In many animals, a phenomenon is often observed in which behaviour depends on population density and many individuals within the group synchronise their state of behaviour to some extent, and theoretical studies have suggested that this synchronisation phenomenon is adaptive for predation avoidance. Moreover, death-feigning behaviour (DF) has been observed as an anti-predator strategy in many animals. There are large individual differences in the duration of DF, and the optimal duration of DF often varies depending on the situation. Therefore, although it is expected that prey may synchronise with others around them for an optimal DF duration, there are few experimental studies testing this hypothesis. This study investigated whether DF duration varies with and without other individuals, and whether it synchronised with the DF duration of other individuals, in the red flour beetle Tribolium castaneum. This study used populations with genetically longer (L-population) and shorter (S-population) DF duration and measured DF duration when maintained alone and cohabitated with individuals from the L- and S-populations, respectively. The results showed that the DF duration of individuals living alone increased significantly compared to pre-treatment. Moreover, individuals that cohabitated with S populations were significantly shorter after cohabitation, but the presence of the L population did not cause any changes in how individuals synchronised their activities. When many individuals had shorter DF durations, DF was synchronised towards shorter durations. This is the first study to illustrate the synchronisation of anti-predator behaviour in terms of DF behaviour.

Insect and culture

A stock culture of T. castaneum, in laboratories for more than 30 years, was used in this study. The beetles were reared on a mixture of wholemeal enriched with yeast in an incubator maintained at 25℃ under a 16:8 h (light: dark) photoperiod (lights on at 7:00; lights off at 23:00). Distinct morphological characters of the abdomen were used in sex determination of beetles at the pupal stage. Then, single-sex groups were stored in a dish (diameter 90 mm, height 15 mm) until experiments were performed. The focal individuals in each treatment were randomly collected from this stock culture.

Selected populations for the duration of death feigning

L-population with long DF duration and S-population with short DF duration in T. castaneum were established by a previous study (Miyatake et al., 2004), and these selection regimes have been maintained for over 30 generations (Matsumura and Miyatake, 2018). More details of this selection were described in previous studies (Matsumura and Miyatake, 2018; Miyatake et al., 2004). Although approximately 20 years have passed since the selected populations were established, large differences in DF duration between the L- and S-populations have been maintained (Fig. 1).

Experimental design

Experimental design of this study is shown in Figure 2. Firstly, I measured DF duration (phase Ⅰ) according to a previously described method (Miyatake et al., 2004). Virgin beetles (21–35 days old) were used. The beetle body was stimulated by a woody stick. If the beetle showed immobility, the duration of this behaviour (until the beetle moved) was recorded by stopwatch. If the beetle did not show DF behaviour by stimuli, the stimulus was repeated up to three times for each beetle.

On the next day of measurements of DF behaviour, each beetle from the stock culture was put into a plastic container (height 15 mm, width 10 mm) with food, and three treatments were performed (alone; male: n = 24, female: n = 24, cohabitation with L; male: n = 38, female: n = 48, cohabitation with S; male: n = 38, female: n = 48). In the alone treatment, each beetle kept alone for an additional day. In the cohabitation treatment, a beetle (focal beetle) was cohabitated with five beetles from L- or S-populations in the plastic container. There was no significant difference in DF duration among the treatments at phase Ⅰ. To identify focal and cohabitant beetles, the elytra of T. castaneum was marked with a paint marker following methods described in previous studies (Matsumura and Miyatake, 2015; Matsumura et al., 2021). To eliminate the effect of marking on results, I marked cohabitants rather than focal beetles. All containers were kept for 30 days in an incubator maintained at 25℃ and 16:8 h photoperiod. Each container was filled with 1.0 g of food, enough for six beetles to live for one month. To avoid influences of copulation, beetles were separated by sex in each treatment. After treatment, measurements of DF duration for focal beetles were repeated (phase Ⅱ).

Statistical analysis

To test the effects of treatment and cohabitated beetles on the DF duration, generalized linear mixed models (GLMMs) with gamma distribution (Ishikawa et al., 2023; Matsumura and Miyatake, 2023), phase, treatment, sex, and interactions of these factors as explanatory variables, beetle ID as a random effect were used for all DF duration data. The GLMM with the same model was also used to analyze the DF frequency (binary data of whether or not shown DF). I used chi-square test for each significance test. When a significant effect of interaction among factors was found, multiple comparisons using Bonferroni correction were performed. All analyses were conducted using R version 4.1.0 (RCoreTeam, 2017) and statistical packages lme4 (Bates et al., 2015) and car (Fox and Weisberg, 2018).