Introduction
Todarodes pacificus (the Japanese flying squid) is distributed in the middle latitudes of the Northwest Pacific, including the Japanese Archipelago and the Korean Peninsula. It migrates between its spawning grounds (in the southwestern Japan Sea, Tsushima Strait, and the northeastern East China Sea) and feeding grounds (the waters around Hokkaido Island including the Sea of Okhotsk) during its 1-year life cycle (Okutani, Reference Okutani and Boyle1983; Arkhipkin et al., Reference Arkhipkin, Rodhouse, Pierce, Sauer, Sakai, Allcock, Arguelles, Bower, Castillo, Ceriola, Chen, Chen, Diaz-Santana, Downey, González, Amores, Green, Guerra, Hendrickson, Ibáňez, Ito, Jereb, Kato, Katugin, Kawano, Kidokoro, Kulik, Laptikhovsky, Lipinski, Liu, Mariátegui, Marin, Medina, Miki, Miyahara, Moltschaniwskyj, Moustahfid, Nabhitabhata, Nanjo, Nigmatullin, Ohtani, Pecl, Perez, Piatkowski, Saikliang, Salinas-Zavala, Steer, Tian, Ueta, Vijai, Wakabayashi, Yamaguchi, Yamashiro, Yamashita and Zeidberg2015; Gleadall et al., Reference Gleadall, Moustahfid, Sauer, Ababouch, Arkhipkin, Bensbai, Elegbede, Faraj, Ferreiro-Velasco, González-Gómez, González-Vallés, Markaida, Morillo-Velarde, Pierce, Pirro, Pita, Roumbedakis, Sakurai, Scheel, Shaw, Veiga, Willette, Winter and Yamaguchi2024). The stock of this species around Japan includes three cohorts based on the spawning period (summer, autumn, and winter), with those caught in the Japan Sea mainly belonging to the autumn cohort (Rosa et al., Reference Rosa, Yamamoto and Sakurai2011). The eggs of this cohort are spawned in the southern Japan Sea and the Tsushima Strait between October and December (Sakurai et al., Reference Sakurai, Kidokoro, Yamashita, Yamamoto, Uchikawa, Takahara, Rosa, Pierce and O'Dor2013). The hatchlings are transported northeastwards along with the Tsushima Warm Current to the waters off Hokkaido Island and grow in subarctic waters during the summer before migrating back to the Tsushima Strait for spawning.
The annual catches of the autumn cohort of T. pacificus increased in the 1990s after a decrease between the mid-1970s and mid-1980s. The long period of poor catches of T. pacificus was perhaps caused not only by excessive fishing effort in the Japan Sea but also by changing environmental conditions (Murata, Reference Murata and Caddy1989; Sakurai et al., Reference Sakurai, Kiyofuji, Saitoh, Goto and Hiyama2000). The recovery of this cohort from the period of poor stock was probably triggered by an expansion of the spawning grounds with the 1989 regime shift (Kidokoro et al., Reference Kidokoro, Goto, Nagasawa, Nishida, Akamine and Sakurai2010); resulting in the catches off Japan increasing to 178,143 tonnes (t) in 1996 (Goto, Reference Goto2002). However, the annual catches of Japan have gradually decreased again since around 2000, dropping to 13,416 t in 2019 (Miyahara et al., Reference Miyahara, Okamoto, Nishijima, Matsukura, Mstsui, Moriyama, Takasaki, Saitoh and Inagake2023). The drastic drop in landings from around Hokkaido could be accounted for not only by the decline of the autumn cohort stock of T. pacificus but also due to changes in distribution (Hokkaido Research Organization, 2024). Overfishing in the northern Japan Sea (outside the Exclusive Economic Zone of Japan) also possibly caused the decline (Park et al., Reference Park, Lee, Seto, Hochberg, Wong, Miller, Takasaki, Kubota, Oozeki, Doshi, Midzik, Hanich, Sullivan, Woods and Kroodsma2020). To clarify the mechanism of distribution changes, it is necessary to investigate more specifically the migration routes of the squid.
Biologging and biotelemetry methods have been developed to improve understanding of migratory routes and behaviour of marine organisms that grow to moderate or large sizes but such methods are still difficult to apply to rather small or physically fragile organisms including T. pacificus.
