A total of 61 surveys were conducted at the intake area of HYHNPP in the eastern part of LDB from May to September of 2019 and 2020. Five sampling stations were set up at the intake area ( Figure 1 ), of which the central station was set up at the inlet area of the NPCS (HYH03), two stations (HYH01 and HYH02) were located at the north-eastern side of the central station, and two stations (HYH04 and HYH05) were set up at the south-western side of the central station. The distance between two stations was approximately 2 km. Large jellyfish species were sampled using a plane anchor drift net (mesh size: 10 mm, length: 110 m, width: 15 m; Figure 2 ). The direction of net casting was perpendicular to the direction of the flow, and the nets were hauled with the current. The soak time usually lasted from 30 minutes to 1 hour and was adjusted according to the size of the catch. As individual N. nomurai and A. coerulea jellyfish are highly fragmented and difficult to count accurately at high density, the relative biomass (RB) was used for expressing the abundance of jellyfish in this study. The crane of the fishing vessel was used for lifting large numbers of jellyfish, which were weighed using a large hanging hook scale. Small numbers of jellyfish and individuals were placed into sample bags and weighed using a small hanging hook scale. The precision of wet weight weighed by the large and small scale was 1 kg and 0.1 g, respectively. The RB of the jellyfish was expressed in kg net^–1 h^–1. The numbers, bell diameters (BDs), and wet weights (WWs) of individual jellyfish were noted; the BD was measured with a straight edge. The surface environmental parameters, including the sea surface temperature (SST), dissolved oxygen (DO), surface salinity (SS), and pH, were measured in situ using a YSI ProQuatro water quality meter. Ephyrae and juveniles were collected using a shallow-water type II plankton net (mesh size: 160 μm, diameter: 31.6 cm) and the ephyrae collected from Jiangjunshi port, located 22 km from HYHNPP, were investigated ( Figure 1 ).

The normality and homogeneity of the variance were confirmed using the one-sample Kolmogorov–Smirnov test and Levene’s test, respectively. The Kruskal–Wallis H-test was performed when the data did not approach normality or homogeneity of variance. Multiple parametric comparisons were performed as the data were abnormal or exhibited non-homogeneous variance. Simple correlation analyses were performed for evaluating the statistical correlation between the RB and environmental factors (SST, SS, DO, and pH) using Spearman’s correlation with SPSS version 16.0. The graphs were prepared with Microsoft 2016 and R version 4.2.2.

In this study, the first ephyrae of A. coerulea appeared at Jiangjunshi port on 15 May 2019 and were collected for further analyses. However, N. nomurai ephyrae were absent during the entire period of the survey. The first juvenile medusae of N. nomurai appeared on 31 May 2019, with a BD of 4–10 cm. The medusae grew rapidly thereafter and reached a maximum BD of 76 cm at the beginning of September. Medusae of A. coerulea were first collected on 8 July, and these had BDs of 12–21 cm and had reached the size of adult organisms. The size of the medusae did not alter much thereafter ( Figure 3 ). The time of appearance of A. coerulea ephyrae in 2020 was the same as that in 2019, and no ephyrae of N. nomurai were detected in 2020. The first juvenile medusae of N. nomurai appeared on 11 June with a BD of 4–6 cm, which was comparable to the BD observed in 2019; however, the medusae appeared at a later period in 2020. There was a difficulty in measuring the BD of the subsequent samples, as the majority of samples were fragments. The results of available survey data revealed that the BD had increased rapidly since the first appearance of juvenile medusa. In this study, the BDs reached a maximum of 130 cm in mid-August 2020, which was significantly greater than the observations in 2019. Compared with 2019, the medusae of A. coerulea appeared earlier the following year, on 11 June 2020. The BDs of A. coerulea medusae collected in 2020 were 3–10 cm. The BDs increased rapidly and were similar to the BDs observed in the same period in 2019 ( Figure 3 ).

