Effects of Temperature on Embryonic Development and Paralarval Behavior of the Neon Flying Squid Ommastrephes Bartramii

Vol. 529: 145–158, 2015 MARINE ECOLOGY PROGRESS SERIES Published June 8 doi: 10.3354/meps11286 Mar Ecol Prog Ser

Effects of temperature on embryonic development and paralarval behavior of the neon flying squid Ommastrephes bartramii

Dharmamony Vijai1,*, Mitsuo Sakai2, Toshie Wakabayashi2,3, Hae-Kyun Yoo1, Yoshiki Kato2, Yasunori Sakurai1

1School of Fisheries Sciences, Hokkaido University, 3-1-1 Minato-cho, Hakodate 041-8611, Japan 2Tohoku National Fisheries Research Institute, 25-259 Shimo-mekurakubo, Samemachi, Hachinohe, Aomori 031-0841, Japan

3Present address: Department of Fisheries Science and Technology, National Fisheries University, 2-7-1 Nagata-Honmachi, Shimonoseki, Yamaguchi 759-6595, Japan

ABSTRACT: Studies on the relationships between the early life stages of the neon flying squid Ommastrephes bartramii and oceanographic conditions are essential for understanding the spatial and temporal distribution patterns of this ecologically and economically important species. In this study, O. bartramii eggs were artificially fertilized and incubated at temperatures found in its known distribution range (16−26°C). Complete organogenesis with normal successive cleavage and the distinct continuity of morphological features was limited to between 18 and 25°C. Exper- imental rearing, cruise-collected specimens, and oceanographic data confirmed that the optimal temperature range for paralarval survival around Hawaii is 18 to 25°C. Embryos reared at 16°C showed abnormal organogenesis; however, normal development resumed when the embryos were transferred to 22 or 24°C after blastoderm formation. Hatchlings showed a variety of swimming behaviors, which may allow squid to regulate their thermal distribution in the wild. Furthermore, ball formation and associated chromatophore expansion might represent an apo - sematic adaptation to imitate unpalatable prey. This study highlights the flexible strategy of O. bartramii embryo development and illustrates how information on paralarval behavior and oceanographic data (sea surface temperature) may be combined to improve our understanding of the factors that influence the survival of this species.

KEY WORDS: Ommastrephidae · Sub-tropical frontal zone · Embryonic development · Sea surface temperature · Paralarval behavior

Resale or republication not permitted without written consent of the publisher

INTRODUCTION Cephalopods are ecological opportunists adapted to exploit favorable environmental conditions (Rod- Ommastrephid squids are the most abundant, house 2013). The embryonic development of cepha - widely distributed, and ecologically active family of lopods is highly temperature dependent, and the ex- cephalopods (Roper et al. 2010). This group supports istence of a minimum temperature for development the largest fisheries for invertebrates in the world; is a general phenomenon among ommastrephids however, little is known about their embryonic (e.g. 12.5°C for Illex illecebrosus, O’Dor et al. 1982; development and paralarval behavior — knowledge 14°C for Todarodes pacificus, Sakurai et al. 1996; and which is critical for better understanding their life 15°C for Dosidicus gigas, Staaf et al. 2011). Neverthe- history (Staaf et al. 2011). less, the thermal limits of true oceanic dwellers, such

*Corresponding author: [email protected] © Inter-Research 2015 · www.int-res.com 146 Mar Ecol Prog Ser 529: 145–158, 2015

as Ommastrephes bartramii, are largely unknown. migration of populations (Bower & Sakurai 1996, Thermal limits determine the spawning grounds, re- Sakurai et al. 2000, Boyle & Rodhouse 2005, Staaf et cruitment levels, and spatiotemporal ranges of para - al. 2008). The planktonic rhynchoteuthion paralarval larvae (O’Dor et al. 1982, Sakurai et al. 2000, Staaf et stage experiences high mortality rates; therefore, al. 2011); thus, it is important to understand these spawning success and paralarval survival rates ranges and the factors that might influence them. (which are in fluenced by large interannual habitat Embryonic development and its de pendence on tem- variability) are the principal factors that determine perature are of particular interest in the context of recruitment levels (Roper et al. 2010, Ichii et al. global warming and the resultant warming of oceans 2011). Growth and ontogenetic changes in squid par- (Rosa et al. 2012, Rodhouse 2013, Xavier et al. 2014). alarvae are largely linked to water temperature O. bartramii is a widely distributed ommastrephid (Sakurai et al. 2000, Staaf et al. 2011). Rosa et al. species with an oceanic bi-subtropical (anti-tropical) (2012) observed the embryonic development of lolig- worldwide distribution, and is fished commercially in inid squid under projected near-future warm condi- the North Pacific Ocean (Roper et al. 2010). The tions, and found that abiotic (hypoxia) conditions North Pacific population of O. bartramii comprises 2 inside the eggs promote metabolic suppression seasonal cohorts (autumn and winter−spring). The followed by premature hatching. Moreover, less- population makes an annual round-trip migration developed hatchlings showed a greater incidence of between its subtropical spawning grounds and its malformations. However, only limited data on envi- northern feeding grounds near the Subarctic Bound- ronment-related changes in survival are available for ary (Yatsu et al. 1998, Bower & Ichii 2005). The main ommastrephid paralarvae. Be cause temperature in - spawning and nursery ground of the autumn cohort fluences the embryonic growth (Sakurai et al. 1995) occurs in the subtropical frontal zone (STFZ) of the and the distribution of the para larvae of this species North Pacific Ocean (Ichii et al. 2009). This location is in the surface layers (Harman & Young 1985, Saito & characterized by enhanced productivity in winter Kubodera 1993, Sakurai et al. 1995), sea surface tem- because of its proximity to the transition zone chloro- perature (SST) is an important factor determining the phyll front. In comparison, the spawning and nursery distribution range and survival of paralarvae. ground of the winter−spring cohort occurs within the Ommastrephid paralarvae are known to begin subtropical domain, which is less productive (Ichii et active swimming immediately after hatching (Staaf al. 2009). The STFZ is defined by a sea-surface salin- et al. 2014). Adult squids generally employ a unique, ity range of 34.6 to 35.2, which generally occurs at dual-mode locomotory system that involves a pulsed ~29 to 34°N, with the subtropical domain occurring jet and fin movements (Bartol et al. 2008, O’Dor et al. south of the STFZ (Roden 1991, Ichii et al. 2009). Pre- 2013). The pulsed jet is formed by first expanding the vailing regional oceanographic features (particularly mantle radially, which causes water to fill the mantle fronts) in combination with temperature impose lim- cavity through intake regions at the anterior margin. its on the geographical dispersal of ce phalopod para - The muscles then contract around the circumference larvae (O’Dor et al. 1982, Moreno et al. 2008, Villa - of the mantle. This contraction increases the pressure nueva et al. 2012). in the mantle cavity, driving water out of the cavity The probability for the successful collection of wild via the funnel. The bending of the funnel below the egg masses and live paralarvae of ommastrephids is body allows squids to swim either arms- or tail-first. very low; moreover, it is difficult to obtain these re- Repetition of these muscle patterns results in a pul - sources by spawning captive brood stock maintained sed jet (Bartol et al. 2008, Staaf 2010, O’Dor et al. in aquaria. As an alternative, artificial fertilization 2013). Squid paralarvae have saccular, rounded bod- techniques have provided valuable information ies, relatively large funnel apertures, and rudimen- about the early development of oceanic squids, and tary fins, which force hatchlings to rely more heavily may provide material for larval rearing experiments on the jet than on the fins for propulsion (Hoar et al. that would help to improve our understanding of the 1994). Paralarvae operate at low and intermediate physiology and ecology of paralarvae (Sakurai et al. Reynolds numbers (Vidal & Haimovici 1998), where- 1995, Villanueva et al. 2012). as juveniles and adults operate at higher Reynolds All ommastrephids have a pelagic lifestyle charac- numbers, and generally have more streamlined bod- terized by the extrusion of fragile, neutrally buoyant ies, larger fins, and relatively small funnel apertures egg masses at depth, the release of larvae into sur- (Bartol et al. 2008).. face plankton, and the use of large-scale water cur- In the first rearing experiments of O. bartramii rent patterns for larval transport and the assisted paralarvae under different temperatures, Sakurai et Vijai et al.: Ommastrephid embryonic development and paralarval behavior 147

