Juvenile hormone, reproduction, and worker behavior in the neotropical social wasp Polistes canadensis

Abstract

Previous studies of the division of labor in colonies of eusocial Hymenoptera (wasps and bees) have led to two hypotheses regarding the evolution of juvenile hormone (JH) involvement. The novel- or single-function hypothesis proposes that the role of JH has changed from an exclusively reproductive function in primitively eusocial species (those lacking morphologically distinct queen and worker castes), to an exclusively behavioral function in highly eusocial societies (those containing morphologically distinct castes). In contrast, the split-function hypothesis proposes that JH originally functioned in the regulation of both reproduction and behavior in ancestral solitary species. Then, when reproductive and brood-care tasks came to be divided between queens and workers, the effects of JH were divided as well, with JH involved in regulation of reproductive maturation of egg-laying queens, and behavioral maturation, manifested as age-correlated changes in worker tasks, of workers. We report experiments designed to test these hypotheses. After documenting age-correlated changes in worker behavior (age polyethism) in the neotropical primitively eusocial wasp Polistes canadensis, we demonstrate that experimental application of the JH analog methoprene accelerates the onset of guarding behavior, an age-correlated task, and increases the number of foraging females; and we demonstrate that JH titers correlate with both ovarian development of queens and task differentiation in workers, as predicted by the split-function hypothesis. These findings support a view of social insect evolution that sees the contrasting worker and queen phenotypes as derived via decoupling of reproductive and brood-care components of the ancestral solitary reproductive physiology.

Fig. 1.

Frequency of task performance by workers in different age groups. There is a statistically significant effect of age on frequency of performance of different tasks (df = 48, X² = 236.65, P < 0.0001) Tasks with different letters show statistically different patterns of performance with age in post hoc comparisons (P < 0.05): A, relatively common in young individuals; B, relatively unchanging across age categories; and C, relatively common in older individuals. The number of tasks observed for each age group is in parentheses. P. Alarm, parasite alarm; Food Ex., food exchange.

Fig. 2.

Age at onset of task performance for five tasks. Number of wasps observed at the first performance of each task are in parentheses. There is a statistically significant effect of task on age of onset (ANOVA, df = 4, F = 19.233, P < 0.0001). Tasks with different letters showed statistically different patterns of onset with age in post hoc comparisons (Tukey's HSD, P < 0.05): A, relatively early onset, including common in newly emerged females; B, relatively early onset but uncommon in newly emerged individuals; and C, relatively late onset. P. Alarm, parasite alarm.

Fig. 3.

Juvenile hormone analog, methoprene, influenced performance of guarding (bite and attack) and foraging behaviors. (A) Treatment group wasps were overrepresented in the group of wasps initiated guarding and foraging behaviors. *, Statistically significant differences (P < 0.05) in Fisher's exact tests for proportions of workers initiating the behaviors. The n is given in the bars. (B) Bite or attack behavior was initiated at younger ages by wasps in the treatment group (control, age at first bite or attack ± SE, 9 ± 0.59 days; treatment, 5.5 ± 0.27; n_control = 8, n_treatment = 8, Wilcoxon signed-rank test, z = -2.536, P < 0.02).

Fig. 4.

Juvenile hormone titers of workers and gynes. Gynes were compared to all workers and had an order of magnitude higher JH titers (SEs for gynes given in text). Worker groups, 1-day-old, nurse, forager, and guard were compared among themselves. Groups labeled “a” were not significantly different from each other. The group labeled “b” was significantly different from all others. The n for each group is indicated at the base of the bars.

Fig. 5.

Correlation of juvenile hormone titers and ovary development in workers (n = 18) and gynes (n = 11). Additional statistics are in the text.