Ephyrae of A. aurita were obtained from stock cultures of polyps derived from matured medusae in the Inland Sea of Japan. Planulae of A. aurita were collected from the oral arms of matured medusae and placed in plastic containers (diameter: 15 cm, depth: 6.5 cm) containing ca. 1 l of filtered (0.2 µm) seawater of salinity 32. These planulae were incubated at 25˚C in darkness to allow them to settle on the walls of containers. After transformation to polyps, they were fed ad libitum with newly hatched Artemia sp. nauplii once or twice weekly, followed by replacement of the seawater. These polyps were maintained as stock cultures.

The stock-cultured A. aurita polyps maintained at ca. 25˚C were transferred to 13˚C to induce strobilation and release of ephyrae. Newly released ephyrae were kept starved in plastic containers containing ca. 1 l of filtered seawater (salinity: 32) and were maintained at three different temperatures (i.e., 9˚C, 12˚C and 15˚C) in darkness for various periods, ranging from 0 to 60 d, prior to respiration measurement experiments. These three temperatures (i.e., 9˚C, 12˚C and 15˚C) represent the lowest, medium and highest temperatures in the winter in the Inland Sea of Japan. The aerated filtered seawater used in the respiration experiments was produced by air stone aeration for ca. 1 h. In each experiment, 20 to 50 ephyrae were pipetted into a small vial (ca. 10 ml volume) in which seawater was completely replaced with aerated seawater and then transferred into a dissolved oxygen (DO) bottle (60.0 ml volume) containing the same aerated seawater. Three bottles were prepared for either respiration measurement or to serve as a control, and they were placed in dark incubators for 72 h with periodical (4 - 8 h intervals) inversions of the bottles by hand. The selection of numbers of ephyrae (20 - 50) and incubation duration (72 h) was performed based on preliminary determination of the respiration rate of an ephyra so that the consumption of DO would be larger than the precision level (0.02 mg O₂ l⁻¹) of a fluorescence-based DO meter (WTW, Multi 3410 with FDO 925 probe). The DO before and after the experiment was measured, and the ephyrae used in the experiment were observed under a stereomicroscope to check their survival. At least 10 of them were photographed by a digital camera to measure their size.

Total of 60 newly released A. aurita ephyrae originating from stock-cultured polyps were prepared, and they were individually placed in wells of 6-well polystyrene culture plates containing 10 ml of filtered seawater of salinity 32. Each lot consisting of 20 ephyrae was transferred to three temperatures (i.e. 9˚C, 12˚C and 15˚C), and kept starved for up to 60 d. The seawater in the wells was replaced twice weekly. Newly released ephyrae at each temperature were photographed under a stereomicroscope to measure their size, and used for the experiment to measure their pulsation rates. Pulsation rate of each ephyra was measured at every 10-day interval.

For the pulsation determination, an ephyra was transferred to a 1-liter plastic bottle filled with seawater of respective temperature, and then its pulsations were counted with a hand-held counter for one min for three successive min. The experiment was conducted under dim light. The mean pulsation rate was determined from 3 counts for each of specimens.

The respiration rates and pulsation rates in different treatments were analyzed by one-way or two-way ANOVA (SPSS 26.0 software). If the overall ANOVA results were significant (P < 0.05), the means were compared using Tukey’s pairwise comparison.

As described in Fu et al. [9] , the size and morphology of ephyrae changed with prolonged duration of starvation, and these changes were more rapid in higher temperature treatments.

The mean (± SD) respiration rates of newly released ephyrae were 0.24 ± 0.07, 0.24 ± 0.03 and 0.19 ± 0.06 µl O₂ ephyra⁻¹ d⁻¹ at 15˚C, 12˚C and 9˚C, respectively (Figure 1), although there was no significant difference between them (one-way

ANOVA, P > 0.05). The respiration rates showed a general trend of decreasing with increasing starvation duration; in the beginning, the decrease was relatively smooth and small, but after 30 to 40 d of starvation, it appeared to be rapid. The rate decreased nearly to or below half of the rate shown by non-starved ephyrae when the starvation was longer than PNR₅₀ (Figure 1). However, the statistical tests failed to show any significant difference in the respiration rates as a function of temperature (two-way ANOVA, P > 0.05), duration of starvation (two-way ANOVA, P > 0.05) or their interactions (two-way ANOVA, P > 0.05) during the first 30 d of starvation. Meanwhile, the respiration rate on Day 0 was significantly higher than the respiration rate on Day 30 (Tukey’s pairwise comparison, P < 0.05). At 9˚C and 12˚C, there was no significant difference in the respiration rates for different durations of starvation (one-way ANOVA, P > 0.05). However, at 15˚C, the duration of starvation did significantly affect the respiration rate between Day 0 and Day 40 (one-way ANOVA, P < 0.05).