The RB of the two large jellyfish species, N. nomurai and A. coerulea, are depicted in Figure 4 . N. nomurai and A. coerulea populations alternated at the intake area of the NPCS, and the RB was not high during the period when they appeared simultaneously. In 2019, large numbers of N. nomurai began appearing in late June, and the average relative biomass (ARB) peaked in early July. The RB was highest on 2 July, when it reached a value of 2,806.00 kg net^–1 h^–1, and decreased rapidly thereafter. The ARB of A. coerulea increased subsequently and A. coerulea gradually became the dominant species from mid-July to mid-August. In 2020, large numbers of large jellyfish started appearing in July, and A. coerulea remained the dominant jellyfish species during the whole of July. The abundance of A. coerulea was highest on 7 July, when the ARB reached 2,234.40 kg net^–1 h^–1. A. coerulea decreased rapidly thereafter and N. nomurai gradually dominated from August, during which the RB remained continually low. The abundance of N. nomurai was highest on 22 August, when the RB reached 412.50 kg net^–1 h^–1. There were no significant patterns in the abundance of the large jellyfish species across the five stations. Overall, the RB of both species remained consistently lower at HYH04 than at the other stations ( Figure 5 ).

The relationship between the BD of the individual jellyfish species (BD_a for A. coerulea, BD_n for N. nomurai) and the WW (W_a for A. coerulea, W_n for N. nomurai) could be described using the following equations ( Figure 6 ): W a = 0.1774 B D a 2.5232 and W n = 0.0916 B D n 2.8265 .

The fluctuations in the SST, DO level, SS, and pH at the intake area of the NPCS between 2019 and 2020 are depicted in Figure 7 . The SST increased from May to early August, gradually stabilized from August, and decreased from September. The annual SST in 2020 was generally lower than that during the same period in 2019. In contrast, the DO levels exhibited a decreasing trend from May to early August, gradually stabilized from August, and decreased from September. Overall, the annual DO level in 2020 was higher than that during the same period in 2019. In 2019, the SS was relatively stable until August, decreased significantly from the beginning of August, and was subsequently stabilized. In 2020, the SS decreased slightly from late August and increased gradually thereafter. The pH fluctuated and remained relatively stable in 2019 and 2020, with a range of 8.34–8.45. The fluctuations in the DO level, SS, and pH exhibited a similar trend across the five stations; however, the SST fluctuated significantly more at HYH04 than at the other stations and was generally higher than at the other stations ( Figure 8 ).

The survival, SST, DO, SS, and pH range of the two large jellyfish species were similar during the survey period. A. coerulea was distributed in waters with SSTs of 16.7–32.2°C, DO levels of 4.37–7.71 mg/l, SS of 31.37–32.82‰, and pH of 8.12–8.50, while N. nomurai was distributed in waters with SSTs of 15.6–32.2°C, DO levels of 4.48–7.71 mg/l, SS of 31.27–32.82‰, and pH of 8.12–8.50 ( Figure 9 ). The RB of these two jellyfish species was similar, and greater than 60 kg net^–1 h^–1. The RB of A. coerulea was relatively high in waters with SSTs of 20.9–30.9°C, DO levels of 4.37–6.69 mg/l, SS of 31.38–32.61‰, and pH of 8.26–8.50. The RB of N. nomurai was relatively high in waters with SSTs of 20.6–32.2°C, DO levels of 4.48–6.69 mg/l, SS of 31.27–32.79‰, and pH of 8.26–8.50 ( Figure 9 ).

The results of simple correlation analyses demonstrated a highly significant negative correlation between the RB of A. coerulea and the RB of N. nomurai ( Figure 9 ) (r = –0.609, p< 0.01). There was no significant correlation between the RB of A. coerulea and the environmental parameters (SST, DO, SS, and pH); however, the RB of N. nomurai had a highly significant positive correlation with the SST and a highly significant negative correlation with the DO level (p< 0.01).

Determining the sources of jellyfish is essential for deciding appropriate measures for the prevention and control of jellyfish disasters at an early stage; it is currently one of most effective approaches for addressing this concern. A pure waterjet and a scraper have been applied to remove polyps after determining the sources in Korean waters, which proved to be effective (Yoon et al., 2018); however, the sources of large jellyfish at the intake area of the HYHNPP remain to be clearly determined to date.

