How memory and motivation modulate the responses to trail pheromones in three ant species

Abstract

Ants are able to modulate their behavior according to private and collective information. Collective information is coded in different concentrations of pheromone deposited in the environment, especially on ant trails, whereas private information is learned and memorized by individual ants. It has been shown that both kinds of information act synergistically on the collective trail-following behavior of ant colonies. Another important factor influencing ant behavior is their motivation to follow pheromone trails. Here, we show how private information and motivation modulate the response to collective information. We investigate these effects using a recently proposed approach that employs psychophysical methods to measure the response to varying pheromone concentrations. We studied the effect of private information (route memory) in the species Lasius niger, Euprenolepis procera, and Linepithema humile. Additionally, the effect of motivation was studied in the species E. procera and L. humile. Using psychophysical methods, we quantified these effects for important biological parameters like behavioral thresholds and error rates. The differential changes in these parameters between the three species imply specific adaptations to their environment.

Appendix

Introduction into psychophysical methods used by von Thienen et al. (2014)

In psychophysics, a psychometric function (PF) describes the relationship between the probability of a positive response p to a stimulus and the stimulus strength x

$$ p(x)=\gamma +\left(1-\lambda -\gamma \right)\cdot F(x). $$

x—stimulus strength, λ—guess rate, γ—lapse rate, F(x)—function describing the probability to detect a stimulus by the underlying sensory mechanism (Kingdom and Prins 2010 p. 74). For F(x) a probability distribution like the Weibull distribution is applied. For mathematical details, please refer to von Thienen et al. (2014).

von Thienen et al. (2014) used the PF to provide a mathematical description for the response of ants to trail pheromones of varying concentrations. They performed two types of psychophysical experiments: Detection experiments and discrimination experiments. The detection experiments measured the ability to detect a trail pheromone of varying intensity against a null stimulus or noise. Discrimination experiments measured the ants’ ability to discriminate a trail pheromone of varying concentration from a reference trail pheromone of constant concentration. Both types are of specific biological importance for understanding the collective behavior of ants. Detection is the ants’ ability to detect a pheromone trail and become attracted to it. It is important, for instance, to describe how ants at the nest entrance respond to a trail that leads out of the nest to a food source. Discrimination is the ability to differentiate pheromone concentrations at a trail bifurcation. This is especially important in situations where ants have to collectively find the better of two alternate solutions, for instance food of different quality. PFs are S-shaped curves providing a probability of response, theoretically ranging from 0.5 to 1 in detection experiments and from 0 to 1 in discrimination experiments. The stimuli were presented as trail pheromone of varying concentrations on one of the two branches of a y-shaped bridge. Parameters of the PF are the lapse rate (the distance between upper asymptote and 1) and the 75 %-detection/discrimination threshold. von Thienen et al. (2014) interpreted the lapse rate biologically in two ways: as a measure for the errors that ants make and as a measure for the independence from pheromone guided behavior. It might be important for ants not to get stuck in suboptimal solutions and explore alternate solutions. The thresholds are standardized measures of the ants’ ability to detect a stimulus or to discriminate between two stimuli, respectively. von Thienen et al. (2014) had shown that differences between the PFs of the three species (L. humile, E. procera, and L. niger) can be attributed to specific adaptations to their environment. L. humile ants coordinate the work of extremely large and interconnected colonies (Holway and Case 2000; Giraud et al. 2002). They depend on persistent as well as short-lived food supplies (Mallis 1942; Suarez et al. 1998). Thus, they depend highly on collective information and must be able to communicate even the smallest changes in their environment. This reflects in a low lapse rate and a low detection threshold. L. niger, on the other hand, has comparably small colony sizes and uses persistent as well as short-lived food sources (Seifert 2007). Thus, collective information is less important compared to L. humile while individual information plays an important part. This reflects in a higher lapse rate and a high detection threshold. E. procera is positioned between these two extremes. Their colonies are also much smaller than those of L. humile. Furthermore, the ants are nocturnal so that visual cues are rare (Witte and Maschwitz 2008; von Beeren et al. 2014). Consequently, they depend more on collective information and show a high trail fidelity, which reflects in the lowest lapse rate of all three species. Since their food sources are much more stable compared to the other two species, they do not have to share as much information and consequently have a comparatively high detection threshold.