Ants learn fast and do not forget: Resistance to extinction may be advantageous when the environment is stochastic and individuals need to switch often from one learned task to another

Ants learn fast and do not forget: associative olfactory learning, memory and extinction in Formica fusca. Baptiste Piqueret, Jean-Christophe Sandoz and Patrizia d'Ettorre. Royal Society Open Science. June 19 2019. https://doi.org/10.1098/rsos.190778

Abstract: Learning is a widespread phenomenon that allows behavioural flexibility when individuals face new situations. However, learned information may lose its value over time. If such a memory endures, it can be deleterious to individuals. The process of extinction allows memory updating when the initial information is not relevant anymore. Extinction is widespread among animals, including humans. We investigated associative appetitive learning in an ant species that is widely distributed in the Northern Hemisphere, Formica fusca. We studied acquisition and memory between 1 h and one week after conditioning, as well as the extinction process. Ants learn very rapidly, their memory lasts up to 3 days, decreases slowly over time and is highly resistant to extinction, even after a single conditioning trial. Using a pharmacological approach, we show that this single-trial memory critically depends on protein synthesis (long-term memory). These results indicate that individual ant workers of F. fusca show remarkable learning and memory performances. Intriguingly, they also show a strong resistance to updating learned associations. Resistance to extinction may be advantageous when the environment is stochastic and individuals need to switch often from one learned task to another.

1. Background

Behavioural flexibility offers significant fitness advantages, especially in environments where resource distribution or threats are characterized by stochasticity. One way to achieve this flexibility is via learning, defined as a change in behaviour occurring as a result of experience. Many learned behaviours can be further modified to suit changing environmental conditions. The ability to learn and memorize allows animals to respond to environmental stimuli in an adaptive way, for instance, by either ignoring them or by giving them a specific value, positive or negative. This helps in predicting the environment when facing new but similar situations [1].

Storing information is costly; therefore, only essential pieces of information should remain available for the individual. For instance, with time, a stimulus which used to predict a certain resource in the environment (e.g. the presence of food) might lose its significance and be no longer associated with the resource. It is beneficial to learn rapidly that such a stimulus is not reliable anymore. Extinction is the process in which a conditioned response gradually decreases through repeated experience with the stimulus in the absence of its outcome. Extinction generally involves the formation of a new inhibitory memory rather than the destruction of the previous memory [2]. Knowledge about the extinction process has important clinical applications, for instance, for the treatment of drugs addiction and abnormal fear of a past event (e.g. war trauma) in humans [3].

The extinction phenomenon was first described by Pavlov in 1927 in experiments with dogs using classical conditioning, the association of an unconditioned stimulus (US, for example a reward) with an initially neutral stimulus that becomes a conditioned stimulus (CS) producing the response in the absence of the US. After a successful conditioning (CS–US association), Pavlov observed the conditioned responses stopped after a few unrewarded CS presentations, leading to the extinction of the conditioned behaviour. Extinction does not erase the old memory. It is rather a new learning (creation of a CS–no US association). Therefore, two memories coexist. When time passes after successful extinction, the original behaviour may reappear (called spontaneous recovery or relapse), through a decay of the extinction memory [2]. Associative learning and extinction are widespread in the animal kingdom and have been intensively studied in several vertebrate species such as mice [4] or zebrafish [5] and also in invertebrates species such as snails [6], crabs [7] or nematodes [8]. Among invertebrates, insects like fruit flies became key model species for learning and memory [9–11]. Insects are well suited for laboratory studies because they are relatively easy to keep, they have short reproductive cycles and offer easy access to brain structures (e.g. crickets, [12]). Learning and extinction have also been investigated in social insects, including bumblebees [13,14] and honeybees [2,15]. Among social insects, ants are the most diverse group with more than 14 000 described species, which represent up to 25% of the total animal biomass on Earth [16]. Visual learning in ants has been intensively studied, also at the individual level, in the context of spatial orientation and navigation [17–19]. Individual olfactory learning has been less investigated [20–27]. Carpenter ants are very efficient in discriminating between different odorants [24,26] and even between different concentrations of the same compound [27]. Recently, workers of Lasius niger were shown to be able to learn odour–reward associations after only one training trial, while more trials were required when using spatial cues instead of odours [23]. However, in this study, the dynamics of memory formation was not investigated. We know that individual ants can form long-term olfactory memories after six CS–US presentations [22], but whether fewer conditioning trials lead to long-term memory (LTM) is unclear. Furthermore, data about extinction of olfactory learned associations are very scarce in ants [28].

In the present work, we present the results of a laboratory study on individual associative olfactory learning, memory and extinction in the ant Formica fusca. Among ants, the genus Formica was described as one of the most advanced from a cognitive point of view (especially concerning communication and learning) [29]. Formica fusca is widely distributed and lives in a variety of environments with a large range of temperatures, resources, predators and competitors. Colonies are populous (hundreds of individuals) and grow well in laboratory conditions. We investigated the acquisition performance of individual ants by changing the number of conditioning trials (from one to six). We tested ants' memory abilities by subjecting them to a memory test between 1 h and one week after training. We then categorized the memory using a pharmacological approach by administrating a protein synthesis inhibitor. Finally, we studied the extinction phenomenon in individual ants, by measuring their behaviour after unrewarded presentations of the CS.