As a possible archive of environmental conditions experienced by an individual, elemental composition of biogenic carbonates precipitated by marine organisms has been studied intensively. In particular, strontium (Sr) is the most useful element because it is incorporated into aragonitic structures at a relatively high concentration and often reflects environmental conditions in many species (reviewed by Doubleday et al., Reference Doubleday, Harris, Izzo and Gillanders2014; Avigliano et al., Reference Avigliano, Volpedo and Walther2020). The strontium-to-calcium ratio (Sr:Ca) deposited in aragonitic structures has been shown to represent a record of the temperatures experienced by an individual in some species (e.g. Schöne et al., Reference Schöne, Zhang, Radermacher, Thébault, Jacob, Nunn and Maurer2011; Shirai et al., Reference Shirai, Schöne, Miyaji, Radarmacher, Krause and Tanabe2014 for the bivalve Arctica islandica); and otolith Sr:Ca has often been used for estimating the age at which diadromous fish species make the transition between marine and freshwater habitats (e.g. Murase et al., Reference Murase, Kawakami, Irie and Iguchi2019). In gastropods and bivalves, the relationship between temperature and shell Sr:Ca is usually reported to be negative (Schöne et al., Reference Schöne, Zhang, Radermacher, Thébault, Jacob, Nunn and Maurer2011; Füllenbach et al., Reference Füllenbach, Schöne and Mertz-Kraus2015), but it may be positive (Wanamaker et al., Reference Wanamaker, Kreutz, Wilson, Borns, Introne and Feindel2008; Irie and Suzuki, Reference Irie and Suzuki2020), and sometimes there may be no clear relationship (Vander Putten et al., Reference Vander Putten, Dehairs, Keppens and Baeyens2000), depending on species. In squids, the statolith is the only calcified structure allowing the Sr:Ca-based reconstruction of age-associated environmental temperatures (Ikeda et al., Reference Ikeda, Arai, Kidokoro and Sakamoto2003; Liu et al., Reference Liu, Cao, Truesdell, Chen, Chen and Tian2016; Jones et al., Reference Jones, Arkhipkin, Marriott and Pierce2018), because daily increment periodicity in statolith growth enables the estimation of individual age by counting growth rings (reviewed by Jackson, Reference Jackson2004). Ikeda et al. (Reference Ikeda, Arai, Kidokoro and Sakamoto2003) suggested that the negative relationship of statolith Sr:Ca and water temperature could be used to monitor migratory behaviour of T. pacificus captured at sea. Yamaguchi et al. (Reference Yamaguchi, Kawakami and Matsuyama2015, Reference Yamaguchi, Aketagawa, Miyamoto, Hirose and Matsuyama2018) estimated the migratory routes of Uroteuthis edulis after an experimental evaluation of the negative relationship between statolith Sr:Ca and water temperature. Recently, Hosono et al. (Reference Hosono, Irie, Yamamoto, Nakaya, Sakurai, Kawamura and Iwata2022) have also shown a similar negative relationship through rearing experiments with Heterololigo bleekeri. However, without a large laboratory rearing tank under controlled temperature, keeping squid under good conditions is difficult even for a short period (Hanlon, Reference Hanlon, Gilbert, Adelman and Arnold1990; Takahara et al., Reference Takahara, Kidokoro and Sakurai2017).
The approach of the present study is to use a regional ocean model and Lagrangian tracer experiments with the aim of evaluating the quantitative relationship between empirical water temperature and statolith Sr:Ca to improve our understanding of the migration of T. pacificus. Although T. pacificus individuals can swim southwestwards in the Japan Sea against the Tsushima Warm Current in the last half of their lifetime (Kidokoro et al., Reference Kidokoro, Goto, Nagasawa, Nishida, Akamine and Sakurai2010), it is considered that, in the first half, they move northeastwards, probably making the best use of the Tsushima Warm Current. Therefore, micro-increments and Sr:Ca were investigated in the statoliths of immature squid collected in July near the west coast of Hokkaido (their feeding area, in the northeastern Japan Sea). Water temperatures on the trajectory of Lagrangian backward tracers were calculated from the collection site and date.
Materials and methods
Squid samples
A total of 33 specimens of Todarodes pacificus were collected at three stations: near the west coast of Hokkaido Island (WH) on 6th July 2020; east of Tsushima Island (ET) on 15th January 2021; and west of the Goto Islands (WG) on 16th January 2021. Biological data and sampling locations are shown in Table 1 and Figure 1.
