Guppies in large groups cooperate more frequently in an experimental test of the group size paradox

Abstract

The volunteer's dilemma, in which a single individual is required to produce a public good, predicts that individuals in larger groups will cooperate less frequently. Mechanistically, this could result from trade-offs between costs associated with volunteering and costs incurred if the public good is not produced (nobody volunteers). During predator inspection, one major contributor to the cost of volunteering is likely increased probability of predation; however, a predator also poses a risk to all individuals if nobody inspects. We tested the prediction that guppies in larger groups will inspect a predator less than those in smaller groups. We also predicted that individuals in larger groups would perceive less threat from the predator stimulus because of the protective benefits of larger groups (e.g. dilution). Contrary to prediction, we found that individuals in large groups inspected more frequently than those in smaller groups, but (as predicted) spent less time in refuges. There was evidence that individuals in intermediate-sized groups made fewest inspections and spent most time in refuges, suggesting that any link between group size, risk and cooperation is not driven by simple dilution. Extensions of theoretical models that capture these dynamics will likely be broadly applicable to risky cooperative behaviour.

Keywords: Trinidadian guppies; anti-predator behaviour; cooperation; group size; volunteer's dilemma.

Figure 1.

Individual inspection frequency and behaviour. The effect of experimental group size on an individual's: (a) number of inspections; (b) duration of inspections; (c) proportion of time spent inspecting; (d) minimum inspection distance; and (e) mean inspection distance. Large boxes show data for all trials of that experimental group size; small boxes show each trial (colours correspond to trials). We found that individuals in large groups inspected more frequently, but that no other measured aspect of inspection behaviour was affected by experimental group size.

Figure 2.

Individual refuge use. The proportion of time that individuals spent in refuges in different experimental group sizes ‘before’ (left) and ‘during’ (right) the appearance of the predator stimulus. Large boxes show data for all trials of that experimental group size; small boxes show data for each trial; colours represent different trials. Individuals in larger groups spent less time in refuges while the stimulus was visible than individuals in small or intermediate-sized groups, and intermediate-size groups spent the most time in refuges.

Figure 3.

Number of inspectors per inspection for each experimental group size. The large boxes show aggregated data for each group size. The small boxes show data for each trial; colours correspond to trials. Number of inspectors in each inspection increases with group size.