Littoral Water in Hong Kong as a Potential Transient Habitat for Juveniles of a Temperate Deepwater Gnomefish, Scombrops boops (Acropomatiformes: Scombropidae)

© 2021 Academia Sinica, Taiwan Open Access Littoral Water in Hong Kong as a Potential Transient Habitat for Juveniles of a Temperate Deepwater Gnomefish, Scombrops boops (Acropomatiformes: Scombropidae) Jiehong Wei1, Jiarui Gu2, Min Liu3, Bai-an Lin3, Gabriel Y. Lee1, Tak-Cheung Wai1,2, Paul K.S. Lam1,2,4 , Meng Yan1,2,5*, and Priscilla T.Y. Leung1,2,5* 1State Key Laboratory of Marine Pollution, City University of Hong Kong, Hong Kong, China. *Correspondence: E-mail: [email protected] (Leung); [email protected] (Yan). Tel: 852-34429438. Fax: 852-34420524. E-mail: [email protected] (Wei); [email protected] (Gu); [email protected] (Lee); [email protected] (Wai); [email protected] (Lam) 2Research Centre for the Oceans and Human Health, City University of Hong Kong Shenzhen Research Institute, Shenzhen, China 3State Key Laboratory of Marine Environmental Science, College of Ocean and Earth Sciences, Xiamen University, Xiamen, China. E-mail: [email protected] (Liu); [email protected] (Lin) 4Department of Chemistry, City University of Hong Kong, Kowloon, Hong Kong, China 5Southern Marine Science and Engineering Guangdong Laboratory (Zhuhai), Zhuhai, China Received 29 December 2020 / Accepted 6 April 2021 / Published 2 July 2021 Communicated by Felipe Ottoni A total of 40 juveniles belonging to a temperate deepwater gnomefish species, Scombrops boops, were sampled from littoral habitats (2–5 m depth) of eastern Hong Kong waters in April and May 2017 and March 2019. The presence of gnomefish juveniles in subtropical southern China is reported for the first time at a record low latitude of 22°11'–22°21'N. The specimens were identified based on the COI gene sequence. The genetic composition between Japan and Hong Kong gnomefish populations were compared by sequencing the mitochondrial Cytb gene, which showed no genetic differentiation. The juveniles ranged from 3.5–10.1 cm (n = 40) in total length, with 35 individuals caught from Sargassum beds and five from rocky reefs. Our findings highlighted that the littoral habitats in Hong Kong waters, in particular the seasonal Sargassum beds, are important for small juveniles of S. boops. Key words: Genetic homogeneity, Northwest Pacific, Nursery habitat, Sargassum beds, Subtropical. BACKGROUND The family Scombropidae used to belong to the order Perciformes (Nelson et al. 2016), but a recent restructuring moved it to a new order, Acropomatiformes (Ghedotti et al. 2018). Scombrops Temminck & Schlegel, 1845 is the only genus in the family, and three members have been reported in the northwestern (NW) Pacific (Oyama et al. 2019), including Scombrops boops (Houttuyn, 1782), Scombrops gilberti (Jordan & Snyder, 1901) and an undescribed species Scombrops sp. (Mochizuki et al. 2017). Scombrops boops is generally a temperate benthopelagic species distributed from the coastal area of Hokkaido (Japan) to the East China Sea (China) at which the southernmost existing record was located (22°49'N, as shown in Fig. 1a) (Shao 1987; Hayashi 2002; GBIF 2020; Froese and Pauly 2020). A couple of studies have also reported that S. gilberti and Scombrops sp. are closely related to S. boops genetically, and are Citation: Wei J, Gu J, Liu M, Lin B, Lee GY, Wai TC, Lam PKS, Yan M, Leung PTY. 2021. Littoral water in Hong Kong as a potential transient habitat for juveniles of a temperate deepwater gnomefish, Scombrops boops (Acropomatiformes: Scombropidae). Zool Stud 60:33. doi:10.6620/ ZS.2021.60-33. Zoological Studies 60:33 (2021) doi:10.6620/ZS.2021.60-33 1 © 2021 Academia Sinica, Taiwan found in the deep water off the Izu Peninsula to the Izu Islands and around the Ryukyu Islands in Japan, respectively (Itoi et al. 2010 2018; Oyama et al. 2019). As a commercially important species