al. (1995) reported that the duration of embryonic MATERIALS AND METHODS development to hatching decreased with increasing temperature. However, because this study focused Sampling on developing an artificial fertilization technique, the optimal thermal limits for embryonic develop- Mature, copulated female specimens of Ommas- ment were not assessed (Sakurai et al. 1995). The trephes bartramii (n = 11; 533 ± 32.4 mm dorsal present study aims to identify the thermal window mantle length [DML], mean ± SD) belonging to the for successful development until hatching in vitro winter−spring cohort were captured with jigs on and verify the results through the field collection of hand lines onboard the RV ‘Kaiyo Maru’ (Fisheries specimens in relation to relevant oceanographic Research Agency, Japan) in December 2013. Sam- data (SST). We have also conducted observations on pling was conducted in the northern Hawaiian the behaviors of the paralarvae that influence their region between 20 and 35°N and 150 and 180°W, survival in the wild. around this species’ spawning ground (STFZ) (Fig. 1). Paralarvae were collected during the same cruise by horizontal towing using a multiple opening and closing net with an environmental sensing system (MOCNESS, 1 m2, 0.335 mm mesh size), a Nackthai net (0.50 mm mesh size), and a ring net (2 m diameter, 0.526 mm mesh size). For details on net operations, see Sa kai et al. (2014). The paralarvae were immediately removed from the catch and identified onboard based on morphological (Sweeney et al. 1992) and DNA ana - lyses (Wakabayashi et al. 2006). In addition to the 2013 sampling, long-term data on mature and post- spawning females of O. bartramii specimens (n = 85; 526.7 ± 33.6 mm DML, mean ± SD) collected onboard the RV ‘Hokusei Maru’ (Hokkaido University) were used to determine the favorable SST range. Reproduc- tive maturity was assessed based on the maturity stages I through VII pro- posed by Nigmatullin (1989). These stages were adjusted for O. bartramii (Vijai et al. 2014), whereby Stages I and II were defined as immature, Stage III as early maturing, Stage IV as late maturing, Stage V as mature, Stage VI as spawning state (after the first spawn), and Stage VII as true spent. Stage VII squid were not collected during any of the crui ses. Sam- pling onboard the RV ‘Hokusei Maru’ Fig. 1. (A) Area surveyed by the RV ‘Kaiyo Maru’ for Ommastrephes bar- was performed by manual jigging, tramii adults and paralarvae in December 2013, and the major surrounding gill-netting, and trawling du ring 11 oceanographic regions (Ichii et al. 2011) in the North Pacific Ocean. (B) Sampling locations and corresponding sea surface temperature (SST) con- cruises from February to March each tours where copulated female squid specimens were captured with jigs on year from 1990 to 2000, between 20 hand lines and 35° N and 150 and 180° W in the 148 Mar Ecol Prog Ser 529: 145–158, 2015

