Evolutionary loss of threat display in more social species: phylogenetic comparisons, natural interactions in the wild, and experiments with models

In: Behaviour
View More View Less
  • 1 Department of Biological Sciences, 78 College Street, Dartmouth College, Hanover, NH 03755, USA
  • | 2 Graduate Program in Ecology, Evolution, Environment, and Society, Dartmouth College, Hanover, NH 03755, USA
Download Citation Get Permissions

Access options

Get access to the full article by using one of the access options below.

Institutional Login

Log in with Open Athens, Shibboleth, or your institutional credentials

Login via Institution


Buy instant access (PDF download and unlimited online access):



Fitness can be increased dramatically by communication. So why, given the benefits of communication, would displays ever be completely lost evolutionarily? Threat displays, in particular, are relatively cheap to produce and are precursors of attack, so should be especially hard for both senders and receivers to lose completely. Here we explore an evolutionary transition in sociality, testing whether complete evolutionary loss of threat display has occurred in ‘more social’ hermit crab species, which interact more frequently with conspecifics. First, we synthesised literature and observations on the presence versus absence of threat display across hermit crab species, mapping this information onto a phylogenetic tree. We found that all ‘less social’ species — marine and terrestrial — produce threat displays, consistent with threat display being the ancestral state. But ‘more social’ terrestrial species, which are highly derived, do not produce a threat display, suggesting an evolutionary loss. Next, we contrasted natural interactions in the wild within a less social species (Pagurus bernhardus) versus within a more social species (Coenobita compressus), finding that the less social species, despite a lower rate of social encounter, had a higher rate of display per encounter (24%). In contrast, the more social species’ rate of display per encounter was negligible (<1%), effectively indicating a loss in production. Finally, we experimentally reanimated threat display in the more social species, using postured models to test whether receivers retained any responsiveness to threat display. Starkly, receivers were not deterred by threat display, showing equal responsiveness across both threat and non-threat models, regardless of whether the models were stationary or dynamically moving. Our results thus reveal a case of complete collapse of communication involving threat display, implicating the social environment in this loss. In more social species, an extreme dependence on conspecific-derived shells likely drove a ‘desperado effect’, with threat displays being lost because they could not stop others from pursuing these valuable resources.

Content Metrics

All Time Past Year Past 30 Days
Abstract Views 283 113 1
Full Text Views 50 14 2
PDF Views & Downloads 45 22 5