in Japan, the distribution, ecology and population genetics of S. boops have been extensively studied (Itoi et al. 2010). The species produces pelagic eggs in the coastal water of Japan during its spawning season from October to March (Mochizuki 1977). The eggs hatch after three days under controlled conditions in captivity (Yamada 1995). Juveniles of S. boops settle and grow in coastal water habitats of Japan and start migrating to the rocky bottom of the continental shelf break (200–700 m at depth) after reaching sexual maturation at ~38 cm Fig. 1. (a) A map indicating winter ocean currents in Northwest Pacific and geographical location of Hong Kong and sampling sites in Noguchi et al. (2012). HK: Hong Kong; IT: Iwate; KC: Kochi; KS: Kagoshima; NG: Niigata; NS: Nagasaki; SO: Shizuoka; TT: Tottori. (b) The six sampling sites in Hong Kong waters and number of Scombrops boops specimens from each site. BFI: Beaufort Island (22°11'N, 114°15'E); BI: Bluff Island (22°19'N, 114°21'E); NFW: Nam Fung Wan (22°21'N, 114°21'E); NP: Ninepins North (22°16'N, 114°21'E); TC: Turtle Cove (22°14'N, 114°13'E); TPC: Tung Ping Chau (22°33'N, 114°25'N). Sites with Sargassum habitat indicated by yellow circle, and with rocky reef indicated by purple triangle. Coastal Current Kuroshio Current Tsushima Current Oyashio Current China NFW (18) TPC (2) NP (11) BI (1) BFI (4) TC (4) 2019 2017 Kowloon Hong Kong Island New Territories 22°21’N HK NS KS TT KC SO NG IT (b) (a) 500 km 5 km 22°49′N Okinawa Trough Yellow Sea Warm Current N page 2 of 11Zoological Studies 60:33 (2021) © 2021 Academia Sinica, Taiwan in total length (TL) (Mochizuki 1977 1997; Hayashi 2002; Takai et al. 2014). Scombrops boops around the Japanese Archipelago may constitute a single population based on a genetic study. The two dominant northward coastal currents, i.e., Kuroshio Current and the Tsushima Current, may transport the planktonic larvae from the south and facilitate the mixing of geographic populations around the Japanese Archipelago (Noguchi et al. 2012). The occurrence of S. boops juveniles beyond Japanese waters is however largely unknown. This study presents the first records of S. boops in Hong Kong waters, which is also the new southernmost occurrence of this species in the NW Pacific region, with only juvenile specimens. This is also the first report of juvenile S. boops occurring in the littoral habitats of the South China Sea region. The potential nursery role of the coastal habitats, particularly the seasonal Sargassum beds, to this deepwater species was also discussed. MATERIALS AND METHODS Sampling and measurement A total of 40 specimens, initially morphologically identified as Scombrops species, were received during the dry-wet transitional season, i.e., April and May 2017 and March 2019, in the eastern waters of Hong Kong based on seasonal surveys. Thirty-six specimens were obtained from five locations in 2017; 31 (~86%) were collected from Sargassum beds (i.e., TC, NP and NFW) and five (~14%) were collected from rocky habitat without any Sargassum spp. (i.e., BI and BFI) (Fig. 1b). Four specimens were collected in 2019 from two Sargassum locations (i.e., NFW and TPC) (Fig. 1b). The fishing gear used was in-shore purse seining (4–6 mm mesh size; top length: 15 m, bottom length: 9.8 m × net depth: 2.5 m) operated by a chartered P4 fishing boat within littoral habitats. The specimens were kept at -20°C in the State Key Laboratory of Marine Pollution (SKLMP) for measurement and genetic analysis. Measurements of fish body length (TL; standard length, SL) and body weight were taken to the nearest 0.1 cm and 0.1 g, respectively. DNA extraction and polymerase chain reaction (PCR) Genomic DNA of the skeletal muscle from each specimen was extracted using Chelex® 100 Resin following the manual’s instruction (Bio-Rad Laboratories, Inc., United States). Polymerase chain reaction (PCR) was conducted