Seasonal and day–night changes in the vertical distribution of Gnathophausia longispina and G. elegans (Peracarida, Lophogastrida) in the East China Sea

Abstract

Seasonal and day–night changes in the vertical distribution and habitat of mesopelagic crustaceans, Gnathophausia longispina and G. elegans were investigated in the East China Sea during four oceanographic cruises carried out between May 2012 and January 2013. The abundance of G. elegans was approximately one-tenth that of G. longispina, although both species appeared throughout the year. The main distribution range of G. longispina was 100–600 m, and G. elegans was 600–700 m. Only G. longispina showed ontogenetic differences in diel vertical migration (ODVM); most individuals inhabited depths of 500–600 m during the day and 100–400 m at night when younger individuals inhabited shallower layers than mature individuals. Overall, both species showed a small overlap in their vertical distribution at around 600-m depth. Vertical segregation of habitat could have facilitated the sympatric biogeographic distribution of these two congeneric species around the West/Central Pacific Ocean.

INTRODUCTION

The genus Gnathophausia (Peracarida, Lophogastrida) includes eight nominal mesopelagic and bathypelagic micronektonic crustaceans (Willemoës-Suhm, 1873; Meers and Meland, 2012; Petryashov, 2015). They share similar ecological niches with large euphausiids, mysids and pelagic decapods (Brodeur and Yamamura, 2005). The genus comprises Gnathophausia zoea (Willemoës-Suhm, 1873), which is cosmopolitan; G. affinis (G.O. Sars, 1883) and G. bergstadi (Meland and Aas, 2013) are restricted to the Atlantic Ocean; G. childressi (Casanova, 1996), G. longispina (G.O. Sars, 1883) and G. elegans (G.O. Sars, 1883) are distributed only in the Pacific Ocean; G. fagei (Casanova, 1996) is confined to the South China Sea; and G. scapularis (Ortmann, 1906) to the Indian Ocean (Meland and Aas, 2013). Gnathophausia elegans and G. longispina have also been recorded in the East China Sea (ECS; Meers and Meland, 2012), a marginal sea of the western North Pacific; sympatric biogeographical distribution of both species is suspected in the Pacific Ocean. Vertical habitat segregation is a major factor that enables the sympatric biogeographic distribution of congenic species or species-groups in oceanic biocoenosis; this is well documented in mesopelagic zooplankton/micronekton including copepods (Kuriyama and Nishida, 2006; Laakmann et al., 2009), euphausiids (Barange et al., 1990, 1991), mysids and decapods (Roe, 1984; Hopkins et al., 1994).

Although there are studies on the vertical distribution of Gnathophausia members in the Atlantic Ocean (Hargreaves, 1989) and East Pacific Ocean (Pequegnat, 1965; Childress, 1968), there are few studies of G. longispina in the Central/West Pacific Ocean (Reid et al., 1991; Wilson and Boehlert, 1993); comparative data for the ECS are limited. We investigated the seasonal and day–night changes in the vertical distribution of Gnathophausia members and their vertical habitats using samples from the ECS.

METHODS

The methods are detailed in the Supplementary Material, Methods. Briefly, four seasonal surveys were conducted at a fixed station (760-m depth) in the ECS basin (8 May, 25–26 July, 21–22 October 2012 and 29–30 January 2013). Day and night vertical stratified net samplings were conducted with a MOCNESS net. All gnathophausiids in the samples were counted, identified, categorized (juvenile, immature male, immature female, mature male and mature female) and their sizes were measured (carapace length, mm). Weighted mean depths (WMD) of abundance for each gnathophausiids stage were estimated for each sampling period (Barange, 1990).

RESULTS

Both G. longispina and G. elegans were found in all seasonal sampling periods, and no other gnathophausiid species was collected (Table I). Gnathophausia elegans abundance (28–65 individuals 100 m⁻²) was approximately one-tenth of G. longispina abundance (213–627 individuals 100 m⁻²; Fig. 1 and Table I). More than 89% of G. longispina individuals appeared between 100 and 600-m depths, although they were collected from all sampled layers throughout the survey except the 0–50-m layer in October 2012 and the 600–700-m layer in January 2013 (Fig. 1a). Contrarily, G. elegans mainly appeared in the 600–700-m layer throughout the year, although some individuals appeared at 300–600-m depths (Fig. 1b). The WMD of G. longispina, were 100–400 m at night and 500–600 m during the day (P < 0.001); it also varied between different stages at night (Kruskal–Wallis test, P < 0.05; Fig. 2a). In contrast, G. elegans showed no day/night or inter-stage differences in WMDs (Kruskal–Wallis test, P > 0.05) with a gross mean of 623 m (Fig. 2b).

