Male behavioural maturation rate responds to selection on pollen hoarding in honeybees

Abstract

Division of labour in social insect colonies relies on behavioural functional differentiation (specialization) of individuals with similar genomes. However, individual behavioural traits do not evolve independently of each other (behavioural syndromes). A prime example is the suite of behavioural differences in honeybee workers that has evolved in response to bidirectional selection on pollen hoarding of honeybee colonies (pollen-hoarding syndrome). More generally, these differences reflect functional differentiation between nectar and pollen foragers. We demonstrate here that this pollen-hoarding syndrome extends to drones. Similar to what has been shown in workers, drones from the high-pollen-hoarding strain had a higher locomotion activity after emergence, and they initiated flight earlier than did males derived from the low-pollen-hoarding strain, with hybrids intermediate. However, these two behavioural traits were unlinked at the individual level. We also found that social environment (the colony) affects the age at which drones initiate flight. The indirect selection responses of male behaviour suggest that male and worker evolution are not independent and may constrain each other's evolution. Furthermore, we identified three distinct peaks in the probability of flight initiation over the course of the experiment and a decreased phenotypic variability in the 'hybrid' males, contrary to quantitative genetic expectations.

Figure 1

Host colonies for measuring the age of first flight in honeybee drones were housed in this modified bee hive, which allowed monitoring of flight activity by a daily census of the bottom box. ➀Drones were introduced into the main hive that contained the brood nest to mature. ➁Once mature, drones left the hive through wire-mesh one-way exits. ➂Upon return to the hive, drones were ‘forced’ to enter through the main entrance into the bottom hive body, where a census was performed at the end of each day.

Figure 2

Onset of flight activity for honeybee drones in the Hybrid, High and Low groups in experiment 1.

Figure 3

Effects of social environment on age at first flight in honeybee drones in the Hybrid, High and Low groups in experiment 1. Means are given with 95% CI.

Figure 4

Age-specific probability of flight initiation in Low (N = 1229), Hybrid (N = 2297) and High (N = 1340) males in experiment 1. Data were smoothed (moving average, 3-day window) and truncated at 21 days of age because the low remaining sample size (fewer than 15 individuals in each group) made the estimates unreliable.

Figure 5

Onset of flight activity for honeybee drones in the Hybrid, High and Low groups in experiment 2. The absence of any drones initiating flight on days 9 (19 May 2002) and 10 (20 May 2002) was due to rain.

Figure 6

Differences in initial locomotor activity of honeybee drones in a walking assay in the light (Light: ○) and the dark (Dark: ▵). Values are means ± 95% confidence interval.