in a total volume of 25 μl, containing 20 ng DNA template, 1 unit of Ex Taq DNA polymerase (Takara), 0.5 µM reverse and forward primers, 0.2 mM mixed dNTPs, 1.5 mM MgCl2, and 1 × PCR buffer (pH 8.4, 200 mM Tris-HCl and 500 mM KCl). Partial sequences of cytochrome c oxidase subunit I (COI) gene were amplified using Fish F1 (forward 5’-TCAACCAACCACAAAGA CATTGGCAC-3’) and Fish R1 (reverse: 5’-TAGA CTTCTGGGTGGCCAAAGAATCA-3’) (Ward et al. 2005). Amplifications of cytochrome b (Cytb) sequences were performed with L14369tuna-Glu (forward: 5’-ACCACCGTTGTTATTCAACTA-3’) and R-DloopY2 (reverse: 5’-CATTAACTTATGCAAGCGT C-3’) (Noguchi et al. 2012). PCR cycling conditions for Cytb and COI are as follow: 95°C for 2 min; 39 cycles of 95°C for 30 s, 51°C for 30 s, and 72°C for 45 s; 72°C for 2 min. PCR product was purified and sequenced by Tech Dragon Ltd. (Hong Kong) with the Sanger sequencing method using an ABI PRISM® 3730xl DNA Analyzer. For each PCR product, sequencings were performed using the same corresponding forward and reverse primers used for PCR reactions. Genetic analyses Raw sequences (sequenced from both forward and reverse primers) were assembled using Geneious v9.0.2 (Biomatters Ltd.). Assembled sequences were aligned with datasets from Japanese studies (COI, Oyama et al. 2019; Cytb, Noguchi et al. 2012) using the MAFFT v7.388 (Katoh and Stanley 2013) plugin of the Geneious program with default settings. Both ends of the aligned sequences for each DNA region were trimmed to obtain a uniform length for subsequent analyses (i.e., 607 bp for COI and 1114 bp for Cytb). Haplotypes of Cytb were identified using DnaSP v6 (Rozas et al. 2017). Unique haplotypes from Hong Kong have been deposited in the National Center for Biotechnology Information (NCBI) Genbank database. Fasta files of COI sequences and Cytb haplotypes are accessible in DataSet S1. Information on Cytb haplotypes was summarized in DataSet S2. HKY+G (Hasegawa et al. 1985) and K80 (Kimura 1980) were selected as the best fit model for COI and Cytb using jModelTest v2.1.10 (Darriba et al. 2012) based on Bayesian Information Criterion (Schwarz 1978). To identify the collected specimens, a Bayesian inference (BI) tree and a maximum likelihood (ML) tree were reconstructed in MrBayes v3.2 (Ronquist et al. 2012) and MEGA7 (Kumar et al. 2016), respectively, using COI sequences of all Hong Kong specimens and scombropid species data from Oyama et al. (2019). Closely related outgroups for BI and ML tree reconstructions were selected based on Oyama page 3 of 11Zoological Studies 60:33 (2021) © 2021 Academia Sinica, Taiwan et al. (2019) (i.e., Epigonus telescopes, accession number: KJ09756; Epigonus denticulatus, JF493426 and AP017435; Epigonus pandionis, KT883637). The tree is rooted by Doederleinia berycoides (accession number: AP009181) selected from Mochizuki et al. (2017). For the BI tree, two independent Markov chain Monte Carlo runs were performed with four chains for 500,000 generations, sampling every 100 generations and discarding the first 25% samples as burn-in. Sufficient convergence of the runs was estimated by summary statistics implemented in MrBayes v3.2 (Effective sampling size > 200, potential scale reduction factors→1). For the ML tree, bootstrap values of nodes were estimated by 1000 bootstrap replicates in MEGA7. The geneology of the of Cytb haplotypes was reconstructed using Popart v1.7 (http://popart.otago. ac.nz) with the TCS algorithm (Clement et al. 2002). Genetic differentiations between the Japanese and Hong Kong populations were estimated using pairwise FST values calculated using the best fit model available in Arlequin v3.5 with 1,000 permutations (Excoffier and Lischer 2010). RESULTS Identification and life stage of the collected specimens The external morphology of the specimens was examined. Pored lateral line scales ranged from 51–58, upper transverse scales 7–9, lower