeju Island, is very exposed to wave action originating in the East China Sea. S. horneri dominated water depths of 5-8 m at Munseom. Kim et al. provided a detailed description of macroalgal assemblages at Munseom [29]. The tidal cycle on the coast is semi-diurnal, and the system is classified as mesotidal, with a maximum tidal range during spring tides of about 3.0 m at Jindo and 2.7 m at Munseom (Tide Tables for the Coast

Study Area
This study was conducted at two sites, Jindo (34°22′37″ N, 126°18′28″ E) and Munseom (33°13′38″ N, 126°34′04″ E), which are on the southwestern coast of Korea ( Figure 1). The Jindo site was located in Supumhang Harbor; it is a sheltered area with a substratum comprising boulders and stones overlaid with muddy sand. The macroalgal community was dominated primarily by large brown algae (Sargassum horneri, S. thunbergii, S. fulvellum, and Undaria pinnatifida) and ephemeral green algae (Ulva spp.). In particular, S. horneri formed dense assemblages at 4-7 m water depth. In contrast, the Munseom site, located on Jeju Island, is very exposed to wave action originating in the East China Sea. S. horneri dominated water depths of 5-8 m at Munseom. Kim et al. provided a detailed description of macroalgal assemblages at Munseom [29]. The tidal cycle on the coast is semi-diurnal, and the system is classified as mesotidal, with a maximum tidal range during spring tides of about 3.0 m at Jindo and 2.7 m at Munseom (Tide Tables for the Coast

Sampling Design
Observations were made monthly at the two sites using Scuba diving in the period from November 2017 to October 2018. In November 2017, we established permanent quadrats at the two sites at water depths in which S. horneri formed dense assemblages. Quadrat sizes were selected according to population densities. Five 1.0 m 2 quadrats and four 0.25 m 2 quadrats were set at Jindo and Munseom, respectively. Quadrats were spaced at three-meter intervals. All thalli in each quadrat were tagged with surveyors' flagging tape and individually numbered for sequential observations on individuals. We marked 100 and 150 thalli at Jindo and Munseom, respectively.

Sampling Design
Observations were made monthly at the two sites using Scuba diving in the period from November 2017 to October 2018. In November 2017, we established permanent quadrats at the two sites at water depths in which S. horneri formed dense assemblages. Quadrat sizes were selected according to population densities. Five 1.0 m 2 quadrats and four 0.25 m 2 quadrats were set at Jindo and Munseom, respectively. Quadrats were spaced at three-meter intervals. All thalli in each quadrat were tagged Sustainability 2020, 12, 2903 4 of 15 with surveyors' flagging tape and individually numbered for sequential observations on individuals. We marked 100 and 150 thalli at Jindo and Munseom, respectively.

Data Collection
The height, density, mortality, biomass, and growth rates of individuals at each site were recorded monthly in the field from November 2017 to October 2018. Heights were measured as the lengths of the main shoots of undamaged thalli. The main shoot is typically the longest; however, lateral branches may be longer than the main axis when it has been severed by wave action [30]. When this was the case, we used the length of the longest lateral branch as a measure of thallus height. Thallus elongation was calculated using the following formula: where L 0 is the initial length (cm) of the main shoot axis, L t is the length (cm) of the main axis after t days, and t is the observation period (days). The growth rates of individuals were calculated using the following equation [31]: The density of thalli was estimated by counting all individuals in the quadrats. The density was converted to units of individuals per m 2 . Densities (individuals per m 2 ) and rates (%) of monthly mortality were estimated from the number of tagged individuals that disappeared each month. The total biomass of thalli per unit area (m 2 ) was estimated by summing the individual predicted weights of thalli in each quadrat. To estimate the individual weights, we collected 60 thalli with a range of main axis lengths monthly from each of the sites to determine the allometric relationship between thallus height (cm) and individual weight (g dry weight). The main axis lengths of all undamaged and damaged thalli in each of the quadrats were measured monthly and converted to dry weights using allometric relationship equations. Sargassum horneri usually grows attached to rocks, and it also continues to grow while free-floating. Unfortunately, we were unable to measure elongation in the floating population. Therefore, we assumed that rates of elongation were identical in attached and floating populations. The weights of detached thalli were estimated using the allometric relationship equations. The total detached biomass per unit area (m 2 ) was estimated monthly by summing the individual predicted weights of thalli that detached in each quadrat.
We measured water temperature at the two sites every 15 min using HOBO Pendant temperature/light data loggers (Onset Computer Corporation, Bourne, MA, USA). Salinity was measured on each sampling occasion using a YSI salinometer (YSI model 63, YSI incorporated, USA). To determine dissolved inorganic nutrient concentrations (NH 4 + , NO 3 − , NO 2 − , and PO 4 3− ) in the water column, we collected four replicate surface water samples each month through the experimental period. Water samples were placed on ice in the field and frozen until analyzed. Water column nutrient concentrations were determined using standard colorimetric techniques following the procedures of Parsons et al. [32]. The sum of NH 4 + -N, NO 3 − -N, and NO 2 − -N concentrations was recorded as total dissolved inorganic N (DIN). PO 4 3− concentrations were used as a measure of total dissolved inorganic phosphate (DIP).