Hawaiian region. The spatial and temporal sampling ing every water change, the developmental stages range of RVs ‘Kaiyo Maru’ and ‘Ho ku sei Maru’ coin- were observed and dead eggs and paralarvae were cided with the spawning grounds and season, counted and removed. The period of embryonic de- respectively, of the winter−spring cohort of O. bar- velopment was defined as the time from fertilization tramii (Bower & Ichii 2005). until 50% hatching (Sakurai et al. 1996). One-way ANOVA, followed by Tukey’s honestly significant difference (HSD) test with α = 0.05, were Artificial fertilization used to analyze the number of 8-d-old paralarvae that survived under the various temperature treat- Artificial fertilization of O. bartramii was performed ments. All data were verified for normality and onboard the RV ‘Kaiyo Maru’ using standard tech- homogeneity of variances using Bartlett’s test and niques (Sakurai et al. 1995, Sakai et al. 2011, Villa- the Fligner−Killeen test with α = 0.01, respectively. A nueva et al. 2012). Fresh seawater collected from 500 generalized additive model was applied on paralar- to 1000 m depth was fortified with antibiotics (25 mg val survival data. The predictor variables (incubation l−1 each of ampicillin and streptomycin) and used as temperatures) were modelled as cubic splines and the experimental medium (Staaf et al. 2011). the degree of smoothing was estimated using the mgcv package in R (Wood 2006). All statistical analy- ses were performed using the R language and envi- Embryonic development ronment for statistical computing (R Development Core Team 2013). A total of 150−300 fertilized eggs per Petri dish were maintained in incubators at 16, 18, 19, 20, 21, 22, 23, 24, 25, and 26°C (±0.2°C). Treatment temper- Swimming behavior atures were chosen based on the ecologically re - levant range for O. bartramii (Roper et al. 2010). For The swimming behavior of O. bartramii paralarvae all experiments, the on−off cycle of white timer- was observed and recorded in Kreisel tanks (38 × controlled halogen lamps coincided with the diurnal 24.5 cm; 28 l), Petri dishes (radius = 4.3 cm; height = photoperiod (14 h light:10 h dark). Water was chan - 2 cm), and long cylindrical tubes (radius = 1 cm; ged twice daily by pipetting out fluid until the height = 1.5 m). After hatching, approximately 20 embryos were minimally covered and then adding paralarvae were maintained in each Petri dish to clean seawater containing antibiotics. Culture exper- avoid overcrowding. Paralarvae attaining Stage 30 or iments were repeated 4 times, for all 10 incubation 31 (3 to 5 d after hatching) were transferred by pi - temperatures. Embryos were characterized accord- pette to a 2 l bottle. Those that showed active hop- ing to the developmental stages described by and-sink behavior (n = 300) were further transferred Wata nabe et al. (1996) for Todarodes pacificus. to circular Kreisel tanks and cylindrical tubes for the Appro ximately 800 developing O. bartramii embryos observation of swimming behavior. Black back- in cubated at 16 and 18°C were transferred to higher grounds were placed behind the Kreisel tanks and temperatures (22 and 24°C) after Stage 15 to observe cylindrical tubes to enhance the visibility of the any variation in development from the initial treat- nearly transparent paralarvae (Staaf et al. 2014). ment temperature. Development of O. bartramii par- Mechanical (flashing light and agitation of water) alarvae after Stage 29 (n = 300) was also observed and chemical (drops of ethanol) stimulations were and recorded in aerated circular Kreisel tanks (38 × applied to elicit startle responses in the paralarvae. 24.5 cm; 28 l). The tanks were kept in temperature- Swimming behavior of paralarvae in the Petri controlled (22 to 25°C) rooms to maintain water tem- dishes was observed under a stereomicroscope (Nikon perature. Paralarvae were maintained under these SMZ 1500). Live videos from the microscope were conditions without feeding for approximately 4 to 7 d recorded with a Canon PowerShot A400 IS digi tal until the internal yolk was completely absorbed. camera. Videos of paralarvae in the Kreisel tanks and Embryonic development rates were compared for cylindrical tubes were taken with a Sony HDR- the different incubation temperatures. Embryonic CX590V handycam. All video footage was annotated, development and mortality rates at each temperature reviewed, and analyzed. Selected sequences from were determined under a dissecting microscope and the 30 frames s−1 videos were captured with VLC 2.1 were based on the percentage of normally develop- and exported as frames into ImageJ (http://rsbweb. ing embryos in each dish (Sakurai et al. 1996). Dur- nih. gov/ij/) to ob serve swimming behavior and ball Vijai et al.: Ommastrephid embryonic development and paralarval behavior 149

formation. Swimming speed was measured as the number of mantle lengths traveled per second, based on frame rate (Staaf et al. 2014). Because the exact distance of squid from the camera could not always be determined, DML was used to calibrate all meas- urements (Staaf 2010).

Sea surface temperature

Satellite-derived data were used to study the interannual variability of SST in winter from 1990 to 2000 and in 2013 in the North Pacific Ocean. Monthly Fig. 2. Generalized additive model plot for Ommastrephes average SST data were compiled from the Advanced bartramii paralarvae showing the effect of temperature on Very High-Resolution Radiometer (AVHRR) Path - survival. Fitted line (solid line) and 95% confidence interval finder v. 5.0 datasets produced by the Physical (area between dashed lines) are shown Oceanography Distributed Active Archive Center (PO.DAAC) at NASA’s Jet Propulsion Laboratory (ftp://podaac-ftp.jpl.nasa.gov/OceanTemperature/ avhrr/L3/pathfinder_v5/monthly/), with a spatial resolution of ~0.25° for both latitude and longitude. SST was also recorded in situ using conductivity−temperature−depth casts at all stations. The GMT software package (http://gmt.soest.hawaii.edu/) was used to produce contour lines depicting SST distribution.

RESULTS

Temperature effects

Embryonic development

The fertilization percentage in the artificial fertilization experiments ranged from 16% (at 16°C) to 98% (at 22°C), and development was highly temperature dependent. At 16°C, normal embryonic development was observed only until Stage 15; beyond that development became abnormal. Normal development with high survival rates (>50%) until hatching was observed between 18 and 25°C. Mean survival rates on Day 8 of larvae reared at different Fig. 3. (A) Relationship between the development of Om- temperatures were significantly different (1-way mastrephes bartramii eggs and incubation temperature (10 ANOVA: F9,21 = 8.245, p < 0.001). Pairwise compar- different temperatures were examined). Horizontal dashed isons revealed that survival rates at 16 and 26°C were line represents the hatching stage. (B) Relationship between significantly lower than those at all other tempera- egg mortality rates and time (d) after fertilization at 10 different incubation temperatures tures (Tukey’s HSD: p < 0.05). The generalized additive model indicated that temperature was significant (p < 0.001), and explained 88.5% of variation in (Fig. 3B). Mortality was low between 18 and 25°C paralarval survival (Fig. 2). Embryonic development (Fig. 3B). between 16 and 18°C was slower (Fig. 3A) than that Hatching occurred at Stages 26 to 28 at all tem - at higher temperatures, with embryos incubated at peratures except 16°C; no embryos survived beyond 26°C suffering from high mortality (84%) after 24 h Stage 21 at 16°C. Abnormal hatching occurred in 150 Mar Ecol Prog Ser 529: 145–158, 2015

