Smells like home: Desert ants, Cataglyphis fortis, use olfactory landmarks to pinpoint the nest

Abstract

Background: Cataglyphis fortis ants forage individually for dead arthropods in the inhospitable salt-pans of Tunisia. Locating the inconspicuous nest after a foraging run of more than 100 meters demands a remarkable orientation capability. As a result of high temperatures and the unpredictable distribution of food, Cataglyphis ants do not lay pheromone trails. Instead, path integration is the fundamental system of long-distance navigation. This system constantly informs a foraging ant about its position relative to the nest. In addition, the ants rely on visual landmarks as geocentric navigational cues to finally pinpoint the nest entrance.

Results: Apart from the visual cues within the ants' habitat, we found potential olfactory landmark information with different odour blends coupled to various ground structures. Here we show that Cataglyphis ants can use olfactory information in order to locate their nest entrance. Ants were trained to associate their nest entrance with a single odour. In a test situation, they focused their nest search on the position of the training odour but not on the positions of non-training odours. When trained to a single odour, the ants were able to recognise this odour within a mixture of four odours.

Conclusion: The uniform salt-pans become less homogenous if one takes olfactory landmarks into account. As Cataglyphis ants associate environmental odours with the nest entrance they can be said to use olfactory landmarks in the vicinity of the nest for homing.

Figure 1

Olfactory landmarks in the Cataglyphis habitat. A. Sample locations. B. Location-specific gas chromatographic profiles collected on consecutive days are displayed next to the corresponding photo. Dashed lines depict identified components that were used for EAG recordings. C. EAG responses of Cataglyphis to the identified components. Horizontal bars indicate the stimulus duration.

Figure 2

Experimental paradigm. Training. Nest situated within blue border strip; channel width and height, 7 cm, length, 16 m; position of feeder (F), 8 m downwind from nest entrance (NE); Training odour, 20 μl of either indole, nonanal, decanal, methyl salicylate (each diluted 1:50 in hexane), or hexane as solvent control. Odours were reapplied every 20 min. Test. Capture site of ants at NE; point of release at RP; position of odour in test channel 1 m upwind of RP; ants were tested with training odour, non-training odour, or solvent control. Analysis. Schematic search run. Six turning points after the ant had passed the odour for the first time were analysed for their median distance to the stimulus.

Figure 3

Discrimination among odours. A. Relative search densities of ants tested with the training odour (red plots), non-training odours (grey plots), or with the solvent as a control (white plots). Diagram columns, training odours; diagram rows, test odours; dashed line, position of odour; black arrow heads, point of release; sample size, 20 ants per plot. Search plots include the first six turning points after the ants had passed the odour for the first time. For details of graph construction see [11]. B. Median distances between the turning points and the position of odour (dashed line). Box plot: black line, median; box, interquartile range; whiskers, 10^th and 90^th percentiles, black dots, outliers. Within each plot, diagram letters indicate significant differences (p < 0.01, Dunn's post hoc test) between the groups.

Figure 4

Recognition of a learned odour in a blend. A. Relative search densities of ants that were trained with indole and tested either with indole (red line), with the solvent (black line), or with a blend of indole, nonanal, decanal, and methyl salicylate (blue area). For details see Figure 3. B. Median distances between the turning points and the position of odour (dashed line). For details see Figure 3. Diagram letters indicate significant differences (p < 0.01, Dunn's post hoc test) between the groups.