Table 1. Collection data of Todarodes pacificus specimens sampled for this study
WH, off the west coast of Hokkaido Island; ET, east of Tsushima Island; WG, west of the Goto Islands; F, female; M, male; SD, standard deviation.
Figure 1. Map of the study area. The three Todarodes pacificus collection sites are indicated with a triangle (off western Hokkaido, WH), a circle (east of Tsushima Island, ET), and a square (west of the Goto Islands, WG).
Micro-increment counts
After removal of the right and left statoliths from all the samples, each was cleaned with 95% ethanol, embedded in Stickwax (Maruto Instrument Ltd., Tokyo, Japan), and mounted on a glass slide with the anterior side upwards. After hardening, the right statolith was ground along the horizontal plane until the nucleus was exposed. It was then turned over and attached to the glass slide with the posterior side up and ground again to the nucleus. Fine-grade waterproof carbide sandpapers and lapping film sheets were used to grind the statolith, and a thin section of the statolith in the lateral plane was obtained for micro-increment counts (Arkhipkin and Shcherbich, Reference Arkhipkin and Shcherbich2012).
The counting of the increments was carried out using a light microscope with a monitor, resulting in a final magnification of 500×. The number of increments on each statolith was counted from the nucleus to the edge of the dorsal dome. Counting of increments was performed three times and the mean value was adopted if the range of counts was within 10% of the mean, and if >10% the statoliths were read again until the range of counts was within 10% of the mean, as determined by a single observer. The hatching date of each squid was estimated from the number of increments and the date of capture, assuming daily deposition of increments (Nakamura and Sakurai, Reference Nakamura and Sakurai1991).
Sr and Ca measurements
The left statoliths from all 33 specimens were prepared for Sr and Ca measurements according to the method of Yamaguchi et al. (Reference Yamaguchi, Kawakami and Matsuyama2015). Each statolith was cleaned, dried, and mounted on a glass slide in a cylindrical silicone stub filled with acrylic resin (Shofu Ortho Palette; SHOFU Inc., Japan) with the anterior side down. After hardening, the polyester block was ground from the anterior side with fine-grade carbide waterproof sandpaper until the nucleus appeared as close to the surface as possible (i.e. the nucleus was not always exposed). Ground statoliths were polished with 3 μm diamond paste before Sr and Ca measurements.
Line analysis was performed to understand the variation in the ratio of Sr and Ca contents in the 33 statoliths using an electron probe microanalyzer (EPMA, JXA-8530F; Jeol Ltd.) installed at Kyushu University. The analytical conditions were as follows: acceleration voltage 10 kV, beam current 100 nA, and beam diameter 5 μm. Sr-Kα and Ca-Kα line intensities were acquired by analysing crystals of thallium acid phthalate and phosphorus electron-transparent heterogeneity, respectively. Measurements were performed from the nucleus to the lateral dome edge of the statolith at measurement intervals of 5 μm and a counting time of 20 s for each spot.
A qualitative analysis of a single squid statolith was conducted using the same analytical conditions as the line analysis to obtain background counts, which were found to be approximately 3500 for Sr and 5500 for Ca. These values were subtracted from the raw intensity (gross counts) of the signal from each statolith measurement and the background and he ratio of Sr and Ca obtained using these values was defined as Sr:Ca.
After measuring the trace elements from the nucleus to the edge of each statolith, the Sr:Ca values were calculated and extracted at 5-day intervals from the hatching date, taking into account the daily age of each squid, assuming that statolith micro-increments were formed at even intervals. The relative change from the hatching date was calculated using the value on the catch date as the standard to cancel the individual differences: this is because it is known from research on Uroteuthis edulis (swordtip squid) that Sr:Ca of several individuals maintained at the same water temperature did not match exactly although the variation trends were similar (Yamaguchi et al., Reference Yamaguchi, Kawakami and Matsuyama2015). Subsequently, 5-day averages were calculated for each sampling area.
Lagrangian backward tracer experiments
The ocean environment was calculated using the high-resolution data assimilation outputs of water temperature and currents produced by Hirose et al. (Reference Hirose, Takayama and Moon2013). Prognostic time integration was modelled by the Research Institute for Applied Mechanics (RIAM) Ocean Model (DREAMS at RIAM [kyushu-u.ac.jp]), which is a hydrostatic, z coordinate, and primitive-equation ocean general circulation model. The spherical coordinate model, with a resolution of 1/12° longitude and 1/15° latitude, covers both the East China Sea and the Japan Sea. Diagnostic analysis, with an approximate Kalman filter, sequentially corrected the modelled variables by assimilating the remotely-sensed surface temperature and height data with in situ temperature and salinity profiles.