It has been demonstrated that artificial structures such as ports provide additional habitats for the asexual stage of jellyfish (Feng et al., 2017). Ephyrae of A. coerulea were observed at Jiangjunshi port for two consecutive years in this study but were not detected at the intake area. The discovery of ephyrae at Jiangjunshi port indicated that the ephyrae had been released at the region, from where they might have gradually migrated outward to regions around the intake area from mid-May. This finding provides evidence regarding one of the habitats of A. coerulea. Moreover, the phenomenon observed at the intake area also coincides with the findings of other studies on the time of appearance of different life history stages of A. coerulea (Dong et al., 2008; Dong et al., 2014; Wang and Sun, 2015; Feng et al., 2018). A previous study indicated that the duration between the release of ephyrae and the development of small medusae of A. aurita could be less than 1 month at 18°C (Båmstedt et al., 2001). The newly liberated ephyrae of A. aurita developed into young medusae in only 20–28 days in the innermost part of Tokyo Bay (Ishii et al., 2004). In this study, juvenile medusae of A. coerulea were collected from the intake area approximately 1.5 and 1 months after the appearance of ephyrae in 2019 and 2020, respectively. Therefore, the findings possibly indicate a potential source of A. coerulea around the intake area. However, it is necessary to assess the association between the released ephyrae and the medusae collected at the intake area by continuous monitoring, and to determine whether there are more sources of A. coerulea.

Notably, N. nomurai ephyrae did not appear at the nearby ports or intake area during the two years of the survey, and the presence of N. nomurai ephyrae in this region has not been reported in similar studies (Wang et al., 2013; Dong et al., 2018). These findings indicate that N. nomurai did not originate in this region. It has been demonstrated that the northern estuarine area of LDB is the main habitat of N. nomurai in the study area. The ephyrae were possibly produced in mid-spring, and the rapid growth period lasted from late spring to early summer. The juveniles migrate toward the central and southern waters during development, reach maturity, and prepare to spawn in mid-autumn, and the populations decrease rapidly in late autumn (Dong et al., 2018). The intake area was located in the central region of LDB, and the overall timing of each stage was delayed by half a month in the central region, compared with that in the northern area of LDB. It was speculated that, following their release, the ephyrae and some juveniles migrated from the northern region of the LDB and continued to grow during their southward migration, and reached the intake area in early summer, which corresponded to the period of delay discovering juveniles at the study area. The RB increased rapidly in this area during the period of continuous migration, and the individuals grew rapidly, with the BDs reaching more than 50 cm in early July. This situation possibly continued until late autumn, following which the RB declined rapidly owing to the lack of source replenishment and the commencement of fishing activities in early September. The results obtained herein support the findings of previous studies which suggested that the estuarine area is possibly one of the sources of N. nomurai in the intake area; however, direct evidence is necessary for confirming the conjecture.

Jellyfish feed heavily on plankton, fish eggs, and juveniles, and populations require large quantities of food for supporting the rapid growth period (Greve, 1994; Hansson et al., 2005). Food availability can become a pivotal variable in limiting the viability and fecundity of jellyfish populations when the quantity of food decreases following a rapid increase in the density of large jellyfish species (Goldstein and Steiner 2020; Kitajima et al., 2020). A partial similarity in food composition induced competition between N. nomurai and A. coerulea (Wang et al., 2021). The study revealed that the RB of large jellyfish increased exponentially at the intake area from late June or early July and reached a plateau thereafter, and finally declined from September. The alternating occurrence and population distribution of these two species with similar niches could indicate passive adaptation and balance under the extremely high biomass and density in a restricted space with limited food availability.

Food composition not only controlled the population size, but also affected the spatial distribution of the two species. The main food for large jellyfish is plankton, which is also the primary species affected by thermal effluents (Bamber and Seaby, 2004; Li et al., 2014). The large influx of thermal effluents in NPPs can alter the hydrodynamic conditions and community structure of local seas (Salgueiro et al., 2015; Lee et al., 2018). In this study, station HYH04 was located near the outlet of the thermal effluents and was characterized by high water velocity and relatively drastic changes in temperature. The SST measured at HYH04 was sometimes higher by 2–3°C or more than that at the other stations during 2019 and 2020, and sometimes there were no differences between the SST at HYH04 and that at other stations. The jellyfish exhibited some, but extremely limited, swimming abilities and migrated actively or passively away from the local seas owing to conditions of strong flow expansion and loss of primary food sources. Comparison of the cumulative RB (CRB) and ARB calculated per hour among different stations revealed that the CRB and ARB were lowest at HYH04, for both A. coerulea and N. nomurai ( Table 1 ). The findings could be attributed to the characteristics of the local environmental where the station was located.