transverse scales 11–12, outer upper gill rakers 5–7, and outer lower gill rakers 13–15 (n = 12). Body was dorsally golden brown and ventrally silvery white. The second dorsal and caudal fins were yellow with white to brown edge (Fig. 2). The mean body lengths were 4.3 ± 1.0 cm SL (mean ± SD) and 5.2 ± 1.2 cm TL, and the mean body weight Fig. 2. Photographs of juveniles of Scombrops boops of different sizes. The upper and middle specimens were 4.6 cm and 7.3 cm total length (TL) collected from the Sargassum beds; the lower specimen was 10.1 cm TL collected from a rocky reef. page 4 of 11Zoological Studies 60:33 (2021) © 2021 Academia Sinica, Taiwan was 1.7 ± 2.0 g (n = 40) (Table 1). The numbers of lower transverse scales and outer lower gill rakers were within the ranges for S. boops juveniles of 8.1–10.3 cm SL (Itoi et al. 2008). The COI sequence dataset for the BI and ML tree reconstructions was 607 bp long, including 149 variable sites and 89 parsimony informative sites. COI sequences obtained in this study are available on GenBank (accession numbers: MK987135–MK987174). COI sequences of specimens from Hong Kong were 99.5–100% identical to S. boops, 98.4–98.7% to S. gilberti, 98.0–98.5% to an undescribed Scombrops sp., and 95.6–96.0% to S. cf. dubius (African Scombrops specimens previously misidentified as Scombrops boops, Oyama et al. 2019). The BI tree indicated four clades within the Scombropidae (Fig. 3), i.e., S. boops, S. gilberti, the undescribed Scombrops sp. and S. cf. dubius. All Hong Kong specimens were nested within the S. boops clade supported by a posterior probability of 0.80 and a bootstrap value of 69 (Fig. 3). Based on the morphological and molecular assessments, the 40 specimens sampled were therefore identified as S. boops juveniles. Similar molecular identification approach has been used to identify fish at early life stages successfully into species (Chu et al. 2019). Genetic comparison between populations from Hong Kong and Japan A trimmed Cytb dataset (1,114 bp) for TCS network was generated from a combined sequence data from this study and Noguchi et al. (2012); it contains a total of 115 haplotypes with 74 variable sites, 48 of which were parsimony informative. Since Sb72 and Sb73 from Kochi (Japan) (Noguchi et al. 2012) were identical after trimming for the combined dataset, the two sequences were regarded as one single haplotype and re-named Cytb72 in our analyses (a specimen from Hong Kong was also identified as Cytb72). Among the 115 haplotypes identified, 15 (i.e., Cytb101– Cytb115) were unique to Hong Kong, 25 were shared by Hong Kong and Japan, and the remaining 75 were unique to Japan (DataSet S2). Unique haplotypes from Hong Kong were deposited to GenBank under accession numbers MK987120–MK987134. The most common haplotype, Cytb2, was observed in eight specimens from Hong Kong and 31 specimens from Japan, which constitutes 20% and 17% of the Hong Kong and Japanese specimens, respectively. Neither haplotypes from Hong Kong nor Japan formed a single monophyletic clade in the TCS network (Fig. 4). Pairwise FST values of the sampling site pairs ranged from -0.02413 to 0.04120. Pairwise FST values between most of the site pairs were non-significant (P > 0.05), while significant values were only reported in pairs between Kochi and Niigata (Japan), and Kochi and Hong Kong (Table 2). DISCUSSION Habitat use of the collected juvenile Scombrops boops Based on the sexual maturation size (~38 cm TL) reported in S. boops (Mochizuki 1977 1997; Hayashi 2002), all specimens collected in this study are at their early juvenile stage. Macroalgal habitats are believed to be important nursery ground for S. boops. A relatively high proportion of the specimens was sampled from Sargassum beds (canopy-forming species) (Fig. 1b). At NFW, juveniles of S. boops were sampled in the same location of the Sargassum bed that occurred in both 