Fig. 5. (A) Number of Ommastrephes bartramii paralarvae collected onboard the RV ‘Kaiyo Maru’ and (B) number of O. bartramii adults (mature and spawned state) collected onboard the RV ‘Hokusei Maru’, plotted against in situ sea Fig. 4. Relationship between the development of Ommas- surface temperature (SST) trephes bartramii eggs and time (h) at different incubation temperatures. (A) Eggs transferred from incubation temperatures of 16 or 18°C to 22°C; (B) eggs transferred from incu- temperatures (22 and 24°C) after Stage 15 developed bation temperatures of 16 or 18°C to 24°C. Dashed lines (de- normally and hatched successfully (Fig. 4A). Devel- rived from Fig. 3A) represent the developmental trajectory opment rates of embryos transferred from 18°C to at a constant (unchanged) temperature higher temperatures (22° and 24°C) increased, and these embryos hatched 4 d earlier than embryos approximately 40% of all embryos at 18°C. At 26°C, maintained at 18°C (Fig. 4B). the mantles of all developing larvae were completely inverted after Stage 15. These paralarvae had ex - posed viscera, were unable to swim, and survived for Sea surface temperature less than 1 d after hatching. Fig. 1B shows the distribution of female squid spe - ci mens collected by the RV ‘Kaiyo Maru’ in relation Differential effects of temperature to SST. Ommastrephes bartramii females and paralarvae were captured at temperatures between 18 Cleavage and blastoderm formation occur until and 25°C. Fig. 5A shows paralarval distribution in Stage 15, and were observed at all treatment temper- relation to in situ SST. Mature and spent O. bartramii atures. Stage 16 represents the onset of organogene- females collected by the RV ‘Hokusei Maru’ were sis, with the length of time of transition from Stage 15 distributed at SST between 16 and 24°C (Fig. 5B). to Stage 16 being highly varied, ranging from a maximum of 20 h at 26°C to a maximum of 75 h at 16°C. At 16°C, there was no chorion expansion after Stage Swimming behavior 15, and de velopment beyond this stage was mostly (90%) ab normal, with retarded arm and mantle de- Muscular contractions began at Stage 23 in development. All eggs transferred from 16°C to higher veloping embryos. Before hatching, developing em - Vijai et al.: Ommastrephid embryonic development and paralarval behavior 151

Fig. 6. (A) Video frames and (B) diagrams of the opening and closure of the mantle aperture around the head during a single jet propulsion in Ommastrephes bartramii paralarvae. Arrows show the direction of movement of the mantle. (C,D) Representative traces from standard videos (30 frames s−1) of maintenance jetting and vertical jetting, respectively. Normalized mantle width and aperture are shown bryos always maintained a ventral-side-up posture in ally facing up. Escape jets were the most rapid form the dish. Soon after hatching (within 8−10 s), the par- of propulsion and appeared to be similar to those of alarvae started swimming by horizontal jetting, adult squid. The maximum speed measured in an always with the ventral side up. Swimming was escape jet was 3.12 cm s−1. Resting paralarvae res - achieved by jet propulsion with associated mantle ponded to physical contact from other paralarvae by contractions, as shown in Fig. 6A,B. From Stage 30, producing escape jets. Pulsing behavior observed in the paralarvae demonstrated a variety of swimming Petri dishes consisted of continuous contractions and behaviors, categorized as maintenance jets, vertical expansions of the mantle with fin flaps, but with zero jets, horizontal jets, escape jets, passive sinking, and velocity gain in any direction. pulsing. The categorization was modified from that Jetting was achieved by mantle contraction. developed by Staaf et al. (2014) for the ommastrephid Fig. 6A,B shows the opening and closing of a mantle Dosidicus gigas. Maintenance jetting (observed in aperture during a jet. The width of the mantle and the Kreisel tanks, Petri dishes, and cylindrical tubes) mantle aperture, normalized by mantle length, are consisted of individual jets counteracting sinking to shown in Fig. 6C,D. The frequency of mantle aper- maintain the vertical position of paralarvae in the ture closure was higher for maintenance jets than for water column. Paralarvae always sank when they vertical jets. Paralarvae moved up during mantle were not actively swimming. With regard to vertical contractions and sank during mantle expansions jetting, we observed prolonged periods of repeated (water refilling), with net vertical displacement being jetting that exceeded the velocity of passive sinking, determined by the duration of the refilling phase. with maximum measured speeds of 0.48 cm s−1. Hori - This behavior was normally followed by periods of zontal jets observed in Petri dishes consisted of sinking that lasted much longer (3−4 s) than a single repeated jets, during which the ventral side was usu- refilling period, and during which no active swim- 152 Mar Ecol Prog Ser 529: 145–158, 2015

DISCUSSION

Temperature effects

Embryonic development

This study identified the potential thermal limits for Ommastrephes bartramii to complete successful in vitro development (18−25°C), in addition to con- Fig. 7. Ball formation. (A) Diagrammatic representation of a Stage-33 Ommastrephes bartramii paralarvae. a: arms I and firming the significance of temperature for embry- II; ap: anal papilla; bm: buccal mass; d: digestive gland; e: onic development. Within the optimal range, the eye; f: fin; fu: funnel; g: gill; i: ink sac; m: mantle; og: optic duration of embryonic development to hatching de - ganglion; p: proboscis; ps: proboscis sucker; sm: stomach; st: creased with increasing temperature and vice versa. statocyst; vg: visceral ganglion; y: inner yolk. (B) Diagram showing the retracted proboscis and head forming the ball The large number of embryos that developed abnor- posture. Internal organs and other parts are similar to those malities outside of these thermal limits may have labeled in A. Arrow shows the direction of retraction (to- been due to hypo- or hyper-embryogenesis (Sakurai wards the ventral side) et al. 1996) coupled with hypoxic stress (Rosa et al. 2012). Because cephalopods are poikilotherms, war - mer water temperatures under a global warming sce- ming was observed. The proboscis was always in a nario would accelerate growth rates (subject to the stretched position during vertical and horizontal jets, availability of sufficient oxygen), shorten life spans, presumably for balance. Spontaneous extension of and increase turnover (Pecl & Jackson 2008, Rod- the proboscis was also frequent. During the ‘rest house 2013), which might, in turn, drive changes in periods’ (10 min to 2 h) of healthy paralarvae, the the life-history parameters of this species. proboscis suckers were attached to the wall of the The identification of the thermal limits for this spe- Kreisel tank, maintaining the body in an inactive cies during the early stages of development ex- state at an inclined angle in the water column. The plained the distribution of ommastrephid species same behavior was observed in the Petri dishes, and around Hawaii. Egg masses of O. bartramii have the sucker grip was sufficiently strong to withstand never been observed in nature; however, based on mild agitation. paralarval distribution data, SSTs ranging from 21 to 25°C are considered favorable for egg hatching in the North Pacific (Hayase 1995, Bower 1996, Mori et Ball formation al. 1999). Our rearing experiment results indicated that the lower limit of favorable temperatures is 18°C. Any sudden provocation (flash of light, agitation of Although we were unable to observe embryonic water, or drop of ethanol) resulted in the rapid retrac- growth at 17°C because of limited resources, Sakurai tion of the head into the mantle, with paralarvae et al. (1995) obtained hatchlings at this temperature, remaining in a ball shape (Fig. 7). Distortions in the but observed frequent developmental abnormalities shape and relative position of body parts while the and cessation of development during the early stages animals were in the ball formation are illustrated in of embryogenesis. Furthermore, the cruises (RV Fig. 7A,B. Ball formation was always accompanied ‘Hokusei Maru’) collected mature (about to spawn) by chromatophore expansion. When the chromato- and spawned state (post-spawning) O. bartramii phores expanded, their surface area increased more females at SSTs ranging from 16 to 25°C (Fig. 8), in than 6-fold (probably to the maximum extent), the addition to paralarvae at SSTs ranging from 18 to head contracted into the mantle (Fig. 7), and the ani- 25°C (RV ‘Kaiyo Maru’; Fig. 9). These observations mal resembled a large orange when viewed under provide evidence that spawning and hatching might the microscope (Young 1972). If the duration of the occur within this temperature range. stimulus was long enough, the ‘ball’ posture was Temperature requirements restrict the spatial and maintained for a longer period and the paralarvae temporal distribution of spawning in ommastrephid sank, remaining in the bottom of the dish. Mainte- squid. Temperature requirements regulate the distri- nance of the ball posture ceased when the stimulus bution and migration ranges of ommastrephids, in - was withdrawn, taking from a few seconds to several cluding Illex illecebrosus (O’Dor et al. 1982), Toda - minutes to recover normal shape. rodes pacificus (Sakurai et al. 2000), and Dosidicus Vijai et al.: Ommastrephid embryonic development and paralarval behavior 153

