Duncan E. Jackson, Mike Holcombe, Francis L.W. Ratnieks
Pheromone trails are used by many ants to guide foragers between nest and food1, 2, 3, 4. But how does a forager that has become displaced from a trail know which way to go on rejoining the trail? A laden forager, for example, should walk towards the nest. Polarized trails would enable ants to choose the appropriate direction, thereby saving time and reducing predation risk. However, previous research has found no evidence that ants can detect polarity from the pheromone trail alone3, 5, 6, 7. Pharaoh's ants (Monomorium pharaonis) produce elaborate trail networks throughout their foraging environment8. Here we show that by using information from the geometry of trail bifurcations within this network, foragers joining a trail can adaptively reorientate themselves if they initially walk in the wrong direction. The frequency of correct reorientations is maximized when the trail bifurcation angle is approximately 60 degrees, as found in natural networks. These are the first data to demonstrate how ant trails can themselves provide polarity information. They also demonstrate previously unsuspected sophistication in the organization and information content of networks in insect societies. Figures at a glance
1. Figure 1: Pheromone trail networks of Pharaoh's ants on a smoked glass surface. a, Part of a network showing bifurcations to smaller trails (scale bar, 1 cm). b, Close-up of a single bifurcation (scale bar, 0.5 cm). c, Analysis of bifurcation angles from four separate trail networks: mean angle = 53.4°, s.d. = 14.8°, n = 321; mean distance between bifurcations = 2.9 cm, s.d. = 2.3 cm, n = 485. Solid line, normal distribution curve. 2. Figure 3: Outcomes of individual trail following tests using straight (0°) and bifurcating (30–120°) trail sections. a, Percentage of foragers making correct and incorrect reorientations. For each point n = 200, except at 0° where n = 100. b, Ratio of correct to incorrect reorientations (y = -0.0012x2 + 0.1449x + 0.9175) with 95% confidence interval (dashed lines, 36–81°). c, Percentage of individual ants departing from test trails at each angle where n = 400 for angles 30–120° and n = 200 for 0° (y = -0.0007x2 + 0.0933x + 17.625) with 95% confidence interval (dashed lines, 33–97°). (Dotted lines show the 95% confidence intervals for the angles with the maximum ratio, in both b and c.) 3. Figure 2: Experimental set-up used to form straight trails and assembly of trail bifurcations. Foraging ant traffic was constrained to produce straight pheromone trails on paper (a), which were cut into sections (b) and reassembled to form the requisite trail bifurcations with variable angle θ = 0–120°. Start locations were designated as 3, the nest direction (left) and 1 or 2, the food directions (right). Arrows show that the original direction to food on the trail was preserved when bifurcations were prepared.
Ants use chemical, visual and even magnetic cues to orient themselves correctly in their foraging environment4, 9. Some pheromone-trail-following ants readily correct their course, when displaced, by using their memory of environmental cues (landmarks and sun compass)3, 10, 11. For example, in the common garden ant Lasius niger10, 11 movement of visual cues by the experimenter results in course adjustments by the ant, while in the Argentine ant Linepithema humilis11 and La. fuliginosus3 pheromone trails are always followed, regardless of whether visual cues are moved. Thus ants using external orientation cues can readily determine direction when on a trail. However, species that are principally guided by the chemical trail itself may find reorientation harder. One solution would be to the follow the trail to its conclusion at nest or food, and then reorientate if necessary. But making such long trips to determine directionality would be costly. A better solution would...