Context‐dependent defences in turtle ants: Resource defensibility and threat level induce dynamic shifts in soldier deployment. (12th July 2017)

• Record Type: Journal Article
• Authors: Powell, Scott; Donaldson‐Matasci, Matina; Woodrow‐Tomizuka, Augustus; Dornhaus, Anna
• Editors: Wilson, Robbie
• Is Part Of: Functional ecology. Volume 31:Number 12(2017)
• Page Start: 2287
• Page End: 2298
• Publication Date: 2017-07-12
• Subjects: bet hedging -- caste -- Cephalotes -- defence deployment -- defence strategy -- defence traits; Ecology -- Periodicals; 574.505
• DOI: 10.1111/1365-2435.12926
• Languages: English
• ISSNs: 0269-8463

Abstract

Abstract: Induced defences involve the dynamic deployment of limited and specialized defensive resources across multiple locations, to maximize organismal defensive function and fitness. They have been studied intensively in plants and solitary animals, but the induced defences of complex animal societies are poorly understood by comparison, despite the coordinated defensive actions of these taxa. Here, we ask whether the level of environmental danger induces shifts in the deployment of limited and morphologically specialized soldiers across multiple nests in colonies of the turtle ant Cephalotes rohweri . Specifically, we test whether less defensible nests induce greater soldier deployment, and whether elevated enemy threat induced further increases in deployment, or reduced deployment consistent with a risk‐limiting strategy. We used colony‐collection data to provide natural ecological context to our experiments, a field experiment to address how nest‐entrance defensibility and soldier number impact defensive performance, and laboratory experiments to test whether differences in nest defensibility and threat level induce dynamic shifts in soldier deployment to new nests. Less defensible nests were lost rapidly in our field experiment, irrespective of soldier number, but soldier deployment significantly increased survivorship of more defensible nests. Concordantly, less defensible nests induced the deployment of more soldiers per nest under low threat in laboratory experiments. Nevertheless, high‐threat conditions revealed a risk‐limiting soldier deployment strategy: with more danger, the number of soldiers per nest was significantly reduced in less defensible nests, as was the overall number of new soldier‐defended nests. Total deployment to new nests was also consistently lower under high threat, dropping from 40% to 30% of all available soldiers across colonies. Induced soldier‐based defences in turtle ants are therefore context‐dependent, and dynamically scaled back at multiple levels when the environment is more dangerous. This dynamic, risk‐limiting strategy is in strong contrast to stable patterns of soldier production in ants, and to typical task‐allocation dynamics in members of the worker caste. Moreover, these findings establish that the evolution of specialized defensive agents can be coupled with sophisticated and inducible deployment strategies in complex social taxa, as we see for organisms at other levels of biological complexity. A plain language summary is available for this article. Abstract : Plain Language Summary