2017 and 2019 surveys. Juveniles of S. boops with 3–12 cm TL were reported to use kelp forest as a seasonal residence in Japan from February to June (Kono et al. 2018). Juveniles of large predatory species, e.g., billfishes, tunas, swordfishes and dolphinfishes, were commonly found in association with Sargassum beds (Hoffmayer et al. 2005). A recent review suggested that tropical macroalgal habitats, including Sargassum habitat, could provide a key middle step in the triphasic life cycle of certain commercially important predatory fishes such as Table 1. Body length and body weight of Scombrops boops juveniles (n = 40) collected in Hong Kong waters. All the data are presented as Mean ± SD (Range) Number of specimens Total length (cm) Standard length (cm) Body weight (g) Sargassum bed 35 5.0 ± 0.9 (3.5–7.3) 4.1 ± 0.8 (2.9–6.1) 1.3 ± 0.8 (0.6–3.6) Rocky reef 5 6.6 ± 2.0 (5.0–10.1) 5.5 ± 1.6 (4.6–8.3) 4.2 ± 5.1 (1.6–13.3) Total 40 5.2 ± 1.2 (3.5–10.1) 4.3 ± 1.0 (2.9–8.3) 1.7 ± 2.0 (0.6–13.3) page 5 of 11Zoological Studies 60:33 (2021) © 2021 Academia Sinica, Taiwan Fig. 3. A COI Bayesian inference tree indicating the identify of the juvenile S. boops specimens from Hong Kong. Individuals from Hong Kong are annotated with green circles. Species of each clade are annotated based on Oyama et al. (2019). Posterior probability and bootstrap value of each clade are shown on the nodes. 0.008 LC388039 HK23 (MK987154) LC388036 Doederleinia berycoides (AP009181) HK19 (MK987150) HK42 (MK987173) HK15 (MK987146) LC388061 HK40 (MK987171) HQ945916 LC388038 LC388047 HK22 (MK987153) LC388057 HK27 (MK987158) LC388037 HK30 (MK987161) HK10 (MK987141) LC388035 LC388043 LC388050 HK43 (MK987174) LC388065 LC388059 HK36 (MK987167) LC388045 LC388060 LC388067 HK39 (MK987170) HK24 (MK987155) HK37 (MK987168) HK11 (MK987142) LC388066 HK04 (MK987138) LC388048 HK26 (MK987157) LC388046 HK35 (MK987166) HK34 (MK987165) LC388041 HK03 (MK987137) LC388054 HK25 (MK987156) HK29 (MK987160) LC388058 HK01 (MK987135) LC388052 HK28 (MK987159) HK41 (MK987172) LC388055 HK09 (MK987140) HK20 (MK987151) HK32 (MK987163) LC388070 HK33 (MK987164) HK14 (MK987145) LC388049 LC388042 HK05 (MK987139) LC388044 HK38 (MK987169) JF494461 HK18 (MK987149) LC388056 LC388063 LC388064 LC388062 LC388051 HK31 (MK987162) HK02 (MK987136) LC388053 LC388040 LC388068 HK13 (MK987144) HK16 (MK987147) LC388069 HK21 (MK987152) HK12 (MK987143) LC388071 HK17 (MK987148) 0.91/87 1/99 1/98 0.53/66 0.59/74 0.80/69 0.73/46 0.92/63 0.62/28 0.82/58 0.84/63 1/100 1/89 1/97 1/98 0.75/68 Epigonus pandionis (KT883637) Epigonus telescopus (KJ09756) Epigonus denticulatus (JF493426) Epigonus denticulatus (AP017435) Scombrops boops Scombrops gilberti Undescribed Scombrops sp. Scombrops cf. dubius page 6 of 11 Zoological Studies 60:33 (2021) © 2021 Academia Sinica, Taiwan species from Cephalopholis Bloch & Schneider, 1801 and Epinephelus Bloch, 1793 (Serranidae), Lethrinus Cuvier, 1829 (Lethrinidae) and Lutjanus Bloch, 1790 (Lutjanidae) (Fulton et al. 2020). Sargassum spp. have been designated as an essential fish habitat (EFH) due to its importance as fish and invertebrate nursery habitats (NOAA 1996). The majority (80%) of seaweed-associated fishes are at their Fig. 4. A haplotype network based on Cytb gene sequences showing the relationships among haplotypes identified in Scombrops boops specimens from Hong Kong (i.e., this study) and Japan (i.e., Noguchi et al. 2012). Green circles refer to haplotypes unique to Hong Kong, red circles refer to haplotypes unique to Japan, and purple circles refer to haplotypes shared by Hong Kong and Japan. 