The carbon weight-specific respiration rates were calculated by dividing individual respiration rates by individual carbon weights. The carbon weight of A. aurita ephyrae was already reported in a previous study of ours [9] . The mean (± SD) carbon weight-specific respiration rates of newly released ephyrae were 35.6 ± 10.3, 37.9 ± 4.2 and 29.4 ± 9.4 µl O₂ mg C⁻¹ d⁻¹ at 15˚C, 12˚C and 9˚C, respectively (Figure 2), without a significant difference between temperatures (one-way ANOVA, P > 0.05). The specific respiration rates were almost constant over the starvation period up to 40 d (Figure 2), as a statistical test did not show a significant difference in the rates with temperature (two-way ANOVA, P > 0.05), duration of starvation (two-way ANOVA, P > 0.05) or their interactions (two-way ANOVA, P > 0.05). Moreover, at each temperature, one-way ANOVA failed to detect any significant difference in the rates with duration of starvation (P > 0.05).

The mean (±SD) disc diameter of newly released ephyrae used in this experiment was 2.19 ± 0.18 mm. Twenty ephyrae died during the experiment due perhaps to the mechanical damage with pipetting. At the end of the experiment, there were 16, 15 and 15 ephyrae remaining at 9˚C, 12˚C and 15˚C, respectively, and they all looked active.

The mean (±SD) pulsation rate of newly released ephyrae was 18.8 ± 6.3, 20.3 ± 11.2 and 28.2 ± 13.8 beats ephyra⁻¹ min⁻¹ at 9˚C, 12˚C and 15˚C, respectively (Figure 3), with significant difference with temperature (one-way ANOVA, P < 0.05). Over the starvation period up to 50 days, two-way ANOVA showed a significant effect of starvation on pulsation rate (P < 0.01), but did not show any significant effect of temperature (P > 0.05) and their interactions (P > 0.05). Tukey’s test showed that the pulsation rate on Day 0 was significantly lower than that on Day 10, Day 20 and Day 30 (P < 0.01), the same test also showed that the rate on Day 20 was significantly higher than that on Day 0, Day 40 and Day 50 (P < 0.01).

Our study is the first study to measure respiration for newly released A. aurita ephyrae (mean disc diameter (DD): 2.02 ± 0.19 mm). Previous work had measured the respiration for slightly advanced ephyra stages whose DD values were 4.2 mm [11] and 5.1 mm [14] . The respiration rates at 15˚C were calculated to be 0.42 µl O₂ ephyra⁻¹ d⁻¹ [14] and 0.74 µl O₂ ephyra⁻¹ d⁻¹ [11] . Compared to these rates, the respiration rate measured at 15˚C in this experiment (i.e., 0.24 µl O₂ ephyra⁻¹ d⁻¹) was significantly lower.

In jellyfish respiration measurements, Purcell et al. [21] pointed out several factors influencing actual measurements, including 1) acclimation to temperature, 2) feeding conditions of specimens before experiment, 3) volume of respiration chamber, 4) acclimation and incubation times, and 5) activity of animals during experiment, although no standardized protocols have been established. In this study, A. aurita ephyrae were treated with great care before and during the respiration experiments to reduce the errors to as small as possible, and the obtained data had substantially wide variations, resulting in a relatively wide SD for each mean value. These might preclude the statistical analysis from detecting the effects of temperature and starvation duration on the respiration rates at a significant level. However, the general trends were clear: the respiration rates decreased with increasing starvation duration, and the temperature effect was not significant.

In addition, all reported respiration rates of A. aurita ephyrae, including those in this study, were much lower than the respiration rates of other zooplankton taxa having similar carbon contents. Assuming a carbon:dry weight ratio of 0.45 [28] , the respiration rate of crustacean zooplankton, primarily copepods, from the Inland Sea of Japan [29] can be estimated as 2.55 µl O₂ ind⁻¹ d⁻¹ at 15˚C. A similar calculation can be made by interpolation of an equation derived from various zooplankton taxa from the world oceans [30] , which resulted in 1.68 µl O₂ ind⁻¹ d⁻¹. The extremely lower respiration rates in A. aurita ephyrae than in other taxa were also demonstrated when the rates were expressed in terms of dry weight-specific respiration rates (see Table A1, Figure 4). Hence, we conclude that the low metabolism rates per unit body (either dry or carbon) mass are specific to A. aurita ephyrae. This conclusion, however, is contradictory to results in publications by Acuña et al. [31] and Pitt et al. [32] . They have reported that the respiration rates of jellyfish are similar to those of other metazoans (e.g., crustaceans and fish) when scaled by carbon content. Interpolation to 6.6 µg C (carbon

weight of an A. aurita ephyra) using equations presented by Acuña et al. [31] and Pitt et al. [32] gives respiration rates of 2.9 and 2.6 µl O₂ animal⁻¹ d⁻¹, respectively, values much greater than those mentioned above for A. aurita ephyrae.