Statistical Analyses
Significant differences in height, density, elongation, growth rate, biomass, detached biomass, mortality, mortality rate of thalli, and water column nutrients (DIN and DIP) among sampling times and between sites were identified by two-way ANOVA. A two-way ANOVA without replication was performed to test for significant differences in daily mean water temperature between the two study sites. Data were tested for normality and homogeneity of variance prior to analysis to determine whether they met the assumptions of parametric statistics. If these assumptions were not satisfied, data were Sustainability 2020, 12, 2903 5 of 15 square root transformed (height, density, biomass, detached biomass, mortality, and elongation) or arcsine transformed (mortality rate and growth rate). When significant differences were detected by ANOVA, we performed Student-Newman-Keuls (SNK) post hoc tests to identify significant pairwise differences between means. An alpha level of 0.05 was used for statistical tests. All statistical analyses were performed using IBM SPSS Statistics ver. 20.0 software (IBM Corporation, Armonk, NY, USA).

Environmental Factors
Water temperature at both sites had distinct seasonal patterns, ranging from 4.6 • C and 14.6 • C in February 2017 to 27.3 • C and 28.4 • C in August 2018 at Jindo and Munseom, respectively ( Figure 2A). Water temperature was significantly higher (p < 0.001) at Munseom than at Jindo. While water temperatures during fall, winter, and spring were higher at Munseom than at Jindo, there was no difference in water temperature between the two sites during late spring-summer. Notably, water temperatures during the late August-early September period decreased temporarily at Munseom following the passage of typhoon Soulik. Salinity was relatively constant throughout the study period, with averages of 31.9 and 30 Practical Salinity Unit (PSU) at Jindo and Munseom, respectively ( Figure 2B).
Water column DIN concentrations at both study sites varied significantly (p < 0.001) with sampling time ( Figure 2C). The DIN concentrations at Jindo were significantly higher (p < 0.001) than those at Munseom. The average DIN concentrations at Jindo and Munseom were 8.8 and 1.8 µM, respectively. Water column DIP concentrations at both sites were usually less than 1.0 µM during the study period ( Figure 2D).

Environmental Factors
Water temperature at both sites had distinct seasonal patterns, ranging from 4.6 °C and 14.6 °C in February 2017 to 27.3 °C and 28.4 °C in August 2018 at Jindo and Munseom, respectively ( Figure  2A). Water temperature was significantly higher (p < 0.001) at Munseom than at Jindo. While water temperatures during fall, winter, and spring were higher at Munseom than at Jindo, there was no difference in water temperature between the two sites during late spring-summer. Notably, water temperatures during the late August-early September period decreased temporarily at Munseom following the passage of typhoon Soulik. Salinity was relatively constant throughout the study period, with averages of 31.9 and 30 Practical Salinity Unit (PSU) at Jindo and Munseom, respectively ( Figure 2B).
Water column DIN concentrations at both study sites varied significantly (p < 0.001) with sampling time ( Figure 2C). The DIN concentrations at Jindo were significantly higher (p < 0.001) than those at Munseom. The average DIN concentrations at Jindo and Munseom were 8.8 and 1.8 μM, respectively. Water column DIP concentrations at both sites were usually less than 1.0 μM during the study period ( Figure 2D).