Table I

Seasonal abundance of Gnathophausia longispina and G. elegans collected during the day and night MOCNESS vertical sampling at a fixed oceanographic station (31°45′ N–129°15′ E, 760-m depth) in the ECS

		Abundance (individuals 100 m−2)
Cruise	Date	Mean ± SD	n	Range [minimum–maximum]
			Gnathophausia longispina
YK1202	8 May 2012	626.6 ± 117.0	4	[475.6–745.5]
YK1206	25–26 July 2012	488.7 ± 66.2	4	[419.0–578.0]
YK1209	21–22 October 2012	402.7 ± 168.4	4	[244.8–604.3]
YK1210	29–30 January 2013	213.1 ± 63.1	4	[140.0–294.2]
			Gnathophausia elegans
YK1202	8 May 2012	53.4 ± 18.8	4	[26.6–70.3]
YK1206	25–26 July 2012	65.4 ± 40.5	4	[43.4–126.0]
YK1209	21–22 October 2012	65 ± 38.4	4	[29.3–116.4]
YK1210	29–30 Janurary 2013	27.7 ± 22.6	4	[8.7–60.4]

Note: SD, standard deviation.

Fig. 1

Seasonal changes in day–night vertical distribution in the basin of the ECS: (a) Gnathophausia longispina and (b) G. elegans.

Fig. 2

Ontogenetic and diel vertical distribution by life stage in the basin of ECS: (a) Gnathophausia longispina and (b) G. elegans. Juv: juvenile, Mim: immature male, Fim: immature female, Mmt: mature male, Fmt: mature female. The blue circles indicate seasonal average WMD during the night and red circles indicate that during the day; error bars represent standard deviation. The dashed lines indicate the daytime grand mean WMD of G. longispina and the grand mean WMD of G. elegans; the solid lines indicate the nighttime grand mean WMDs of G. longispina immature and mature individuals, with data for the sexes pooled.

DISCUSSION

The appearance of G. longispina in the ECS basin coincided with the known near-coast distribution of this species (Reid et al., 1991; Wilson and Boehlert, 1993). Our study demonstrates that G. elegans is distributed sympatrically with G. longispina throughout the year at depths up to 700 m, whereas the maximum depth in a previous study exceeded 2000 m in southern ECS (Lee et al., 1980). Therefore, surveying deeper regions can elucidate the spatio-temporal distribution of gnathophausiids in the ECS.

The habitat of G. longispina is shallower than that of G. elegans, and the main vertical distribution range of both species is segregated by a depth of almost 600 m. Species-specific vertical habitat segregation of pelagic Lophogastrida crustaceans have been found in several mesopelagic regions; for example, a 29-species assemblage of midwater decapods and mysids in the eastern Gulf of Mexico showed vertical habitat segregation together with niche (prey) partitioning between species for reducing competition (Hopkins et al., 1994). The vertical habitat segregation for reducing competition and possibly prey partitioning between G. longispina and G. elegans demonstrated here could have facilitated the sympatric biogeographical distribution of these congenic in the Central/West Pacific Ocean (Mauchline and Murano, 1977). However, the relationship of G. longispina and G. elegans with other congeners, G. zoea and G. childressi, in the Pacific Ocean (Meland and Aas, 2013) is still unknown.

The WMD of G. longispina tended to deepen with maturation; juveniles, 186 m; immature individuals, 229 m; and mature individuals, 317 m, with significant differences between those of immature and mature stages at night (Wilcoxon rank-sum test, P = 0.005). The lack of ontogenetic and diel vertical migration of G. elegans may result from its adaptation to a dark and stable habitat (seasonal variations in water temperature were 6.1–6.3°C and salinity was 34.36 annually at 700-m depth: data not shown). Only G. longispina showed ontogenetic distinct diel vertical distribution pattern (ODVM); therefore, younger individuals spent the nighttime at shallower depths than older ones. The ODVM is presumed to be a common ecological behavior in G. longispina, as similar findings have been reported in the Central Pacific Ocean, although it was studied only during summer (Wilson and Boehlert, 1993). The trade-off between predation risk and food intake (Brodeur and Yamamura, 2005) may have caused young G. longispina to migrate to shallow layers, where preys (mainly copepods) are abundant for growth, whereas mature G. longispina are distributed in deeper layers to avoid predation in shallow layers when mesopelagic fish migrate at night (Wang et al., 2019).

CONCLUSION

We hypothesized that vertical habitat segregation of G. longispina and G. elegans enables their sympatric biogeographical distribution. The seasonal vertical distribution range of G. longispina is shallower than that of G. elegans. Although G. longispina showed ontogenetic diel vertical migration where younger individuals spend the night at shallower depths than older ones, the range of migration was between 100 and 600-m depths, supporting our hypothesis.

FUNDING

This work was partially supported by grants from the Dynamics of Commercial Fish Stocks (DoCoFis) program of the Fisheries Agency of Japan and from a research project of the Ministry of Agriculture, Forestry, and Fisheries, Japan.

ACKNOWLEDGEMENTS

We acknowledge the captain and crew of RV Yoko-Maru of Japan Fisheries Research and Education Agency for their assistance in deploying the coupled MOCNESS. We acknowledge Dr S. Ohshimo for providing useful comments on earlier versions of this manuscript. We also thank the two anonymous reviewers for their helpful suggestions and valuable comments on the manuscript. We would like to thank Editage (www.editage.com) for English language editing.