Experiment 1: Effects of food conditions on molting
Fig. 3 shows the cumulative molting rate of solitary deutonymphs. Under moderate nutrient conditions (control), solitary deutonymphs exhibited a low molting rate (Fig. 3d). The cumulative molting rate on the 10th day was 3 times higher in males (26/99) than in females (10/135). Survival analysis (Cox proportional hazards model) detected a highly significant difference between sexes (no. parameters = 1, d. f. = 1, likelihood ratio Chi-square = 15.3555, P < 0.0001). No individuals died during the 10-day observation period.
However, when food resources were drastically increased, the solitary individuals quickly responded (Fig. 3a~c). More than 50% of individuals molted to adulthood within a few days. Males showed a higher molting rate than females. Survival analysis (Cox proportional hazards model) detected a highly significant difference among the four groups (Fig. 3) (no. parameters = 3, d. f. = 3, likelihood ratio Chi-square = 38.4473, P < 0.0001) and between sexes (no. parameters = 1, d. f. = 1, likelihood ratio Chi-square = 21.3086, P < 0.0001). Even if the control samples (n = 234, 84% of all samples) were excluded from the analysis, the differences were still significant among the three treatments (Fig. 3a~c) (no. parameters = 2, d. f. = 2, likelihood ratio Chi-square = 7.1877, P = 0.0275) and between sexes (no. parameters = 1, d. f. = 1, likelihood ratio Chi-square = 7.1954, P = 0.0073).
Experiment 2: Effect of pairing on molting
Fig. 4 shows the cumulative molting rate of a deutonymph (DN) paired with another individual (two individuals in a vial). In the pairs of two deutonymphs, molting of both individuals was tracked. Thus, “DN♀ + DN♀, n = 38” (Fig. 4a) means that 16 pairs (38 individuals) of female deutonymphs (16 vials) were investigated. In the pairs with one deutonymph and one adult (A), only deutonymphs were tracked so that “DN♀ + A♂, n = 16” (Fig. 4c) means that 16 deutonymphs (16 vials) were examined.
When paired with a deutonymph of the same sex (Fig. 4a), approximately 50% of female deutonymphs and 40% of male deutonymphs finally molted within a month. Survival analysis (Cox proportional hazards model) detected no significant difference in the molting rate between sexes (no. parameters = 1, d. f. = 1, likelihood ratio Chi-square = 0.8426, P = 0.3587).
Out of 18 heterosexual pairs of deutonymphs (Fig. 4b), three females molted within a day but died before male molting. Males of these pairs were excluded from the analysis, so the male sample size decreased to 15. Females and males in almost all pairs molted to adulthood within a few days. The molting rate of females was slightly higher than that of males, but survival analysis (Cox proportional hazards model) detected no significant difference between sexes (no. parameters = 1, d. f. = 1, likelihood ratio Chi-square = 0.3631, P = 0.5468).
When paired with an adult of the opposite sex (Fig. 4c), female deutonymphs immediately molted, but male deutonymphs did not. Survival analysis (Cox proportional hazards model) detected a highly significant difference between sexes (no. parameters = 1, d. f. = 1, likelihood ratio Chi-square = 16.7439, P < 0.0001). The molting rate in these males was as low as that in the solitary condition (ca. 35%; Fig. 4d).
The solitary deutonymphs (control; Fig. 4d) survived for longer than a month without molting. The cumulative molting rate at the end was ca. 35% for both sexes. Survival analysis (Cox proportional hazards model) detected no significant difference between sexes (no. parameters = 1, d. f. = 1, likelihood ratio Chi-square = 0.1048, P = 0.7460).
In the condition in which the deutonymph was alone, males tended to molt earlier than females (protandry; Figs. 3, 4d), whereas in the condition in which the deutonymph was combined with other individuals, females tended to molt earlier than males (protogyny; Fig. 4a-c).
In summary, the pairwise comparison of the maximum cumulative molting rate in all 6 combinations among 4 treatments using Fisher’s exact probability test with Holm‒Bonferroni sequential correction (Fig. 5) detected a sex difference in the response. All female deutonymphs molted in combination with males (irrespective of stage), but in pairs with another female deutonymph, they showed a low molting rate that did not differ from the solitary condition (Fig. 5a). In contrast, the molting rate of male deutonymphs was still low at pairing with female adults (Fig. 5b).