Density and Mortality
Sargassum horneri had a clear annual density cycle. Recruitment occurred in the fall ( Figure 3A). Thallus density at Munseom was significantly higher (p < 0.001) than that at Jindo ( Table 1). The species had the highest densities during late fall and early winter (November/December), 1-3 months after recruitment; thereafter, densities decreased. All S. horneri thalli had completely disappeared by

Density and Mortality
Sargassum horneri had a clear annual density cycle. Recruitment occurred in the fall ( Figure 3A). Thallus density at Munseom was significantly higher (p < 0.001) than that at Jindo ( Table 1). The species Sustainability 2020, 12, 2903 6 of 15 had the highest densities during late fall and early winter (November/December), 1-3 months after recruitment; thereafter, densities decreased. All S. horneri thalli had completely disappeared by July. The patterns of mortality density (i.e., thallus losses per unit area) and mortality rate differed markedly between the two sites ( Figure 3B,C). The mortality density at Munseom was significantly higher (p < 0.001) than that at Jindo due the differences in density between the two sites ( Table 1). The mortality rate at Munseom fluctuated greatly with time; it was highest during the periods December-January, March-April, and June-July. The mortality rate of the Jindo population gradually increased over time, with the highest values in April and the period June-July.
Sustainability 2020, 12, x FOR PEER REVIEW 6 of 15 July. The patterns of mortality density (i.e., thallus losses per unit area) and mortality rate differed markedly between the two sites ( Figure 3B,C). The mortality density at Munseom was significantly higher (p < 0.001) than that at Jindo due the differences in density between the two sites ( Table 1). The mortality rate at Munseom fluctuated greatly with time; it was highest during the periods December-January, March-April, and June-July. The mortality rate of the Jindo population gradually increased over time, with the highest values in April and the period June-July.

Height, Growth and Biomass
The seasonal fluctuations in thallus height differed greatly between sites ( Figure 4A). Thallus height at Munseom dramatically increased from November to April, when thalli reached a maximum length (857 cm). The heights at Jindo gradually increased over the study period, reaching a maximum of 137 cm in April. After May, the apices of all thalli were damaged or severed by wave action. Elongation varied significantly (p < 0.001) between sites ( Figure 4B; Table 1). During the growth period, the elongation rate at Munseom was significantly higher (p < 0.001) than that at Jindo ( Table 1). The highest elongation rate, 12.7 cm day -1 , occurred during February at Munseom. The elongation rate was very low at Jindo (<0.5 cm per day). The growth rate was also significantly higher (p < 0.001) at Munseom than at Jindo. The highest rate at Jindo occurred in January, and the lowest rate during the period March-April ( Figure 4C; Table 1).
Thallus biomass also varied significantly (p < 0.001) by season, and patterns differed between sites (p < 0.001) ( Figure 5A; Table 1). The biomass of the Munseom population increased during fall and winter and decreased during spring and early summer. The biomass of the Jindo population also increased during fall and early winter (September-December), was almost constant from December to April, and decreased between mid-spring and summer (April-July). The maximum population biomasses at Jindo and Munseom were 925 and 3100 g fresh weight m -2 , respectively. The biomass of detached thalli at Munseom varied significantly (p < 0.001) with sampling time but did not track a clear seasonal trend ( Figure 5B; Table 1). The detached biomass of the Jindo population increased gradually through the study period ( Figure 5B). The biomass of detached thalli was highest during January and March (average: 1600 g fresh weight m -2 ) at Munseom and lowest during April and July (average: 250 g fresh weight m -2 ) at Jindo. Thallus biomass also varied significantly (p < 0.001) by season, and patterns differed between sites (p < 0.001) ( Figure 5A; Table 1). The biomass of the Munseom population increased during fall and winter and decreased during spring and early summer. The biomass of the Jindo population also increased during fall and early winter (September-December), was almost constant from December biomasses at Jindo and Munseom were 925 and 3,100 g fresh weight m -2 , respectively. The biomass of detached thalli at Munseom varied significantly (p < 0.001) with sampling time but did not track a clear seasonal trend ( Figure 5B; Table 1). The detached biomass of the Jindo population increased gradually through the study period ( Figure 5B). The biomass of detached thalli was highest during January and March (average: 1600 g fresh weight m -2 ) at Munseom and lowest during April and July (average: 250 g fresh weight m -2 ) at Jindo.

