Data-driven analyses of social complexity in bees reveal phenotypic diversification following a major evolutionary transition

Abstract

The evolution of social complexity, from solitary to highly social species, represents one of the most enigmatic phenotypic transitions in the animal kingdom, with insects being a prime example. However, current classification methods for discrete social classes are oversimplified, limiting our ability to quantitatively study the evolutionary changes across the phenotypic space of social complexity. Here, we propose a data-driven approach that is well-suited to overcoming these limitations, allowing us to avoid constraining our inference to a narrow evolutionary trajectory, and to study the full diversity of social phenotypes without condensing social complexity into putative discrete classes. To investigate the evolution of social complexity in bees, we constructed and analyzed a comprehensive dataset encompassing 17 social traits for 77 bee species. We find that corbiculate bees — honey bees, stingless bees, and bumble bees — underwent a major evolutionary transition ∼70 mya, which does not follow the commonly assumed “social ladder” route in phenotypic space. We also show that this major transition was followed by a phase of substantial phenotypic diversification of social complexity. In contrast, other bee lineages display a continuum of social complexity, ranging from solitary to simple societies. Bee evolution, therefore, provides a unique demonstration of a macroevolutionary process in which a major transition removed biological constraints and opened novel evolutionary opportunities, driving the exploration of the space of social phenotypes. Our approach can be readily applied to illuminate the evolution of social complexity in additional animal groups.