Numerical sex ratio
In each population, the total number of male individuals was higher than that of female individuals, regardless of caste (binomial tests for Okinawa alates, total female:male individuals = 2,427:3,468, p < 0.0001; for Okinawa nymphs, 9,391:14,305, p < 0.0001; for Okinawa pseudergates, 12,362:21,687, p < 0.0001; for Okinawa soldiers, 1,126:1,984, p < 0.0001, for Ishigaki/Iriomote alates, 2,951:3,144, p = 0.0139; for Ishigaki/Iriomote nymphs, 8,143:8,779, p < 0.0001; for Ishigaki/Iriomote pseudergates, 8,365:13,226, p < 0.0001; for Ishigaki/Iriomote soldiers, 1,126:1,984, p < 0.0001, and for Yonaguni alates, 2,951:3,144, p = 0.0139; for Yonaguni nymphs, 8,143:8,779, p < 0.0001; for Yonaguni pseudergates, 8,365:13,226, p < 0.0001; for Yonaguni soldiers, 1,126:1,984, p < 0.0001; Supplementary Table 1). The proportion of males in each colony was clearly biased towards males in all of the castes (Fig. 1b; Wald tests in the glm analysis with sequential Bonferroni all deviated from the equal ratio, p < 0.0001), except for marginally male-biased alates in Ishigaki/Iriomote (z = 2.4717, p = 0.0403) and Yonaguni (z = 2.4018, p = 0.0326), and no bias was detected in Yonaguni soldiers (z = − 0.3303, p = 0.7412). Compared to the nymphs, the pseudergates showed more male-biased sex ratios (Fig. 1b; glmm analysis for the data of each population with sequential Bonferroni, Okinawa, z = 10.6800, p < 0.0001; Ishigaki/Iriomote, z = 17.6840, p < 0.0001; Yonaguni, z = 4.4210, p < 0.0001). In the Okinawa and Ishigaki/Iriomote populations, the pseudergates showed a similar sex ratio to that of the soldiers, but in the Yonaguni population, the pseudergates were more male-biased than the soldiers (Fig. 1; glmm analysis with sequential Bonferroni, Okinawa, z = 0.9411, p = 0.6940; Ishigaki/Iriomote, z = − 2.2610, p = 0.1185; Yonaguni, z = − 2.8640, p = 0.0251). In all populations, the nymphs showed a similar sex ratio to that of the alates (Fig. 1; glmm analysis with sequential Bonferroni, Okinawa, z = 1.1560, p = 0.7440; Ishigaki/Iriomote, z = − 0.6086, p = 0.5430; Yonaguni, z = 1.5780, p = 0.4600). No significant effect of the colony size on the proportion of males was detected in any caste of any population (Supplementary Fig. 1). Note that the data included two colonies with poor food availability (C171 and C172), even in which males were more abundant than females in all castes (Supplementary Table 1).
Relationship between caste developmental pathway and numerical sex ratios in the three populations. (a) The caste developmental pathway of Neotermes sugioi 26. E egg, L larva, Pse pseudergates, PS pre-soldier, S soldier, N nymph, A alate. (b) Comparisons of numerical sex ratios between castes. In each box, the cross symbol indicates the mean, bar in the box indicates the median, and the box top and bottom indicate the first and third quartiles, respectively. The asterisk next to each bar between boxes indicates p < 0.05 (glmm analysis with sequential Bonferroni), and “n.s.” indicates “not significant”, i.e., p ≥ 0.05. The gray boxes indicate deviations from equal sex ratios at p < 0.05 (Wald tests in glm analysis with sequential Bonferroni), and the white boxes indicate p ≥ 0.05.
Measurement of alate body size and sex allocation
In the Okinawa and Ishigaki/Iriomote populations, the head width of male alates was 0.8% and 0.7% larger than that of female alates, respectively (Supplementary Table 2). No sexual differences in dry weight were found between these two populations (Supplementary Table 2). In the Yonaguni population, no sexual difference was detected in head width, but the dry weight of males was 3.3% lower than that of females (Supplementary Table 2).
In the Okinawa population, alate biomass as an index of sex allocation was calculated by multiplying the mean dry weight by the number of alates for each sex, and it was biased toward males (proportion of male biomass = 0. 549 ± 0.027 (mean ± SE), paired t-test, n = 18, t = 2.445, p = 0.026; Fig. 2a). Such biased sex allocation was not detected in the Ishigaki/Iriomote (proportion of male biomass = 0. 538 ± 0.036 (mean ± SE), paired t-test, n = 9, t = 0.079, p = 0.939), and Yonaguni populations (proportion of male biomass = 0. 528 ± 0.018 (mean ± SE), paired t-test, n = 11, t = 1.835, p = 0.097). In all the populations, both male and female alate biomass increased linearly with the increase in total alate biomass in the colonies (Fig. 2b; Supplementary Table 3).
Characteristics of sex allocation and results of the test for intrasexual competition among siblings (ICS) and sex-asymmetric longevity corresponding to sex-asymmetric reproductive value (SRV) in the three populations. (a) Histogram of the proportion of biomass of male alates in the colonies of the three populations. (b) Plot analysis of the model of ICS. (c) Pattern-frequency of reproductives in the three populations. Solid lines (males) and dashed lines (females) in (b) show the correlations between the total biomass of alates in a colony and the biomass of alates of each sex in a colony, and numerals on the graph (c) indicate the number of colonies. CI confidence interval.
Estimating the lifespans of kings and queens
Out of a total of 364 colonies sampled from the Okinawa population, 56 had a queen but no king, whereas 15 had a king but no queen. Out of a total of 44 colonies from the Ishigaki/Iriomote population, ten had a queen but no king, while three had a king but no queen. Out of a total of 34 colonies from the Yonaguni population, five had a queen but no king, while no colonies with a king but no queen were found (Fig. 2c). Thus, in all the populations, colonies with a queen but no king tended to be more frequent than those with a king but no queen (binomial tests, p < 0.0001 for Okinawa, p = 0.0923 for Ishigaki/Iriomote, p = 0.0625 for Yonaguni). There was no difference in the ratio of these two colony types among the populations (Fig. 2c; chi-squared test, χ2 = 1.369, d.f. = 2, p = 0.504).
Source: Ecology - nature.com