Relationship between Water Temperature and Thallus Biomass
The biomass of the Jindo population was highest at a water temperature of ca. 5-12 °C between December and early April; thereafter, it decreased with increasing water temperature ( Figure 6A). Elongation peaked at temperatures of about 8-12 °C in January and during the period March-April, but slowed during May when water temperatures exceeded 16 °C ( Figure 6C). Thallus biomass at Munseom was highest at temperatures ca. 14-16 °C between January and early April, and decreased thereafter with increasing water temperatures ( Figure 6B). The peak elongation in the Munseom

Relationship between Water Temperature and Thallus Biomass
The biomass of the Jindo population was highest at a water temperature of ca. 5-12 • C between December and early April; thereafter, it decreased with increasing water temperature ( Figure 6A). Elongation peaked at temperatures of about 8-12 • C in January and during the period March-April, but slowed during May when water temperatures exceeded 16 • C ( Figure 6C). Thallus biomass at Munseom was highest at temperatures ca. 14-16 • C between January and early April, and decreased thereafter with increasing water temperatures ( Figure 6B). The peak elongation in the Munseom population occurred at a temperature of about 15 • C in February, but it fell when temperatures exceeded 16 • C in May ( Figure 6D).

Population Dynamics of S. horneri Populations
The present study showed that two S. horneri populations on the southwestern coast of Korea had obligate annual life cycles. The demographic parameters of both populations exhibited strong seasonal variation: Recruiting during September-November, growing and maturing during November-April, and disappearing during July. These patterns reflected seasonal changes in water temperatures, which have conspicuous impacts on the growth of S. horneri. Seasonal trends in S. horneri populations (increasing during fall and spring and decreasing during spring and early summer) have been reported for the northwest Pacific coasts [30,33,34]. Water temperature is a vital driver of seasonal variation in Sargassum spp. biomasses within the seaweed beds at Ningaloo [35]. We found that S. horneri at both study sites had maximum biomass and elongation at temperatures in the range 5-15 °C during winter and spring, although two sites had different seasonal trends in water temperature. This finding was similar to those of previous studies on persistent S. horneri beds along the coast of Japan; the optimal temperature for S. horneri growth in Japan ranged between 10 and 15 °C [30,33,36]. In our study, the growth rate of the two populations was highest at water temperatures within the 5-15 °C range during the winter-spring period, similar to the findings of Yu et al., who showed that the specific growth rates for length and weight in S. horneri increased when water temperature rose from 5 °C to 20 °C [37]. Thus, we showed that water temperature can be considered a major factor controlling the growth and maturation of Sargassum spp. [33,38].
The two populations we studied exhibited obviously different trends in height, density, biomass, and mortality. The height, density, biomass, elongation, and growth rate at Munseom were 5-10 times higher than those at Jindo. Nutrient availability influences the growth, productivity, and community structure of seaweeds [39,40]. The growth patterns of kelp species are strongly affected