5 _30 _ a 2 8 19 93 22 96 64 38 106 107 112 113 110 80 75 39 81 59 85 27 63 86 52 10 99 44 108 111 104 89 102 97 68 16 41 60 34 92 57 71 74 80 88 83 90 62 69 66 58 82 103 100 115 25 53 55 78 109 79 65 48 26 45 37 35 21 36 12 72 18 1 24 47 77 34 7 42 95 14 73 49 94 46 87 11 13 5 9 114 101 105 84 56 51 50 20 54 31 29 76 91 67 17 23 28 70 33 98 61 15 6 45 32 43 Japan Share haplotypes Hong Kong 1 10 40 Sampling size Table 2. Estimated pairwise FST (below the diagonal) and p value (above the diagonal) of eight locations: seven from Japan (Noguchi et al. 2012) and one from Hong Kong (this study). Significant (p < 0.05) FST values are indicated by * NG TT NS KS KC SO IT HK Niigata (NG) - 0.47656 0.76855 0.92383 0.04590 0.80371 0.62500 0.29785 Tottori (TT) -0.00174 - 0.92773 0.58496 0.16406 0.63672 0.54199 0.53516 Nagasaki (NS) -0.00887 -0.01130 - 0.96680 0.18945 0.99121 0.29199 0.61133 Kagoshima (KS) -0.01603 -0.00379 -0.01412 - 0.06250 0.96387 0.11035 0.16406 Kochi (KC) 0.03516* 0.01339 0.01086 0.02074 - 0.05176 0.11133 0.01465 Shizuoka (SO) -0.00935 -0.00448 -0.01528 -0.01274 0.02200 - 0.44043 0.14355 Iwate (IT) -0.00578 -0.00302 0.00395 0.01185 0.01951 -0.00025 - 0.19922 Hong Kong (HK) 0.00382 -0.00287 -0.00327 0.00827 0.02859* 0.00865 0.00716 - page 7 of 11Zoological Studies 60:33 (2021) © 2021 Academia Sinica, Taiwan juvenile stages, benefiting from the availability of food sources and shelter provided by the complex canopy structures (Castro et al. 2002). Sargassum fronds can in fact act as a habitat for various epifaunal crustaceans such as copepods and amphipods (Mukai 1971; Tararam and Wakabara 1981). Juveniles of S. boops less than 4.0 cm SL are known to feed mainly on copepods and decapods (Kimura et al. 1982). Additionally, a previous study reported that the diet of S. boops shifted as their size increased, i.e., the smaller juveniles (4.0 to 5.0 cm SL) feed on clupeoid larvae while the larger juveniles (> 5.0 cm SL) on the juveniles and adults of clupeids (Kimura et al. 1982). Clupeoid juveniles, which could be one of the important food sources of S. boops juveniles, are also commonly found in the coastal habitats in Hong Kong (PTY Leung, unpublished results). Our findings suggest that the seasonal canopyforming Sargassum habitat in Hong Kong could play a potential nursery role as a transient juvenile habitat of S. boops. Low genetic differentiation between populations from Hong Kong and Japan There was no clear genetic structure between Hong Kong and Japan populations in Cytb haplotype network (Fig. 4). Both Hong Kong and Japan populations were rich in haplotype Cytb2. All pairwise FST values were low (-0.02413–0.04120) and similar to the FST values (-0.0411–0.0615) in Noguchi et al. (2012) and can be classified as “little genetic differentiation (FST)” following Hartl and Clark (1997). Low but significant FST value was obtained between Kochi and Hong Kong, suggesting that the Kochi population might be significantly differentiated from that of Hong Kong. But such results might also be due to excessive singleton haplotypes in geographic populations affecting the results of the permutation test. Indeed, Cytb is relatively conservative compared to the control region of the mitochondrial genome, RFLP, and microsatellites (genetic markers commonly used for population genetics), which might result in the observed low and insignificant pairwise FST values. Genetic structure of the S. boops populations need to be further studied with hypervariable gene markers. The low differentiation among all localities suggests that juveniles of S. boops in Hong Kong might be either the offspring of the Japanese population or derived from a single genetic population around the NW Pacific. Based on records of immediate post-birth larvae, S. boops is assumed to utilize the Okinawa Trough as its spawning ground (Uchida et al. 1958; Mochizuki 1977; Noguchi et al. 2012; Takai et al. 2014; Sassa and Konishi 2015). During the spawning season (October to March), the warm Kuroshio Current and the Tsushima Current might contribute to the transport of S. boops offspring from the Okinawa Trough to both sides of the Japanese Archipelago, which in turn may facilitate genetic homogeneity