In the backward tracer experiments, 10,000 Lagrangian tracers were released at 43.50°N, 140.68°E (WH) on 6 July 2020, the location and date of capture of eight immature specimens (Figure 1). Tracer advections and diffusions were calculated using the method of Kato et al. (Reference Kato, Sakai, Mmasujima, Okazak, Igarashi, Masuda and Awaji2014). The Smagorinsky coefficient was set at 0.1.
For these backward tracer experiments, the squid were assumed to move passively with the ocean currents and the tracers were set to move continuously in the exact opposite direction of the ocean current. There is very little information about vertical migrations of squid, so previous numerical experiments have assumed constant depths of the passive tracers based on an appropriate water temperature range for this species (Fujii et al., Reference Fujii, Hirose, Watanabe and Kidokoro2004). In the present study, an appropriate depth for the tracers was determined based on the results of comparative experiments. The results were evaluated using the data from Sr:Ca and water temperatures in the northeastern Japan Sea for the first half of the squid lifetime.
Lagrangian backward (LB) tracer experiments were conducted at depths of 6 m (LB06), 15 m (LB15), and 30 m (LB30). Only the tracers that reached south of 39°N and west of 136.5°E (SW) in the Japan Sea were extracted as migrating squid data. This is because the species was identified in that region during a recruitment survey conducted by the Fisheries Agency of Japan and the Japan Fisheries Research and Education Agency from 14 to 20 April 2020 (Kubota et al., Reference Kubota, Miyahara, Kaga, Okamoto, Nishijima, Matsukura, Tkasaki, Saitoh and Inagake2021). The survey was carried out at 33 locations east of 133.5°E and south of 39°N, but squid were collected only at survey sites located west of 136.5°E. Although recruitment surveys were not conducted north of around 39°N, the tracers to the north of 39°N and west of 136.5°E (NW) on 17 April 2020, were also found to be useful. Preliminary experiments showed very few or no tracers in the above target areas on 17 April 2020, at a depth of 50 m or deeper (data not shown).
Lagrangian forward tracer experiments
Lagrangian forward tracer experiments were conducted at a depth of 15 m from 1 November 2019, and from 1 January 2020, using the same conditions as those for the Lagrangian backward tracer experiments. The release point was in the eastern Tsushima Strait (33.5°N, 129°E). The number of paralarvae caught in autumn research surveys in the eastern Tsushima Strait has recently been fewer than during the 1990s (Kubota et al., Reference Kubota, Miyahara, Kaga, Okamoto, Nishijima, Matsukura, Tkasaki, Saitoh and Inagake2021). To examine the fate of the lower number of paralarvae, Lagrangian forward tracer experiments were conducted at a depth of 15 m from 1 November 2019, and from 1 January 2020. The release point was in the eastern Tsushima Strait (33.5°N, 129°E).
Water temperature estimation along the migration trajectory
The median points of the Lagrangian tracers that reached the SW Japan Sea on 17 April 2020 were calculated backwards every 5 days from the release day (SW group), and then water temperature at each point was determined using DREAMS_M. Those reaching the NW coordinates on 17 April 2020 (NW group) also were calculated and each water temperature was determined. The correlation between the mean Sr:Ca of the eight squid statoliths and the water temperatures of the SW and NW groups was estimated through the backward tracer experiments of LB06, LB15, and LB30 and regression expressions were calculated.
The water temperature variations experienced by T. pacificus collected at the three sampling locations were estimated through regression expressions that relate the Sr:Ca data to the water temperatures of the SW and NW groups across the three LB experiments. To evaluate the validity of these estimated water temperature variations, they were assessed as a proportion of the appropriate habitat water temperature range known for this species (12–23°C; Sakurai et al., Reference Sakurai, Kidokoro, Yamashita, Yamamoto, Uchikawa, Takahara, Rosa, Pierce and O'Dor2013) and expressed as an appropriate temperature (AT) ratio.
Results
Age at date of capture and estimated hatching date
The age at capture and estimated mean date of hatching for the three sampling areas are listed in Table 1. The squids captured to the west of Hokkaido (WH) in July had hatched about 6 months earlier, while the ET and WG Japan Sea groups caught in January were about 9 months old and had hatched in April the previous year.