Population Dynamics of S. horneri Populations
The present study showed that two S. horneri populations on the southwestern coast of Korea had obligate annual life cycles. The demographic parameters of both populations exhibited strong seasonal variation: Recruiting during September-November, growing and maturing during November-April, and disappearing during July. These patterns reflected seasonal changes in water temperatures, which have conspicuous impacts on the growth of S. horneri. Seasonal trends in S. horneri populations (increasing during fall and spring and decreasing during spring and early summer) have been reported for the northwest Pacific coasts [30,33,34]. Water temperature is a vital driver of seasonal variation in Sargassum spp. biomasses within the seaweed beds at Ningaloo [35]. We found that S. horneri at both study sites had maximum biomass and elongation at temperatures in the range 5-15 • C during winter and spring, although two sites had different seasonal trends in water temperature. This finding was similar to those of previous studies on persistent S. horneri beds along the coast of Japan; the optimal temperature for S. horneri growth in Japan ranged between 10 and 15 • C [30,33,36]. In our study, the growth rate of the two populations was highest at water temperatures within the 5-15 • C range during the winter-spring period, similar to the findings of Yu et al., who showed that the specific growth rates for length and weight in S. horneri increased when water temperature rose from 5 • C to 20 • C [37]. Thus, we showed that water temperature can be considered a major factor controlling the growth and maturation of Sargassum spp. [33,38].
The two populations we studied exhibited obviously different trends in height, density, biomass, and mortality. The height, density, biomass, elongation, and growth rate at Munseom were 5-10 times higher than those at Jindo. Nutrient availability influences the growth, productivity, and community structure of seaweeds [39,40]. The growth patterns of kelp species are strongly affected by local nutrient availability [41]. Nitrogen enrichment enhances the specific growth rates of length and weight in S. horneri and promotes high biomass accumulation of drifting thalli [37]. The water column DIN and DIP concentrations at Jindo were significantly higher than those at Munseom. It is likely that the ambient nutrient level at Jindo provided adequate nutrients for S. horneri growth. Nevertheless, the Jindo population had lower height, density, biomass, elongation, and growth rate than the Jeju population. Therefore, the difference in growth between the two populations cannot be attributed to differences in nutrient availability.
The variability in the growth rate and morphology of S. horneri has been linked to water temperature [36]. Warming promotes higher Sargassum growth rates by increasing rates of metabolism and photosynthetic activity before the thalli decay [42]. During winter, S. horneri thalli exposed to high water temperatures have increased growth rates and lengths, and earlier maturation than those in cooler waters [30]. The average water temperature at Munseom (16.2 • C) was higher than at Jindo (8.9 • C) during the growing season (December-April). After January, the Munseom thalli grew more rapidly and reached larger sizes than the Jindo thalli. Thalli with receptacles were found in Munseom during early March; receptacles were first observed on Jindo thalli in mid-April (S.R. Park, personal observation).
Wave action is also considered an important factor in the growth of large brown algae when there are minimal differences in other environmental factors between sites, such as light, temperature, and nutrients [43]. Wave action has an impact on the growth and productivity of macroalgae [44,45]. The morphology and survival of seaweeds is related to water flows created by waves or currents [18,46]. Thus, wave activity or wave-induced water motion is considered an important controlling factor in algal community structure and composition [47,48]. Munseom is located on the southern coast of Jeju Island, where it is directly open to the East China Sea. Thus, the Munseom population was likely exposed to higher wave activity than the Jindo population. The degree of exposure to wave action may account for the large differences in morphology, density, biomass, and mortality that we observed between the two populations. Therefore, wave exposure and warmer water temperatures during the winter season were probably key drivers of S. horneri growth.