Fig. 8. RV ‘Hokusei Maru’ sampling locations (black dots) where mature and spawned state Ommastrephes bartramii females were captured. Sea surface temperature (SST) data are composited monthly data for February of each year, except 1991, for which March data are shown 154 Mar Ecol Prog Ser 529: 145–158, 2015

temperature increases (O’Dor et al. 1982). Further research is required to determine the duration and temperature range under which such stasis may be maintained. The normal development of embryos transferred from 16°C to 22 or 24°C also suggests that O. bartramii may be able to withstand larger temperature fluctuations. The adaptation of embryos to these varying conditions may allow squid to survive for extended periods, even when their egg masses are released at temperatures below 16°C, and to continue normal development when favorable temperature conditions return, which is a likely scenario Fig. 9. Locations and sea surface temperature (SST) contours where Omma- in nature. strephes bartramii paralarvae were collected in December 2013 Depending on their density, egg masses accumulate in the mid-water gigas (Staaf et al. 2011). As an r-strategist species, pycnocline (Bower & Sakurai 1996, Sakurai et al. this requirement may be a limiting factor in the range 2000, Staaf et al. 2008). After hatching, paralarvae of O. bartramii, and may explain why it migrates to are recruited to the surface waters (Bakum & Csirke subtropical regions for spawning. SST varies inter - 1998, Sakurai et al. 2000). At the spawning grounds annually in the North Pacific (Tanimoto et al. 1993, near Hawaii, paralarvae of O. bartramii are distrib- Sheng 1999, Deser et al. 2010, Kosaka & Xie 2013), uted in the surface layers, where the pycnocline tem- with major anomalies occurring in the STFZ (Qiu et perature is approximately 15 to 17°C (Bingham & al. 2014). Because the lifespan of ommastrephid Lukas 1996, Seki et al. 2002) during the putative squids is just 1 yr, this variation affects the distribu- spawning season of the winter−spring cohort tion and spawning area of these species. Year-to- (December−March), which is lower than the ideal year fluctuations in oceanographic factors and pro- range of SST identified here. This observation indi- duction cycles also account for large variations in cates that paralarvae may hatch at lower tempera- recruitment success (Jennings et al. 2001). Sinclair’s tures in the pycnocline, but survive in habitats at (1988) hypothesis regarding the variation in recruit- higher SSTs. Our findings indicate that putative ment emphasizes the importance of the relationship spawning and hatching temperatures in the waters between the time of spawning and stable oceano- around Hawaii range from 16 to 25°C, and that par- graphic features that retain larvae in favorable envi- alarval distribution in the surface layers occurs from ronments. 18 to 25°C. Spawning may also occur at lower temperatures (<16°C). The ability of O. bartramii paralarvae to adapt to Differential effects of temperature higher temperatures has not been previously stud- ied; however, such information is critical for under- Several phases of squid development have been standing how early developmental stages respond demonstrated to be differentially affected by temper- to global warming. This study identified the upper ature (O’Dor et al. 1982). We recorded evidence of temperature limit beyond which deformity of body the differential effects of exposure to low tempera- structures occurs, and would have deleterious ture. Although cleavage and blastoderm formation effects on embryonic survival and growth by creat- were normal at all temperatures, organogenesis ap- ing challenges to feeding (Rosa et al. 2012). The peared to be sensitive to temperatures outside the inability of O. bartramii paralarvae to survive at optimal thermal limits. Experiments investigating high temperatures (26°C) may explain its absence embryonic growth at 16°C with subsequent transfer from tropical waters. A temperature increase in the to higher temperatures (22 and 24°C) indicated that present subtropical spawning area as a result of underdeveloped eggs may remain in stasis until the global warming may result in a shift in the distribu- Vijai et al.: Ommastrephid embryonic development and paralarval behavior 155

tion range of this squid species towards higher lati- the same position (ventral side up) in the Petri tudes (Zeidberg & Robison 2007, Cheung et al. dishes used in the present study, probably for the 2009, Chen et al. 2011, Hoving et al. 2013, Stewart same reason. et al. 2014).

