foraging efficiency in ants Synergy between social and private information increases

Insect societies integrate many informationsources to organize collective activities such asforaging. Many ants use trail pheromones toguide foragers to food sources, but foragers canalso use memories to find familiar locations ofstable food sources. Route memories are oftenmore accurate than trail pheromones in guidingants, and are often followed in preference totrail pheromones when the two conflict. Whythen does the system expend effort in producingand acquiring seemingly redundant and low-quality information, such as trail pheromones,when route memory is available? Here we showthat, in the ant Lasius niger, trail pheromonesand route memory act synergistically duringforaging; increasing walking speed and straight-ness by 25 and 30 per cent, respectively, andmaintaining trail pheromone deposition, butonly when used together. Our results demon-strate a previously undescribed major role oftrail pheromones: to complement memory byallowing higher confidence in route memory.This highlights the importance of multiple inter-acting information sources in the efficientrunning of complex adaptive systems.Keywords: synergy; trail pheromone; route memory;multiple information sources1. INTRODUCTIONThe integration of multiple information sources isnecessary for the functioning of adaptive biological sys-tems at cell, organ, organism and society levels [1–4]and also in technological systems [5]. The successfulcoordination of the many individuals in a colony ofsocial insects typically involves the gathering and trans-fer of information from several sources [6–8]. Workersmay gain information either by interacting with theirenvironment (private information) or by interactingwith their nest-mates (social information) [9,10].During foraging, workers commonly transfer infor-mation in order to enhance the ability of the colonyto forage efficiently in a constantly changing environ-ment. Honeybees, for example, use the waggle danceto inform nest-mates about the location of foodpatches in the environment [11], whereas many antsuse trail pheromones. Pheromone trails, however, donot provide perfect information and naive foragers fre-quently make mistakes when using trail pheromones attrail bifurcations [12–15]. Gru¨ter et al.[12] found thatin Lasius niger, only 70 per cent of foragers chose a trailwith high levels of trail pheromone at a bifurcation.Route memories, on the other hand, may be acquiredrapidly, leading to 95 per cent correct choices byL.nigerworkers at a bifurcation after only three previousvisits to a food source on one branch [12]. In situationswhere these information types conflict, many speciesfollow route memories in preference to trail phero-mones [12,13,16–18], as do honeybees when socialand private information conflicts [19]. Aphid-tendingants may return to stable food sources for days,weeks or months [16,20,21]. Why then do ants con-tinue laying pheromones on a trail past the initialrecruitment phase when they have more reliableroute memories? One possibility is that pheromonesare only used by naive ants. However, this does notexplain why pheromone deposition continues past theinitial recruitment. Could pheromones interact withroute memory to play a yet unknown function for thevast majority of foragers that are travelling alongfamiliar routes? We propose an alternative hypothesisthat trail pheromones may complement routememory. Foraging ants may be using the trail phero-mone as a reassurance marker, like a white line on aroad, allowing ants to reduce the speed/accuracytrade-off by reducing the foragers’ need to spendtime checking their location or avoiding straying fromtrails. Foraging efficiency could thus be increased onfamiliar trails. We therefore tested whether the pres-ence of trail pheromone causes an increase in walkingspeed and path straightness, and a decrease in therate of U-turning in both experienced and naive ants.2. MATERIAL AND METHODS