Tragedy of the commons in Melipona bees revisited

Abstract

Melipona stingless bees display a paradoxical overproduction of queens, which are later eliminated by nest-mate workers. Mechanistically, it was suggested that the monoterpenoid geraniol deposited into newly provisioned cells by adult bees would cause larvae to develop into queens in Melipona beecheii. This system could be evolutionarily stable if many of these new queens were to leave the nest and parasitize other genetically unrelated colonies nearby, as was shown to occur in a congeneric species. Here, we use microsatellite markers to test whether queen overproduction could be a strategy by which adult workers control the caste fate of the developing larvae to export copies of their own genes to the rest of the population via queen parasitism in M. beecheii. In addition, we re-examined whether artificially increasing the levels of geraniol indeed caused larvae to develop as queens rather than workers. Contrary to our prediction, we found no evidence for queen parasitism in M. beecheii and observed no effect of geraniol on the rearing of new queens. Together, these results support the original 'tragedy of the commons' hypothesis for queen overproduction in Melipona bees, where individual larvae selfishly bias their development towards the queen pathway according to their best evolutionary interests.

Keywords: Melipona; caste fate conflict; social insects; social parasitism; tragedy of the commons.

Figure 1.

Egg-laying queens are always replaced by own offspring in M. beecheii. The full likelihood maternal parentage analysis using five microsatellite markers shows that all replacing queens (blue diamonds) were assigned as daughters of the previous egg-laying queen with a probability of 1. Photos by Jorge Ramirez Pech.

Figure 2.

The terpenoid geraniol has no effect on caste development in M. beecheii. A binomial GLMM shows no significant differences in the probability of queens being reared in geraniol-treated versus control cells in the stingless bee M. beecheii (n = 342 for treatment and n = 308 for control). In fact, there is a slight, though nonsignificant, trend in the opposite direction.