Sampling. H. cf. fasciata were sampled from an artificial fish reef located at a depth of approximately 10 m in Nagasaki (Nagasaki Prefecture, Japan). Two experiments were performed in this study. Sixteen individuals were used in Experiment 1, including 12 individuals collected by scuba divers on September 27, 2019, and December 9, 2020, two collected by fishermen, and two that were provided by other researchers (both collected in Nagasaki Prefecture). In Experiment 2, six individuals collected from the same artificial fish reef on September 13, 2021, and November 17, 2021, were used. Each octopus was housed individually in a small rearing case (W 175 × D 105 × H 105 mm) and submerged in a 300-L trough filled with artificial seawater (salinity: 33‰, water temperature: 20°C).
Experiment 1. We examined whether exposure to predators in an aquarium over a 3-day period reduced TTX concentrations in the arms of H. cf. fasciata. The experiment was conducted in a 90-cm-wide glass aquarium (W 900 mm × D 300 mm × H 360 mm) divided into two sections of 30 and 60 cm each by a clear acrylic plate with small holes. Before the experiment commenced, TTX was orally administered to 16 H. cf. fasciata to sufficiently increase TTX levels in the body of all individuals and to compensate for any individuals with low TTX levels at the time of collection. If the TTX level was too low, then changes under the experimental conditions might be below the level of detection. TTX was administered by injecting crab surimi mixed with TTX into frozen crabs with a carapace length of 1–2 cm, and all octopuses were fed the TTX-injected crabs until they stopped eating. The final amount of TTX administered ranged from 60 to 150 µg, depending on the number of crabs the octopus has eaten. In confirming the level of TTX in the muscle and skin before the experiment, all octopuses were anesthetized 24 h after TTX administration with 1% ethanol seawater (Andrews and Tansey 1981; Ikeda et al. 2009), and tissue samples from the arm tips (approximately 0.1 g) were collected to measure TTX levels. In previous research, TTX levels in the arm and mantle were not markedly different (Yamate et al. 2021). Therefore, TTX levels in the arm may reflect TTX levels in the muscle and skin of the whole body. Once the octopuses had recovered from the anesthesia, they were allocated to the predator exposure group (n = 8) or the control group (n = 8). In the predator exposure group, a moray eel (Gymnothorax kidako) was introduced into the 60-cm compartment of the experimental tank, and the octopus was exposed to this predator for 3 days. In the control group, the octopus was kept for 3 days without a predator in the 60-cm compartment. At the end of the experiment, all octopuses were euthanized in 5% ethanol seawater (Andrews and Tansey 1981; Ikeda et al. 2009), and a portion of the arm tissue (approximately 0.1 g) was collected for TTX extraction.
Experiment 2. This experiment aimed to investigate the presence of TTX in the mucus on the body surface after predator exposure. It was conducted in an experimental tank (W 450 mm × D 300 mm × H 300 mm) at a depth of 25 cm. The two experimental tanks of the system described in Experiment 1 were prepared per individual, one for predator exposure and the other for predator nonexposure (control). The same moray eels were used as predators in both experiments. Each octopus was housed in a small case (W 55 mm × D 45 mm × H 90 mm) made of clear acrylic plates with small holes, and the lid was closed with Parafilm to prevent them from escaping. Then, the case was hung by a string and suspended in the nonexposure tank for 10 min as the acclimation time and withdrawn from the tank after an additional 10 min. Next, the case was placed in a bath filled with seawater to a depth of approximately 2 cm. Then, the case was opened, and mucus was collected from the body surface of the octopus, with only the dorsal mantle of the octopus exposed out of the seawater. The mucus was collected using gauze (1 cm × 1 cm) held by tweezers, which was slowly rubbed over 1 cm of the dorsal mantle of the octopus before turning the gauze over and rubbing the mantle again in the same manner. After collecting the mucus, the lid of the case was closed and sealed with Parafilm, and then the case was suspended in the predator exposure tank for 10 min for acclimation. Thereafter, a moray eel was introduced as a predator and exposed to the octopus for 10 min. Then, the case was pulled up again, and mucus was collected as previously described. In this experiment, the octopus may unintentionally secrete TTX because of contact stimulation by mucus collection rather than the presence of a predator. Therefore, after the first predator presentation, the series of steps from the nonexposure condition to the predator exposure condition was repeated. If no TTX was detected in the second nonexposure condition and TTX was detected in the predator presentation condition, then TTX was secreted by the octopus in response to the presence of the predator rather than in response to the contact stimuli associated with mucus collection.
TTX Analysis. TTX was extracted by adding 300 µL of 0.1% acetic acid per 0.1 g of arm tissue collected in Experiment 1 and 200 µL of 0.1% acetic acid to each mucus sample collected in Experiment 2. The mixture was homogenized by using an ultrasonic crusher and then heated for 10 min at 100°C in a heating block (CTU-Neo, Taitec, Aichi, Japan). After cooling the sample, it was centrifuged at 13,000 × g for 15 min. Then, the supernatant was filtered through a filter (FILTSTAR Syringe Filter Hydrophilic Nylon, Hawach Scientific, Xi’an City, China), transferred to a vial, and stored at − 30°C until assayed. The amount of TTX in each sample was quantified using liquid chromatography mass spectrometry (LC/MS, Online Resource 1) system. The limit of detection and the limit of quantification were 0.0009 nmol/mL (0.003 nmol/g tissue; signal-to-noise ratio [S/N] = 3) and 0.003 nmol/mL (0.09 nmol/g tissue; S/N = 10), respectively.
Statistical Analysis. In Experiment 1, multiple regression analysis was conducted using a generalized linear mixed model (GLMM) with TTX concentration in the arm tissue as the response variable, the presence/absence of predator exposure and pre- and post-experiment as explanatory variables, and individual ID as the random effect. P values were calculated using likelihood ratio tests to determine whether the presence or absence of predator presentation was associated with pre- and post-experiment.
In Experiment 2, the TTX concentration in the mucus was assumed to be zero if it was below the limit of detection. Multiple regression analysis was conducted using a GLMM with TTX concentration in mucus as the response variable, the presence or absence of predator exposure and the experimental cycle indicating the first (none, 1; presence, 1) and second (none, 2; presence, 2) presentation conditions as explanatory variables, and individual ID as the random effect. P values were calculated by performing likelihood ratio tests for significant effects of predator presentation and experimental cycle.