among the geographic populations (Noguchi et al. 2012; Takai et al. 2014). During the spawning season of S. boops, the typical East Asian monsoon (northeasterly monsoon) in winter usually drives the colder and fresher flows southward along the coast of China, i.e., the southward China Coastal Current (Fig. 1a), and consequently this current holds back the northward intrusion of the Kuroshio Branch Current (Jan et al. 2010). Scombrops boops larvae were reported in the southern East China Sea (Sassa and Konishi 2015). Therefore, it is plausible to assume that larvae of S. boops might drift along the China Coastal Current, disperse through the Taiwan Strait and arrive in the coastal waters of Hong Kong. After the disappearance of the seasonal Sargassum habitat since the arrival of the wet season, these juveniles may move to their next transient habitats or migrate back to the deeper water habitat as adults, i.e., the mesopelagic zone in temperate or subtropical areas, e.g., Japan and Taiwan. Likewise, a similar migration pattern was also suggested for S. gilberti, which is a closely related species of S. boops. The juveniles of S. gilberti undertake a long-distance migration from northern Japan, where they use high productive waters during their early life stage, to southern Japanese waters, where they complete their development into adulthood (Itoi et al. 2011). Our findings on S. boops have provided additional evidence on the life history characteristics of this genus. The South China Sea has a known maximum depth of 5,559 m. The continental shelf of the South China Sea might provide suitable habitats for the mesopelagic adults of S. boops. Our finding of its small juveniles in Hong Kong suggests that there could be an unreported population of S. boops in the mesopelagic area of the South China Sea. If so, then the low genetic differentiation between Hong Kong’s juvenile samples and Japanese populations might due to the high migratory rate between the populations that overcome the effect of genetic drift; or insufficient time for subpopulations to reach genetic equilibrium after a recent range expansion event (Slatkin 1993). However, no adult specimen of S. boops has been reported in the mesopelagic zone of the South China Sea. To bridge this knowledge gap, further research on the diversity and ecology of deepwater fishes in the South China Sea region is needed. page 8 of 11Zoological Studies 60:33 (2021) © 2021 Academia Sinica, Taiwan CONCLUSIONS This study represents the first report of a deepwater temperate fish, S. boops, in littoral waters of Hong Kong, making it the southern-most record of Scombrops boops. Our findings showed that the occurrences of these juveniles were limited to the drywet transitional season (March to May), which is also the peak season for the canopying-forming Sargassum beds along the coastal water of Hong Kong. The high genetic relatedness between Hong Kong specimens and Japan populations, suggesting that these juveniles S. boops found in Hong Kong might be either the offspring of the Japanese population or derived from a single genetic population around the North West Pacific. This paper highlighted that subtropical littoral habitats, in particular the seasonal Sargassum habitat, could be one of the important transient juvenile habitats for this temperate deepwater species. Acknowledgments: The authors are thankful to the fisherman Mr. MK Ma, the owner of the chartered fishing vessel (Registration: C704515), who operated purse seine fishing in this study. The authors thank The Agriculture, Fisheries and Conservation Department of The Government of the Hong Kong SAR for providing permit to perform field sampling inside the Marine Parks areas (Permit number: (47) in AF GR MPA 08/9 Pt.1). The authors thank Prof. Shiro Itoi and Shunsuke Noguchi for providing haplotype information in Noguchi et al. (2012) and sharing their