Ball formation Swimming behavior When disturbed, the strong head-retractor mus- The vertical jets produced by the paralarvae may cles enable paralarvae to rapidly withdraw their explain their ascent mechanism from the mid-water head and proboscis ventrally into the mantle cavity hatching area to the surface water where they feed. (Shigeno et al. 2001). Ball formation in gonatid par- These jets may also be used to drive diel vertical alarvae is considered a defensive body posture that migrations, allowing young cephalopods to sink to mimics the appearance (size and color) of the small dee per depths during the day, to avoid visual preda- stinging hydromedusan jellyfish Aglantha digitalis tors, and then rise to the surface at night (Staaf et (Arkhipkin & Bizikov 1996). Ball formation has also al. 2008). Staaf (2010) calculated that a paralarval been observed in chemically preserved samples migration of 15 m would take less than 1 h for D. (e.g. Cranchia scabra, Young 1972; T. pacificus, gigas at the observed swimming speed of 0.5 cm s−1. Shigeno et al. 2001). Withdrawal of the head into Vertical swimming experiments conducted by Yoo et the mantle cavity during fixation illustrates the par- al. (2014) with T. pacificus paralarvae at several alarval response to irritants (alcohol and formalin) temp eratures also revealed that the paralarvae al - by adopting the ball posture. O’Dor et al. (1985) ways have a preference for surface layers with high - observed the complete withdrawal of the head into er temperatures. In this context, hop-and-sink be - the mantle with simultaneous contraction of the havior may be an adaptation for conserving energy mantle lip in the rhynchoteuthion of I. illecebrosus, (Staaf 2010). and interpreted this as a feeding behavior that en- Maintenance jetting may be achieved by leaking abled the squid to feed upon detritus adhered to the water through the mantle aperture rather than ex - mantle’s mucous covering. We observed the suckers pelling it solely through the funnel (Staaf et al. of the proboscis close to the mouth in the ball state, 2014). Escape jet locomotion is the fastest swimming possibly supporting that this activity is a feeding behavior (3.12 cm s−1), and probably enables squid behavior. The sinking nature of paralarvae that to avoid predation. Staaf et al. (2008) reported that adopt the ball posture may enable them to move swimming speed in 3- to 6-d-old D. gigas paralarvae away from danger zones. Since head-withdrawal varies from simply maintaining their position in the behavior is caused by me chanical (flashing light water column (net velocity of 0 cm s−1, generated by and water agitation) and chemical (alcohol and for- the animal’s swimming velocity counteracting its malin) stimulation, this behavior may be a general sinking velocity) to 0.51 cm s−1, which is the fastest defensive reflex of various oceanic ommastrephid net upward swimming speed maintained over sev- squid paralarvae. eral seconds. These researchers also observed a dis- Chromatophore expansion associated with the ball turbance-provoked cessation of normal vertical posture may generate aposematic coloration. Warn- swimming and commencement of jetting in a tight ing colors and patterns are an adaptation of many circle, one to several circuits, before resuming the prey species, providing a signal to predators of the usual jetting up and sinking down. Although fin potential cost of making an attack. Orange coloration flapping was observed, its role in movement was is associated with chemical defenses in at least 3 limited, and directional swimming was achieved by phyla (e.g. the ascidian Ecteinascidia turbinata, the controlling the funnel nozzle (Staaf 2010). Pulsing holothurian Psolus chitonoides, and the octocoral Bri- may be a predatory behavior (Staaf 2010), as areum asbestinum; Young & Bingham 1987), with observed in Doryteuthis opalescens hatchlings, in 60% of unpalatable marine invertebrate larvae being which predation on Artemia spp. is sometimes pre- red or orange colored (Lindquist & Hay 1996). ceded by frequent mantle contractions (Preuss et al. Aguado & Marin (2007) reported orange coloration as 1997). Before hatching, the embryo’s anatomy and aposematic, while Valdés et al. (2013) reported center of gravity orient it with the dorsal side facing bright colors as being aposematic in marine mol- towards the inner wall of the chorion (Marthy 1994). lusks. Because ommastrephid paralarvae lack mor- After hatching, swimming paralarvae maintained phological defenses, the orange color and ball shape 156 Mar Ecol Prog Ser 529: 145–158, 2015

