In a study that has just been published in Frontiers in Psychology, we show that beetles (Tenebrio molitor) use numerical cues to evaluate the number of rival males they encounter before mating. Due to their sophisticated sperm competition mechanisms, T. molitor males are in a lot of (selective) pressure to asses the number of males around during mating to assess the risk that the female they are mating with may re-mate quickly with a second male. In this specie, as in many other insects, males transfer their sperm wrapped up in a package called spermatophore. Once the spermatophore is inside the female, it will begin to inflate by osmosis (like a ballon) until it bursts open and individual sperm are released into the female reproductive tract. This process lasts between 7-10 minutes, after which sperm will be stored in the female spermatheca (an organ specialised in storing sperm) until used to fertilise eggs. That is, if a second male doesn’t prevent it… In this species, the male aedeagus (insect jargon for ‘penis’) is covered in spines that will puncture the spermatophore of a rival male, preventing it from inflating and hence completely avoiding sperm release. By this mechanism, called spermatophore inhibition, a male is thus in risk of loosing all his progeny if the female it has mated with re-mates with a second male before its own sperm has been released (< 7-10 minutes).
Because the risk of this happening directly depends on the average encounter rate between males and females, we thought this was a perfect biological system to test whether beetles are spontaneously capable of assessing quantities based on numerical cues. We set out on this task by staging matings between virgin females and virgin males in which we varied the number of rival males the experimental male had access to immediately preceding mating (1-4 rival males). Rival males were presented sequentially, and we controlled for continuous cues by ensuring that males in all treatments were exposed to the same amount of male–male contact. Briefly, in the ‘one male’ treatment the experimental male was sequentially presented with the same rival male four consecutive times (i.e. each presentation consisting of 3 minutes of male-male contact followed by 2 minutes of isolation) during the 20 minutes immediately before being presented with a virgin female. In the ‘two males’ treatment, experimental males were first presented with male A, then male B, then male A again, and then male B again, following the same procedure as in the ‘one male’ treatment; and so on with the ‘three males’ (male A, male B, male C, male A) and the ‘four males’ (male A, male B, male C, male D) treatments. Crucially, the only difference between the four treatments was the number of different males presented, which varied from 1 to 4. We found that males triplicated the time they devoted to mate guarding females in the ‘four males’ treatment, precisely in the treatment simulating an encounter rate reflecting a real risk of suffering from spermatophore inhibition! Since males could not rely on continuous cues we conclude that they kept a running tally of the number of individuals they encountered serially, which meets the requirements of the basic ordinality and cardinality principles of proto-counting. Our results thus offer good evidence of “true” numerosity estimation or quantity estimation and, along with recent studies in honey-bees, suggest that vertebrates and invertebrates share similar core systems of non-verbal numerical representation. You can download the pdf of this article here.