latest researches on genus Scombrops. The authors also thank Miss Zoe Ho and Maggie Au for their technical support, and Miss Sharon Chan for copyediting the draft. This study was funded by Environment and Conservation Fund, The Government of the Hong Kong SAR (ECF project 15/2015), and supported by the State Key Laboratory of Marine Pollution, City University of Hong Kong. Authors’ contributions: Jiehong Wei: Data curation, Formal analysis, Investigation, Visualization, WritingOriginal draft. Jiarui Gu: WritingOriginal draft. Min Liu: Conceptualization, Data curation, WritingReview & Editing. Bai-an Lin: Investigation. Gabriel Y. Lee: Investigation. Tak-Cheung Wai: WritingReview & Editing. Paul K.S. Lam: Supervision, WritingReview & Editing. Meng Yan: Investigation, WritingOriginal draft. Priscilla T.Y Leung: Conceptualization, Data curation, Investigation WritingReview & Editing, Project administration, Supervision, WritingReview & Editing. Competing interests: The authors declare that they have no conflict of interests. Availability of data and materials: The collected specimens were deposited in the State Key Laboratory of Marine Pollution, City University of Hong Kong, Hong Kong, China. COI sequences and Cytb haplotypes of the Hong Kong specimens could be found in the NCBI Genbank. Accession number of COI sequences: MK987135–MK987174; Accession number of Cytb haplotypes: MK987120–MK987134. Fasta files of COI sequences and Cytb haplotypes used in this study could be found in DataSet S1. Cytb haplotype information could be found in DataSet S2. Consent for publication: Not applicable. Ethics approval consent to participate: All applicable international, national, and/or institutional guidelines for the care and use of animals were followed by the authors. All necessary permits for sampling and observational field studies have been obtained by the authors from the competent authorities and are mentioned in the acknowledgments. REFERENCES Castro JJ, Santiago JA, Santana-Ortega AT. 2002. A general theory on fish aggregation to floating objects: an alternative to the meeting point hypothesis. Rev Fish Biol Fish 11(3):255–277. doi:10.1023/A:1020302414472. Chu C, Loh KH, Ng CC, Ooi AL, Konishi Y, Huang SP, Chong VC. 2019. Using DNA barcodes to aid the identification of larval fishes in tropical estuarine waters (Malacca Straits, Malaysia). Zool Stud 58:30. doi:10.6620/ZS.2019.58-30. Clement M, Snell Q, Walker P, Posada D, Crandall K. 2002. TCS: estimating gene genealogies. In: Proceeding 16th International Parallel Distributed Processing Symposium, Fort Lauderdale, Florida, 15–19 April 2002. Darriba D, Taboada GL, Doallo R, Posada D. 2012. jModelTest 2: more models, new heuristics and parallel computing. Nat Methods 9(8):772. doi:10.1038/nmeth.2109. Excoffier L, Lischer HEL. 2010. Arlequin suite ver 3.5: a new series of programs to perform population genetics analyses under Linux and Windows. Mol Ecol Resour 10(3):564–567. doi:10.1111/ j.1755-0998.2010.02847.x. Froese R, Pauly D. 2020. FishBase. World Wide Web electronic publication. http://www.fishbase.org/, version (12/2020). Fulton CJ, Berkström C, Wilson SK, Abesamis RA, Bradley M, Åkerlund C, Barrett LT, Bucol AA, Chacin DH, Chong-Seng KM, Coker DJ, Depczynski M, Eggertsen L, Eggertsen M, Ellis D, Evans RD, Graham NAJ, Hoey AS, Holmes TH, Kulbicki M, Leung PTY, Lam PKS, van Lier J, Matis PA, Noble MM, Pérez-Matus A, Piggott C, Radford BT, Tano S, Tinkler P. 2020. Macroalgal meadow habitats support fish and fisheries in diverse tropical seascapes. Fish Fish 21(4):700–717. doi:10.1111/ faf.12455. GBIF: The Global Biodiversity Information Facility. 2020. What is GBIF? Available from https://www.gbif.org/what-is-gbif. Accessed 13 January 2020. Ghedotti MJ, Gruber JN, Barton RW, David MP, Smith WL. 2018. Morphology and evolution of bioluminescent organs in the page 9 of 11Zoological Studies 60:33 (2021)