Variations of Sr:Ca in statoliths
The mean numbers and standard deviations (SDs) of Sr:Ca measurement points from the nucleus to the lateral dome edge in statoliths of the three groups of sampled squids were 71 ± 9 (WH), 82 ± 13 (ET), and 86 ± 9 (WG). However, even among individuals collected at the same station, their daily ages varied slightly. Additionally, the number of points where Sr and Ca were measured in the statolith did not necessarily correspond to the daily age. Therefore, to represent the Sr:Ca variation relative to the average daily age, the Sr:Ca values for each squid were extracted at 5-day intervals. Consequently, the mean numbers of measurements were 38 (WH), 53 (ET), and 54 (WG).
Changes in Sr:Ca from the centre (hatching date) to the edge (captured date) of the statolith for each of the eight squid captured at WH are shown in Figure 2, and the mean Sr:Ca are shown in Figure 3 for each of the WH, ET, and WG groups. The mean Sr:Ca of the immature squids collected at WH on 6 July (Figure 3A) remained high until the 120th day after hatching (around May) and then declined towards zero around the time of capture. For the mature squids caught at ET on 15 January (Figure 3B), the mean Sr:Ca was positive between the 40th and 130th day after hatching (around June to September), and thereafter was negative until the 180th day (around November), becoming positive again by the day of capture. The changes in mean Sr:Ca for the WG group, captured on 16 January (Figure 3C) decreased steadily from the hatching date to just before the 120th day, becoming negative until the 190th day (around November), just before capture.
Figure 2. Lifetime variation of Sr:Ca in the statoliths of the eight T. pacificus specimens (①–⑧) collected at WH on 6 July 2020. Ordinate: Sr:Ca ratio relative to the value obtained at date of capture. Abscissa: Days before collection date.
Figure 3. Mean relative Sr:Ca variation (±SDs) of T. pacificus collected at WH, ET, and WG. Left ordinate: Sr:Ca relative to that at date of capture. Right ordinates: empirical water temperature estimated through regression expressions for the SW group using Lagrangian backward tracer depth experiments LB06, LB15, and LB30. The lower limit of the water temperature appropriate for T. pacificus (12°C) is indicated in bold. Note that the scales are such that temperature increases downwards, towards the foot of the page. Abscissa: Days since estimated hatching date.
Lagrangian backward tracer experiments
The Lagrangian backward tracers released from WH on 6 July 2020, generally moved back southwestwards in the Japan Sea, and some tracers reached the Tsushima Strait before 1 January 2020 (the estimated hatching date of the squid collected at WH on 6 July 2020; Figure 4). The distribution of tracers on 17 April 2020 is shown in Figure 5. The trajectories of the SW and NW groups of LB15 are shown in Figure 6 (data not shown for trajectories of LB06 and LB30 since they were little different from that of LB15).
Figure 4. Distributions of Lagrangian backward tracers on (A) 17 April and (B) 1 January 2020 for LB15. The release point is indicated by the triangle (location of WH sampling). Todarodes pacificus were identified to the west of 136.5°E and the south of 39°N (SW quadrant) around 17 April 2020 through a recruitment survey.
Figure 5. Histogram of the location of Lagrangian backward tracers that reached further west than 136.5°E on 17 April (SW and NW) for experiments LB06, LB15, and LB30.
Figure 6. Five-day interval median coordinates (±SDs) of Lagrangian backward tracers that reached further west than 136.5°E and south (SW) and the north (NW) of 39°N on 17 April 2020, with LB15.
Lagrangian forward tracer experiments
The result of the forward tracer experiment from 1 January 2020 showed the easternmost tracer reached just west of 136.5°E on 17 April 2020 (Figure 7A), corresponding to the location of the westernmost tracer with the LB15 experiment (Figure 4A). Meanwhile, many forward tracers released on 1 November 2019, advected to the east of 136.5°E by 17 April 2020 (Figure 7B).
Figure 7. Distributions of Lagrangian forward tracers on 17 April 2020, released at a depth of 15 m on (A) 1 January 2020 and (B) 1 November 2019. The release point is indicated by the triangle in the eastern Tsushima Strait (33.5°N, 129°E).
Relationship between Sr:Ca and water temperatures
The correlations between the individual and mean Sr:Ca in the statoliths of the eight immature squid collected at WH on 6 July 2020 and water temperatures at the median points of trajectories of the SW and NW groups on LB06, LB15, and LB30 were assessed (the three LBs vs the two geographical groups, for a total of six comparisons; Table 2). The strongest negative correlation between the mean Sr:Ca and water temperature was in the SW group for LB15 (r = −0.917), and the second strongest in the SW group for LB06 (r = −0.904).
