Song diversity predicts the viability of fragmented bird populations.

In the global scenario of increasing habitat fragmentation, finding appropriate indicators of population viability is a priority for conservation. We explored the potential of learned behaviours, specifically acoustic signals, to predict the persistence over time of fragmented bird populations. We found an association between male song diversity and the annual rate of population change, population productivity and population size, resulting in birds singing poor repertoires in populations more prone to extinction. This is the first demonstration that population viability can be predicted by a cultural trait (acquired via social learning). Our results emphasise that cultural attributes can reflect not only individual-level characteristics, but also the emergent population-level properties. This opens the way to the study of animal cultural diversity in the increasingly common human-altered landscapes.