adopted by paralarvae when threatened by preda- nisms. Annu Rev Mar Sci 2: 115−143 tors may imitate unpalatable organisms that share Harman RF, Young RE (1985) The larvae of ommastrephid squids (Cephalopoda, Teuthoidea) from Hawaiian wa- the same habitat. ters. Vie Milieu 35: 211−222 In conclusion, this study provides insights into the Hayase S (1995) Distribution of spawning grounds of flying flexible strategy of O. bartramii embryo development squid, Ommastrephes bartrami, in the North Pacific and demonstrates how information on paralarval Ocean. Jpn Agric Res Q 29:65−72 Hoar J, Sim E, Webber D, O’Dor RK (1994) The role of fins in behavior and oceanographic parameters may be the competition between squid and fish. In: Maddock L, combined to improve our understanding of the fac- Bone Q, Rayner J (eds) Mechanics and physiology of ani- tors that influence the survival success of this group mal swimming. Cambridge University Press, Cambridge, of squid. p 27−33 Hoving HJT, Gilly WF, Markaida U, Benoit-Bird KJ and others (2013) Extreme plasticity in life-history strategy Acknowledgements. We are grateful to the captains and crew allows a migratory predator (jumbo squid) to cope with a of RV ‘Hokusei Maru’ and RV ‘Kaiyo Maru’ and all cruise changing climate. Glob Change Biol 19: 2089−2103 participants. We thank J. Bower, P. Puneeta, and I. Alabia for Ichii T, Mahapatra K, Sakai M, Okada Y (2009) Life history their critical readings of the manuscript. This study was sup- of the neon flying squid: effect of the oceanographic ported by the Research Program on Climate Change Adap- regime in the North Pacific Ocean. Mar Ecol Prog Ser tation (RECCA) under the Ministry of Education, Culture, 378: 1−11 Sports, Science and Technology (MEXT), Japan, and by the Ichii T, Mahapatra K, Sakai M, Wakabayashi T and others ‘Project on the Evaluation of the Status of International Fish- (2011) Changes in abundance of the neon flying squid ery Resources’ by the Fisheries Agency of Japan. This study Ommastrephes bartramii in relation to climate change in was supported in part by JSPS KAKENHI grant no. 26450276. the central North Pacific Ocean. Mar Ecol Prog Ser 441: 151−164 Jennings S, Kaiser M, Reynolds JD (2001) Marine fisheries LITERATURE CITED ecology. Wiley-Blackwell, Oxford Kosaka Y, Xie SP (2013) Recent global-warming hiatus tied Aguado F, Marin A (2007) Warning coloration associated to equatorial Pacific surface cooling. Nature 501:403−407 with nematocyst-based defences in aeolidiodean nudi- Lindquist N, Hay ME (1996) Palatability and chemical branchs. J Molluscan Stud 73: 23−28 defense of marine invertebrate larvae. Ecol Monogr 66: Arkhipkin A, Bizikov V (1996) Possible imitation of jellyfish 431−450 by the squid paralarvae of the family Gonatidae Marthy HJ (1994) Telolecithal eggs need attention for eval- (Cephalopoda, Oegopsida). Polar Biol 16: 531−534 uating the morphogenetic role of gravity in embryogene- Bakum A, Csirke J (1998) Environmental processes and sis. In: Oser H, Guyenne TD (eds) Life sciences research recruitment variability. In: Rodhouse P, Dawe E, O’Dor in space. Proceedings of the Fifth European Symposium, RK (eds) Squid recruitment dynamics: the genus Illex as 26 September−1 October 1993, Arcachon, France. ESA a model, the commercial Illex species and influence on SP-366. European Space Agency, Arcachon, France, variability. FAO, Rome, p 105−124 p 177−180 Bartol IK, Krueger PS, Thompson JT, Stewart WJ (2008) Moreno A, Dos Santos A, Piatkowski U, Santos AMP and Swimming dynamics and propulsive efficiency of squids others (2008) Distribution of cephalopod paralarvae in throughout ontogeny. Integr Comp Biol 48: 720−733 relation to the regional oceanography of the western Bingham FM, Lukas R (1996) Seasonal cycles of tempera- Iberia. J Plankton Res 31: 73−91 ture, salinity and dissolved oxygen observed in the Mori J, Tanaka H, Yatsu A (1999) Paralarvae and adults of Hawaii Ocean Time-series. Deep-Sea Res II 43:199−213 neon flying squid (Ommastrephes bartrami) occurred in Bower JR (1996) Estimated paralarval drift and inferred the subtropical North Pacific Ocean in autumn. Heisei 9 hatching sites for Ommastrephes bartramii (Cephalo - nendo ikarui shigen kenkyuu kaigi houkoku [Report of poda: Ommastrephidae) near the Hawaiian Archipelago. the 1997 Meeting on Squid Resources]. Tohoku National Fish Bull 94: 398−411 Fisheries Research, Tohoku Bower JR, Ichii T (2005) The red flying squid (Ommas- Nigmatullin CHM (1989) Las especies de calamar mas trephes bartramii): a review of recent research and the abundates del Atlantico Sudoeste y sinopsis sobre la eco- fishery in Japan. Fish Res 76: 39−55 logia del calamar (Illex argentinus). Rev Frente Maritimo Bower JR, Sakurai Y (1996) Laboratory observations on 5: 71−81 Todarodes pacificus (Cephalopoda: Ommastrephidae) O’Dor RK, Balch N, Foy EA, Hirtle RWM, Johnston DA egg masses. Am Malacol Bull 13:65−71 (1982) Embryonic development of the squid, Illex ille - Boyle P, Rodhouse P (2005) Cephalopods: ecology and fish- cebrosus, and effect of temperature on development eries. Wiley-Blackwell, Oxford rates. J Northwest Atl Fish Sci 3: 41−45 Chen IC, Hill JK, Ohlemüller R, Roy DB, and others (2011) O’Dor RK, Helm P, Balch N (1985) Can rhyncoteuthions sus- Rapid range shifts of species associated with high levels pension feed? (Mollusca: Cephalopoda). Vie Milieu 35: of climate warming. Science 333:1024−1026 267−272 Cheung WWL, Lam VWY, Sarmiento JL, Kearney K and oth- O’Dor R, Stewart J, Gilly W, Payne J, Borges TC, Thys T ers (2009) Projecting global marine biodiversity impacts (2013) Squid rocket science: how squid launch into air. under climate change scenarios. Fish Fish 10: 235−251 Deep-Sea Res II 95:113−118 Deser C, Alexander MA, Xie SP, Phillips AS (2010) Sea Pecl G, Jackson G (2008) The potential impacts of climate surface temperature variability: patterns and mecha- change on inshore squid: biology, ecology and fisheries. Vijai et al.: Ommastrephid embryonic development and paralarval behavior 157