Table 2. Correlation coefficients between relative Sr:Ca ratios in statoliths of T. pacificus (specimens ①–⑧) collected off the west coast of Hokkaido and ambient water temperatures of the SW and NW groups calculated from Lagrangian backward tracer experiments at three different depths LB06, LB15, and LB30 (see section ‘Materials and methods’ for details)
The regression expressions of relationships between the 17 mean Sr:Ca and 17 ambient water temperatures of the same days of the SW and NW groups on LB experiments LB06, LB15, and LB308 (Figure 8) are plotted in Figure 9. The strongest determination coefficient of the six regression expressions was in the SW group for LB15 (R 2 = 0.841).
Figure 8. Ambient water temperature variations of the 5-day-interval median points of the SW and NW groups with Lagrangian backward tracer experiments LB06, LB15, and LB30.
Figure 9. Regression lines and equations (with determination coefficient) for mean relative Sr:Ca in statoliths of T. pacificus vs ambient water temperatures of the SW and NW groups with Lagrangian backward tracer experiments LB06, LB15, and LB30.
Estimation of water temperature variations experienced by the squid
Using the proportionality factors for each regression expression of the six conditions in Figure 9, the empirical variations in lifetime water temperature of the squids were estimated (see the extrapolated temperatures of the six ordinates to the right in Figure 3): the estimated water temperatures differed depending on the proportionality factors of the regression expressions and the water temperatures when and where the squid were caught. The highest AT ratio was observed with LB30, and to some extent with LB15 (Table 3).
Table 3. Estimated AT ratios (%) (ratio of empirical water temperatures lying within appropriate water temperatures of 12–23°C; see final section of ‘Materials and methods’) for T. pacificus collected at each of the three sampling sites (WH, ET, and WG), as estimated through six regression expressions (the SW and NW groups vs the three Lagrangian backward tracer depths LB06, LB15 and LB30)
In the SW group of LB15 (whose regression expression has the strongest determination coefficient), the estimated mean empirical water temperature experienced by the immature squids collected at WH was between 13 and 15°C from January to May, then gradually increased, reaching 17–18°C during May and June (Figure 3A). The water temperature experienced on the collection date by the 10 mature ET squid was estimated to be 14.2°C, and by the 15 mature WG squid to be 15.2°C. The vertical profile of the water temperature in the Tsushima Strait was almost uniform from the surface to a depth of about 150 m from autumn to spring because of seasonal vertical mixing (Shibano et al., Reference Shibano, Morimotoa, Takayama, Takikawa and Ito2019). The mean empirical water temperature variation of the ET squid remained between 13 and 14°C from May to September after a steady decrease from 15°C as juveniles (Figure 3B). Then, the variation increased to 16°C in October and decreased until January. The mean empirical water temperature variation of the 15 squid collected at WG remained more or less between 14 and 15°C from April to August and increased to 18°C in November before a decrease until January (Figure 3C).
To estimate water temperatures the squid were experienced accurately, the most reasonable combination of the six conditions was determined through four grades, A(1) > B(2) > C(2) > D(1), where each numerical value of its grade shows the number of its grade. Overall, the SW group with the LB15 simulation showed the strongest negative correlation (between the mean statolith Sr:Ca and water temperatures at the median points of the tracer trajectories; Table 2), and the third best AT ratio after the SW and NW groups on LB30 (Table 4). For the SW and NW groups with LB30, relatively very few Lagrangian tracers appeared on 17 April 2020. Moreover, the negative correlations and AT ratios of the SW and NW groups with LB06 were lower than those of the SW group on LB15. Therefore, of the six combinations, the regression expression for the SW group with LB15 provided the best estimate of the lifetime empirical water temperature variation of the squid.
Table 4. Score sheet for six combinations of three experiments and two groups across three results of analyses
Grade (the number of each grade), A(1) > B(2) > C(2) > D(1).
