The Evolution of the Algorithms for Collective Behavior

Abstract

Collective behavior is the outcome of a network of local interactions. Here, I consider collective behavior as the result of algorithms that have evolved to operate in response to a particular environment and physiological context. I discuss how algorithms are shaped by the costs of operating under the constraints that the environment imposes, the extent to which the environment is stable, and the distribution, in space and time, of resources. I suggest that a focus on the dynamics of the environment may provide new hypotheses for elucidating the algorithms that produce the collective behavior of cellular systems.

The Role of Information Acquisition and Processing in Decision‐Making by Individual within Insects Colonies

“…These information allow the insects to evaluate the nest and environmental conditions and to decide the adequate response to the dynamics of the environment and colony needs. Gordon () defines the collective behaviour as the outcome of a network of local interactions and uses a computational perspective to argue about the evolution of the collective behaviour. According to Gordon (), collective behaviour is a phenotype that evolves to fit the dynamics of a particular environment and three factors are important in the relationship between environment and collective behaviour: (i) Operating costs : cost, in terms of energy, to perform a specific task under environmental constraints; (ii) Stability : indicates how stable the environment is; and (iii) Distribution of resources : indicates the uniformity and concentration of the resources.…”

“…The interest of researches in understanding the complex behaviours of social insects is not new. Several researches have been presented with the objective of understanding the physiological, neural and hormonal features of social insects and how they contribute to the formation of complex behavioural patterns (Greene and Gordon, 2003;Gadau and Fewell, 2009;LeBoeuf et al, 2013;Gordon, 2016b;Feinerman and Korman, 2017). The main points of these researches are as follows: (i) Understand how the individuals use or combine different types of information (Grüter and Leadbeater, 2014); (ii) Understand how the individual behaviour of a social insect affects the group-level decision-making and; (iii) Understand how the ability of solving complex problems emerges from the individual behaviours (Feinerman and Korman, 2017).…”

Ant collective behavior is heritable and shaped by selection

“…However, the degree to which collective behaviors are heritable and how genetic variation contributes to population-level variation in individual and collective behaviors remains largely unknown. Furthermore, it is often assumed that, like individual behavior and other individual-level traits, collective behavior and other group-level traits are strongly shaped by natural selection (Gordon 2013(Gordon , 2016. However, very little is actually known about how natural selection acts on collective behaviors or group-level traits more generally (Wright et al 2019).…”

A distributed algorithm to maintain and repair the trail networks of arboreal ants

“…This suggests that a biological algorithm evolved to deal with the constraints of the tropical forest canopy may be useful in other applications, such as in swarm robotics or molecular robots [67,68,69,70]. Our work contributes to the growing intersection of distributed algorithms used by natural biological processes [74,11]. 14…”

“…This is an important problem in many network applications [9] and can be solved efficiently using numerous graph algorithms, such as Dijkstra's algorithm or the Bellman-Ford algorithm [10]. However, these classic algorithms require significantly more computation and memory than is likely available to simple biological agents such as turtle ants, who regulate their behavior using local interactions rather than central control [11].Repairing breaks requires overcoming three challenges. First, the ants must succeed in finding an alternative path by exploring new edges that currently have no pheromone and avoiding deadends in the network.…”