Rev Fish Biol Fish 18:373−385 Early ontogeny of the Japanese common squid Toda - Preuss T, Lebaric ZN, Gilly WF (1997) Post-hatching devel- rodes pacificus (Cephalopoda, Ommastrephidae) with opment of circular mantle muscles in the squid Loligo special reference to its characteristic morphology and opalescens. Biol Bull 192:375−387 ecological significance. Zool Sci 18: 1011−1026 Qiu C, Kawamura H, Mao H, Wu J (2014) Mechanisms of the Sinclair M (1988) Marine populations: an essay on popula- disappearance of sea surface temperature fronts in the tion regulation and speciation. University of Washington subtropical North Pacific Ocean. J Geophys Res Ocean Press, Seattle, WA 119: 4389−4398 Staaf DJ (2010) Reproduction and early life of the Humboldt R Development Core Team (2013) R: A language and en- squid. PhD dissertation, Stanford University, CA vironment for statistical computing. R Foundation for Staaf DJ, Camarillo-Coop S, Haddock SHD, Nyack AC and Statistical Computing, Vienna. www.r-project.org/ others (2008) Natural egg mass deposition by the Hum- Roden G (1991) Subarctic-subtropical transition zone of the boldt squid (Dosidicus gigas) in the Gulf of California North Pacific: large-scale aspects and mesoscale struc- and characteristics of hatchlings and paralarvae. J Mar ture. NOAA Natl Mar Fish Serv Tech Rep 105: 138 Biol Assoc UK 88:759−770 Rodhouse PGK (2013) Role of squid in the Southern Ocean Staaf DJ, Zeidberg LD, Gilly WF (2011) Effects of tempera- pelagic ecosystem and the possible consequences of cli- ture on embryonic development of the Humboldt squid mate change. Deep-Sea Res II 95:129−138 Dosidicus gigas. Mar Ecol Prog Ser 441: 165−175 Roper CFE, Nigmatullin C, Jereb P (2010) Family Ommas- Staaf DJ, Gilly WF, Denny MW (2014) Aperture effects in trephidae. In: Jereb P, Roper CFE (eds) Cephalopods squid jet propulsion. J Exp Biol 217: 1588−1600 world. An annotated illustrated catalog of species known Stewart JS, Hazen EL, Bograd SJ, Byrnes JEK and others to date. Vol 2. Myopsid oegopsid squids. FAO Species (2014) Combined climate- and prey-mediated range Catalogue for Fishery Purposes No. 4, Vol 2. FAO, Rome, expansion of Humboldt squid (Dosidicus gigas), a large p 269−347 marine predator in the California Current System. Glob Rosa R, Pimentel MS, Boavida-Portugal J, Teixeira T, Change Biol 20: 1832−1843 Trübenbach K, Diniz M (2012) Ocean warming enhances Sweeney MJ, Roper CFE, Mangold KM, Clark MR, Boletzky malformations, premature hatching, metabolic suppres- SV (1992) ‘Larval’ and juvenile cephalopods: a manual sion and oxidative stress in the early life stages of a key- for their identification. Smithson Contrib Zool 513 stone squid. PLoS ONE 7: e38282 Tanimoto Y, Hanawa K, Toba Y, Iwasaka N (1993) Charac- Saito H, Kubodera T (1993) Distribution of ommastrephid teristic variations of sea surface temperature with multi- rhynchoteuthion paralarvae (Mollusca, Cephalopoda) in ple time scales in the North Pacific. J Clim 6: 1153−1160 the Kuroshio Region. In: Okutani T, O’Dor RK, Kubodera Valdés Á, Ornelas-Gatdula E, Dupont A (2013) Color pattern T (eds) Recent advances in fisheries biology. Tokai Uni- variation in a shallow-water species of opisthobranch versity Press, Tokyo, p 457−466 mollusc. Biol Bull 224:35−46 Sakai M, Brunetti NE, Ivanovic ML, Elena B, Sakurai Y Vidal EAG, Haimovici M (1998) Feeding and the possible (2011) Useful techniques for artificial fertilization of the role of the proboscis and mucus cover in the ingestion of ommastrephid squid Illex argentinus. Jpn Agric Res Q microorganisms by rhynchoteuthion paralarvae (Cepha- 45: 301−308 lo poda: Ommastrephidae). Bull Mar Sci 63:305−316 Sakai M, Wakabayashi T, Kato Y, Vijai D, Sakurai Y (2014) Vijai D, Sakai M, Kamei Y, Sakurai Y (2014) Spawning pat- The R/V Kaiyo Maru 2013 cruise report: spawning stock tern of the neon flying squid Ommastrephes bartramii of neon flying squid Ommastrephes bartramii in the (Cephalopoda: Oegopsida) around the Hawaiian Islands. North Pacific Ocean during the 2013 winter. Fisheries Sci Mar 78:511−519 Agency of Japan, Hachinohe Villanueva R, Staaf DJ, Argüelles J, Bozzano A and others Sakurai Y, Young RE, Hirota J, Mangold K, Vecchione M, (2012) A laboratory guide to in vitro fertilization of ocea - Clarke MR, Bower J (1995) Artificial fertilization and nic squids. Aquaculture 342-343: 125−133 development through hatching in the oceanic squids Wakabayashi T, Suzuki N, Sakai M, Ichii T, Chow S (2006) Ommastrephes bartramii and Sthenoteuthis oualaniensis Identification of ommastrephid squid paralarvae collec - (Cephalopoda: Ommastrephidae). Veliger 38: 185−191 ted in northern Hawaiian waters and phylogenetic impli- Sakurai Y, Bower JR, Nakamura Y, Yamamoto S and others cations for the family Ommastrephidae using mtDNA (1996) Effect of temperature on development and sur- analysis. Fish Sci 72: 494−502 vival of Todarodes pacificus embryos and paralarvae. Watanabe K, Sakurai Y, Segawa S, Okutani T (1996) Devel- Am Malacol Bull 13:89−95 opment of the ommastrephid squid Todarodes pacificus, Sakurai Y, Kiyofuji H, Saitoh S, Goto T, Hiyama Y (2000) from fertilized egg to rhynchoteuthion paralarva. Am Changes in inferred spawning sites of Todarodes pacifi- Malacol Bull 13:73−88 cus (Cephalopoda: Ommastrephidae) due to changing Wood S (2006) Generalized additive models: an introduction environmental conditions. ICES J Mar Sci 57: 24−30 with R. CRC Press, Boca Raton, FL Seki MP, Polovina JJ, Kobayashi DR, Bidigare RR, Mitchum Xavier JC, Allcock AL, Cherel Y, Lipinski MR and others GT (2002) An oceanographic characterization of sword- (2014) Future challenges in cephalopod research. J Mar fish (Xiphias gladius) longline fishing grounds in the Biol Assoc UK (in press), doi: 10.1017/S0025315414 springtime subtropical North Pacific. Fish Oceanogr 11: 000782 251−266 Yatsu A, Tanaka H, Mori J (1998) Population structure of Sheng J (1999) Correlation between the Pacific / North the neon flying squid, Ommastrephes bartramii, in the American pattern and the eastward propagation of sea North Pacific Ocean. In: Okutani T (ed) Contributed surface temperature anomalies in the North Pacific. papers to International Symposium on Large Pelagic J Geophys Res Atmos 104(D24):30885−30895 Squids. Japan Marine Fishery Resources Research Shigeno S, Kidokoro H, Goto T, Tsuchiya K, Segawa S (2001) Center, Tokyo, p 31−48 158 Mar Ecol Prog Ser 529: 145–158, 2015

➤ Yoo HK, Yamamoto J, Saito T, Sakurai Y (2014) Laboratory nia. Smithson Contrib Zool No 97 observations on the vertical swimming behavior of ➤ Young CM, Bingham BL (1987) Chemical defense and apo - Japanese common squid Todarodes pacificus paralarvae sematic coloration in larvae of the ascidian Ecteinascidia as they ascend into warm surface waters. Fish Sci 80: turbinata. Mar Biol 96:539−544 925−932 ➤ Zeidberg LD, Robison BH (2007) Invasive range expansion Young RE (1972) The systematics and areal distribution of by the Humboldt squid, Dosidicus gigas, in the eastern pelagic cephalopods from the seas off southern Califor- North Pacific. Proc Natl Acad Sci USA 104:12948−12950

Editorial responsibility: Christine Paetzold, Submitted: August 22, 2014; Accepted: March 16, 2015 Oldendorf/Luhe, Germany Proofs received from author(s): May 19, 2015