by Federico D'Agostino - Submitted to the University of Sussex on 25/01/2021
Download the full PDF with Appendix material here
<aside> <img src="https://www.halberesford.com/content/images/2018/07/null.png" alt="https://www.halberesford.com/content/images/2018/07/null.png" width="40px" /> Mechanisms underlying ants’ collective behaviour gained considerable research interest in recent years. Taking an ecological perspective, here we investigate cooperative nest maintenance in the wood ant Formica rufa, whose behaviour was elicited by blocking nest entrances with pseudo-randomly placed matchsticks. A broad, group-level perspective was chosen to analyse the behavioural output. Rather than individual ants, carried sticks were tracked and their movement metrics analysed to determine cooperation effectiveness. Interestingly, results show how, contrary to initial expectations, cooperation is inefficiently directed at both small and larger sized sticks, causing in both cases excess movement with the exclusive benefit of a moderate increase in transport speed. We integrate these findings with ecological considerations to propose that coordinated-type cooperative transport is not needed in wood ant nest maintenance and building, and that observed inefficiencies might even have advantageous effects on nest structure. This would add to the evidence for the robustness and capabilities of self-organising and stigmergic complex systems.
</aside>
Eusocial insects, and ants in particular, are well known for their ability to produce astonishing feats of collective behaviour emerging from simple and often indirect interactions among individuals without the requirement of a central controller. Complex systems’ processes such as self-organisation (organised patterns emerging from simple quantitative interactions in an initially disordered system; Invernizzi & Ruxton, 2019) and stigmergy (indirect coordination or communication between agents through traces left in the environment; Theraulaz et al., 1999) permeate the colony, to the point that it has frequently been likened to a superorganism (Hölldobler & Wilson, 2009).
Nests are arguably the most appropriate example of an emergent phenomenon, with no ant having a blueprint or teleological plan of how the final construction will look (Theraulaz et al., 1999). Nonetheless, sophisticated structures are built with numerous functional units, from air-conditioning systems to royal chambers (Hölldobler & Wilson, 1990). Studying nests and nest-building behaviour not only gives a window onto cognition, both at the individual and “super-organismic” level (Gould & Gould, 2007, Chapter 10). Since nests are considered part of an animals’ extended phenotype (Dawkins, 1983) – being in this case possibly as diverse as the over 11000 species of ants on earth are – studying nest-building also immediately sets us within an evolutionary perspective that forces us to consider ecological constraints and environmental relationships.
At the finer-grained level of individual interactions, complexity and self-organising properties can also be observed in the behavioural patterns which underlie the fascinating capacity of eusocial insects for cooperation. Among the ants, this is best exemplified by cooperative transport, which is the act of “two or more workers grasping an item and moving it intact to another location” (McCreery & Breed, 2014). Seen in over 40 genera in the Formicidae (Czaczkes, Nouvellet & Ratnieks, 2011), and likely evolved multiple times, cooperative transport abilities and techniques vary widely among ant species (Czaczkes & Ratnieks, 2013).
In short, of interest to our exploration, cooperative transport is broadly divided into coordinated and uncoordinated type. The former can be of the forward-facing or encircling kind (Czaczkes & Ratnieks, 2013) and is characterised by the fast and efficient moving of objects. The latter, uncoordinated type, is inefficient, deadlock-producing and often resolved by random changes in carriers’ composition. Importantly, it has been shown how the type of transport depends on the cognitive and physical characteristics of single ants as well as on the emergent physical properties of their interaction (Feinerman et al., 2018).
Studying ants has proved influential in unveiling characteristics of complex systems, to the point of inspiring fields as distant as robotics (Hamann, 2018), but interestingly enough only around 50 of 11000 species of ants have been studied in detail (Gordon, 2010). This is problematic because generalisation is not immediate from one specie to another, both because seemingly similar behaviours can have inherently different algorithmic causes (Gordon, 2016) and because idiosyncratic ecological factors and environmental synergies have a fundamental importance in networks of local interactions, as it is the ant colony (Gordon, 2019).
Given how fundamentally understudied details of ant collective behaviour are, in this study we focus on one specie, the common European wood ant Formica rufa, by investigating the characteristics of collective nest maintenance behaviour. The appeal of the specie resides in the impressive dimensions that its mound nests can reach (with a mean of 146 cm in width and 60 in height; Rybnikova & Kuznetsov, 2015). Is there any cooperation, specifically in the form of cooperative transport, involved in the construction and maintenance of such remarkable structures? Is that cooperation efficient, like in coordinated-type cooperative transport, and if not, does it still have a functionally valuable role?
To answer these questions – and to assess the degree of cooperation and/or self-organisation displayed by Formica rufa – nest-maintenance behaviour was elicited by blocking nest entrances with randomly placed wooden matchsticks, which the ants promptly removed. Stick movements were tracked and their metrics analysed to determine the characteristics of exhibited transport behaviour. We hypothesised to observe some degree of specialisation during cooperation in the form of selective recruitment to heavier loads and coordinated type cooperative transport.
Our study aims to extend the literature on cooperative transport, which is not only surprisingly limited but also dominated by studies on the transport of food items (McCreery & Breed, 2014). Studying cooperative nest maintenance in an ecologically realistic scenario has the ambition to widen the current perspective, given the valuable cognitive insights to be learned from construction behaviour (Hansell & Ruxton, 2008). Our investigation also intends to provide detailed information of collective behaviour for a precise specie, hoping to be a step of many others towards a targeted and clearer characterisation of cooperative phenomena.
Experiments were performed with queen-right colonies of the ant Formica rufa L., the common wood ant of the Palaearctic region, which were collected on local forests near the Campus of the University of Sussex (UK), and housed in wide tanks with natural nest material in the laboratory at 20-25 °C.