Discussion
As a result of the statolith analyses and Lagrangian tracer experiments, the best combination out of the six LB simulations was the condition of the SW group on LB15, and the second best was that of the NW group on LB06 (Table 4). Considering what is known of the ecology of this species (Sakurai et al., Reference Sakurai, Kidokoro, Yamashita, Yamamoto, Uchikawa, Takahara, Rosa, Pierce and O'Dor2013), the fixed depth of 15 m on the tracer experiment is reasonable, although the squid probably undergo diurnal migrations (Kawabata et al., Reference Kawabata, Yatsu, Ueno, Suyama and Kurita2006). Figure 8 suggests that, from April to July, the water temperatures in the northeastern Japan Sea below a depth of 30 m were probably 13°C or lower. The appropriate water temperature range of this species is reported to be between 12 and 23°C (Sakurai et al., Reference Sakurai, Kidokoro, Yamashita, Yamamoto, Uchikawa, Takahara, Rosa, Pierce and O'Dor2013), so a mean depth of 15 m is consistent with its known normal temperature preference range.
Therefore, the Sr:Ca in the statoliths of Todarodes pacificus species showed a negative correlation with empirical water temperatures in the present study, consistent with the findings of Ikeda et al. (Reference Ikeda, Arai, Kidokoro and Sakamoto2003). Furthermore, a quantitative relationship between the Sr:Ca variation and empirical water temperatures for this species was established for the first time.
The temperature variations obtained from statolith Sr:Ca variations of the squid collected in the Tsushima Strait in January generally corresponded to the known seasonal variation of seawater temperature in the sub-surface layer of the southern Japan Sea and Tsushima Strait from summer to winter (online Figure S1). Although the highest sea temperature is usually during September, the squid collected at ET and WG experienced their highest water temperature in October and November, respectively. This time lag suggests that the squid had possibly moved towards the warmer waters, which is consistent with a hypothesis explaining the spawning migration from the Japan Sea to the Tsushima Strait (Kidokoro et al., Reference Kidokoro, Goto, Nagasawa, Nishida, Akamine and Sakurai2010; Rosa et al., Reference Rosa, Yamamoto and Sakurai2011). The results for long-lived squid also clearly indicated that the statolith Sr:Ca variation was affected by the empirical water temperatures.
Thus, the migratory route of T. pacificus can be determined using Lagrangian tracer experiments, assuming that ocean currents influence the migration, at least during its early life stages. This method also has the potential to reveal the relationship between the hatching grounds and distant fishing grounds. A squid catches of round 400 kg in the eastern Tsushima Strait were reported for this week (Nagasaki Prefecture Local Government, 2020) on January 2020. The result of this experiment showed the easternmost tracer reached just west of 136.5°E on 17 April (Figure 7A), corresponding to the location of the westernmost tracer with the LB15 experiment (Figure 4A). Meanwhile, many forward tracers released on 1 November 2019 advected to the east of 136.5°E by 17 April 2020 (Figure 7B). If a larger spawning population had been produced in autumn (as occurred a few decades ago), it is expected that more squid would have approached WH in the following July. In reality, the number of squid present around western Hokkaido has been decreasing catastrophically. These results suggest that the presence of fewer paralarvae in the eastern Tsushima Strait in autumn is one cause of the recent poor catches around western Hokkaido in the following summer, which marks the first success in elucidating the relationship between the hatching grounds and the fishing grounds through the use of a numerical ocean model.
In conclusion, this study successfully demonstrated the efficacy of using a combination of interdisciplinary methods to achieve a better understanding of some of the main features determining the migration of T. pacificus. However, further sampling of squids in the field is required to confirm their migratory routes.
Supplementary Material
The supplementary material for this article can be found at https://doi.org/10.1017/S0025315425000098.
Data
The data that support the findings of this study are available from the corresponding author, upon reasonable request.
Acknowledgements
We thank Dr Tatsuro Adachi for providing technical advice during the EPMA analysis; and Mr Kenichiro Omote of Marino Research Co., Ltd, for his cooperation during the statolith increment analysis.
Author Contributions
T. Yamaguchi performed statolith Sr:Ca measurements, Lagrangian tracer experiments, data analyses, and writing. H. Matsui, H. Miyahara, and H. Kubota collected and measured squids. N. Hirose managed the regional ocean model and revised the manuscript.
Financial support
This study was supported by a grant from the Research and Assessment Program for Fisheries Resources, the Fisheries Agency of Japan to H. Matsui, H. Miyahara, and H. Kubota and the Nakatsuji Foresight Foundation Research Grant to T. Yamaguchi.
Competing interests
None.
Ethical standards
All procedures performed in the present study followed the ethical standards of the Life Science Research Ethics.