It is worth mentioning the differences between the measured respiration rates of newly released ephyrae (i.e., 0.24, 0.24 and 0.19 µl O₂ ephyra⁻¹ d⁻¹ at 15˚C, 12˚C and 9˚C, respectively) and those determined indirectly from decreases in carbon content (i.e. 0.40, 0.28 and 0.27 µl O₂ ephyra⁻¹ d⁻¹ at 15˚C, 12˚C and 9˚C, respectively), which was reported by Fu et al. [9] . We suspect that the placement of ephyrae into a small volume of the DO bottles might reduce their swimming activity and thus decrease the respiration rates compared to those obtained by indirect estimation, where ephyrae were allowed to swim freely.

Since the carbon weight-specific respiration rates were stable regardless of starvation duration, the decrease in body carbon weight was responsible for reduced individual-level respiration rates for starved ephyrae. This fact also suggests that the basal metabolism was maintained at a certain fixed level for A. aurita ephyrae so far as their metabolic substrate is available. It is noteworthy that even ephyrae starved almost to PNR₅₀ could maintain these metabolic kinetics.

The mean carbon weight-specific respiration rates of newly released A. aurita ephyrae were converted to dry weight-specific respiration rates by using a dry weight and carbon conversion factor of 0.33 (the measured value in this study) to yield 11.7, 14.5, and 14.6 µl O₂ mg DW⁻¹ d⁻¹ at 9˚C, 12˚C and 15˚C, respectively, for comparison to the rates for planktonic larvae of other marine animals (i.e., molluscs, crustaceans and fish, Figure 4). The dry weight-specific respiration rates ranged from 139.2 to 333.8 µl O₂ mg DW⁻¹ d⁻¹ for molluscs, from 9.6 to 504 µl O₂ mg DW⁻¹ d⁻¹ for crustaceans, and from 39 to 154.1 µl O₂ mg DW⁻¹ d⁻¹ for fish (see Table A1). Compared to these values, the rates of A. aurita ephyrae are, in general, much lower.

Based on the respiration rates, the minimum food requirement can be estimated as

MFR = k × R × RQ/A,

where MFR is the minimum food requirement (µg C g⁻¹ ephyra⁻¹ d⁻¹), k is a constant (0.375 µg C µg⁻¹ O₂), R is the respiration rate (µg O₂ ephyra⁻¹ d⁻¹), RQ is the respiratory quotient (assumed to be 0.8 due to the protein-dominated metabolism, [33] ), and A is the assimilation efficiency (assumed to be 0.8, [18] ). The MFR for a newly released ephyra was calculated to be 0.13, 0.13 and 0.10 µg C ephyra⁻¹ d⁻¹ at 15˚C, 12˚C and 9˚C, respectively, corresponding to 2.0%, 2.0% and 1.6% of the ephyra carbon weight.

As A. aurita ephyrae are tactile predators without any ability to detect prey remotely, swimming and feeding are closely related each other. In this study, the maximum pulsation rate recorded was 49.2 beats min⁻¹, indicating that A. aurita ephyrae are cruising predators, which swim almost continuously. If one pulsation can generate a movement of 1.8 mm, as was observed by Sullivan et al. [34] , the maximum average swimming speed of A. aurita ephyrae would be 8.87 cm∙min⁻¹. The ephyrae cannot chase prey with swimming speed higher than them, but can encounter with swimming prey. However, even in the latter case, the prey with higher swimming speeds can escape from the feeding current generated by ephyra’s disc pulsation [34] and even shake off the manubrium. Therefore, A. aurita ephyrae are expected to capture primarily slow-moving prey, such as fish eggs, barnacle nauplii, veliger larvae and hydromedusae. Copepod nauplii (escape speed: 120 cm∙min⁻¹) and copepodites and adults (escape speed: >300 cm∙min⁻¹) may not be the primary prey for A. aurita ephyrae (Figure 5).

In this study, ephyrae could actively swim at all three temperatures, and their pulsation rates did not differ significantly among the temperatures tested (an exception was for newly released ephyrae), which is contrary to the previous results. Both Mangum et al. [26] and Dillon [27] found that the pulsation rate increased with increasing temperature up to 25˚C, and then decreased with further temperature increase. A similar temperature effect was also found for A. aurita medusae [23] [24] [35] . The difference between this study and the previous ones might be attributed to longer time adaption of ephyrae to experimental temperatures in this study compared to the previous ones, where ephyrae were exposed to experimental temperatures rather suddenly. Over the temperature range at least between 9˚C and 15˚C, where A. aurita ephyrae usually experience in the field, they can swim always actively to seek food and escape from predators.

It was an interesting finding that the pulsation rates of A. aurita ephyrae after 10 and 20 days of starvation were significantly higher than those of ephyrae of newly released. This result indicates that starvation may accelerate pulsation in order to search for more prey. The pulsation rate culminated on 20 d after starvation,

and thereafter the rate decreased. The decrease of pulsation rates together with body size shrinkage may significantly reduce the feeding capability of A. aurita ephyrae. In addition, decrease of pulsation can also cause the reduction of escaping ability from predators, leading to higher mortality of A. aurita ephyrae in the field.