The interannual patterns and bloom dynamics of N. nomurai were linked to processes of the regional climate (Lee et al., 2021). Temperature is an important factor in influencing the growth and development of jellyfish; temperature affects the production of ephyrae during the asexual stage and the growth of medusae during the sexual stage (Baumsteiger et al., 2018; Feng et al., 2020). Optimal temperature is beneficial to the growth of juveniles, and previous studies have demonstrated a significant positive correlation between the RB and SST in many waters (Baumsteiger et al., 2018; Dong et al., 2018). In this study, the results of correlation analyses revealed a significant positive correlation between the RB of N. nomurai and the SST. The RB and BDs of N. nomurai increased rapidly from mid- to late June. The SST decreased slowly after August, while the RB of N. nomurai remained relatively stable and gradually decreased thereafter. The rapid warming in late spring and early summer could lead to the rapid growth and development of jellyfish.

In this study, the results of correlation analyses revealed a significant negative correlation between the DO level and the RB of N. nomurai, which was consistent with the findings of previous studies (Riyas et al., 2021). The DO level decreased sharply from May to early August, which corresponded with the rapid growth and increased RB, and coincided with a rapid increase in the daily interception and capture of jellyfish. This could be attributed to the increased consumption of DO caused by the rapid increase in RB within a short period of time. The phenomenon could also be attributed to artificial jellyfish control measures at the intake area of the NPCS. The presence of large numbers of fragmented and dead jellyfish could affect the local biogeochemical cycle and rapidly reduce the DO levels owing to the high oxygen consumption by microorganisms during decomposition (Pitt et al., 2009; Condon et al., 2011). The practice of fragmenting and discarding jellyfish pieces into the sea water to avoid the risk of blockages inevitably increases the mass of jellyfish fragments and corpses, which in turn accelerates the consumption of DO. As jellyfish have a low metabolism (Rutherford and Thuesen, 2005), artificial jellyfish control measures could play a more significant role in reducing the DO level.

Salinity is another important factor that affects the RB and distribution of jellyfish, and the effect of salinity was more pronounced in the asexual stage. The ephyrae and juveniles of N. nomurai exhibited a preference for low-salinity areas such as estuaries (Yoon et al., 2008; Dong et al., 2018). In this present study, salinity varied in a narrow range, and no significant correlation between SS and RB was detected. Therefore, it is unclear whether distributions of these two species at sexual stages are constrained by salinity. The pH has a limited effect on the RB and distribution of jellyfish, and jellyfish statoliths become significantly smaller at lower pH values (Winans and Purcell, 2010). A reduction in pH affects the pulsing behavior and size of A. coerulea ephyrae (Tills et al., 2016); however, it is generally accepted that acidification is not significantly associated with the abundance of medusae (Richardson and Gibbons, 2008). In this study, the pH values fluctuated between 8.26 and 8.50 during the period of investigation, and there was no significant correlation between the pH and RB of both species of jellyfish.

First, the analysis of environmental factors is an effective strategy for predicting jellyfish blooms (Baumsteiger et al., 2018; Dong et al., 2018; Riyas et al., 2021). Further analyses of metagenic life cycles and real-time online monitoring of environmental indicators would aid in predicting and designing effective measures against the aggregation of different jellyfish species for ensuring the operational safety of NPCSs. Second, ecological theories can provide a theoretical basis for the prevention and control of disasters caused by large jellyfish species. The release of competing species without disturbing the balance of the ecosystem, such as economic fishes, could serve as a suitable strategy for suppressing jellyfish blooms and providing additional economic value. Third, it is necessary to conduct studies on jellyfish blooms in regions beyond the local area of NPCSs. Further efforts are necessary for identifying the sources of jellyfish and understanding their migration patterns for increasing the efficacy of various preventive and control measures. Fourth, the treatment of jellyfish should be enhanced for avoiding negative effects on marine environments (Pitt et al., 2009; Condon et al., 2011), as these would in turn increase the possibility of other related disasters. Fifth, the increased use of unofficial sources or other joint monitoring approaches would increase access to information, provide further evidence for future studies, and aid in ensuring the safety and security of NPCSs (Gutiérrez-Estrada et al., 2021).