Contribution of Mortality to the Formation of Golden Tides
Since the early years of the current century, massive drifting mats of S. horneri have been frequently observed along the East China Sea coast. They have aggressively spread along the length of the southwestern coast of Korea (including Jeju Island) and the western coast of Japan [11,[49][50][51]. Floating S. horneri rafts originated from attached thalli in the subtidal zones of islands off Zhejiang Province, China, and were transported northeastwards with oceanic currents and wind [5,11,13,14]. The floating S. horneri biomass therefore depends on the biomass of the attached source populations. The biomass of attached benthic populations of S. horneri off Gouqi Island, China, where the massive drifting S. horneri rafts in the East China Sea originated, ranged from 0.54 kg fresh weight m -2 in August to 4.42 kg fresh weight m -2 in June [52]. The maximum biomass of attached S. horneri in Obama Bay, Japan, averaged 555 g dry weight m -2 (ca. 2.8 kg fresh weight m -2 , using a dry weight: fresh weight ratio of 0.2) [33]. In contrast, non-native populations of S. horneri in Todos Santos Bay, Mexico, had very low biomass, ranging from 3 g dry weight m -2 (ca. 15 g fresh weight m -2 ) to 78 g dry weight m -2 (ca. 390 g fresh weight m -2 ) [53]. S. horneri biomass at Munseom and Jindo ranged from 0.6 kg fresh weight m -2 to 3.1 kg fresh weight m -2 during the growing season (December-May); this range was similar to those measured in S. horneri populations in China and Japan [33,52]. The maximum biomasses of S. horneri fragments detached from attached thalli at Munseom were 1.5 ± 0.4 and 1.7 ± 0.5 kg fresh weight m -2 in January and March, respectively, when mortality was high. The floating S. horneri biomass in the East China Sea was ca. 2 kg fresh weight m -2 . The estimated floating biomass of Sargassum in the Florida straits and the Gulf of Mexico ranged from 1.26 kg fresh weight m -2 to 6.74 kg fresh weight m -2 , averaging 3.34 kg fresh weight m -2 [10,20]. Given that detached S. horneri thalli continue to grow during the floating period [21], our estimates of floating biomass are similar to those reported for the East China Sea, the Florida Straits, and the Gulf of Mexico [10,20]. Our results provide a quantitative assessment of the overall floating Sargassum biomass. Similar to those around the islands off Zhejiang Province, the S. horneri population at Munseom was exposed to high wave action and had high mortality rates and densities during the December-January period and March, suggesting that large drifting seaweed rafts may be produced by detaching thalli. When high mortality occurs around the islands off Zhejiang Province and floating S. horneri mats form in the Yellow Sea and East China Sea, drifting S. horneri rafts can be observed in the southern part of Jeju Island, Korea, within approximately a month [51]. Large mats of drifing S. horneri were introduced to the Korean coasts by the Kuroshio Current and Taiwan Warm Current in the late January-February and April-early May periods of 2015-2019. These rafts accumulated along the southern coast of Korea, including Jeju Island [14,54]. High mortality rates and/or densities likely also occurred among the islands off Zhejiang Province during the December-January period and March. If so, this would strongly support our second postulate, namely that the biomass of drifting S. horneri is strongly associated with mortality in attached S. horneri populations. Unfortunately, there are few data for the S. horneri populations among the islands off Zhejiang Province, China, which severely limits more detailed interpretation of the extent and formation of golden tides in the East China Sea. Therefore, further studies of the populations among the islands off Zhejiang Province are required to better understand the extent and formation processes of drifting S. horneri mats.

Conclusions
Recently, massive S. horneri mats have resulted in devastating impacts on the local economy and coastal ecosystems in Korea. A dangerous combination of wave and tide conditions can exacerbate economic losses due to Sargassum beaching events. Thus, predicting the timing and extent of golden tides will provide early-warning of Sargassum beaching. The results of this study confirm that S. horneri had a clear annual life cycle and exhibited seasonal variations in density, which reflected seasonal changes in water temperature. Furthermore, the difference in height, density, growth, and mortality between the two sites may be affected by the degree of exposure to wave action. High mortality density at the study site was observed during the December-January period and in March. The detached thalli had an effect on the extent and formation of floating S. horneri mats. This result showed that the biomass of S. horneri thalli detached from substrata was approximately 1.6 kg fresh weight m -2 during high mortality periods. This will help us to estimate the extent and formation timing of golden tides. However, most seaweeds show inter-annual variation in their biomass, and environmental conditions were temporally not constant. Thus, long-term monitoring, with more study sites, is needed to fully predict the extent and formation timing of drifting S. horneri mats. writing-original draft preparation, S.K.C., S.K., and S.R.P.; writing-review and editing, S.K.C., S.K., and S.R.P. All authors have read and agreed to the published version of the manuscript.
Funding: This research was a part of the project titled 'Development of the collecting treatment system and resources technology for Sagrassum horneri' funded by the Ministry of Oceans and Fisheries, Korea. This work was also supported by a grant from the National Institute of Fisheries Science (R2020045).