Skip to main page content
Access keys NCBI Homepage MyNCBI Homepage Main Content Main Navigation
. 2015 Nov 11;10(11):e0141269.
doi: 10.1371/journal.pone.0141269. eCollection 2015.

On Heels and Toes: How Ants Climb With Adhesive Pads and Tarsal Friction Hair Arrays

Free PMC article

On Heels and Toes: How Ants Climb With Adhesive Pads and Tarsal Friction Hair Arrays

Thomas Endlein et al. PLoS One. .
Free PMC article


Ants are able to climb effortlessly on vertical and inverted smooth surfaces. When climbing, their feet touch the substrate not only with their pretarsal adhesive pads but also with dense arrays of fine hairs on the ventral side of the 3rd and 4th tarsal segments. To understand what role these different attachment structures play during locomotion, we analysed leg kinematics and recorded single-leg ground reaction forces in Weaver ants (Oecophylla smaragdina) climbing vertically on a smooth glass substrate. We found that the ants engaged different attachment structures depending on whether their feet were above or below their Centre of Mass (CoM). Legs above the CoM pulled and engaged the arolia ('toes'), whereas legs below the CoM pushed with the 3rd and 4th tarsomeres ('heels') in surface contact. Legs above the CoM carried a significantly larger proportion of the body weight than legs below the CoM. Force measurements on individual ant tarsi showed that friction increased with normal load as a result of the bending and increasing side contact of the tarsal hairs. On a rough sandpaper substrate, the tarsal hairs generated higher friction forces in the pushing than in the pulling direction, whereas the reverse effect was found on the smooth substrate. When the tarsal hairs were pushed, buckling was observed for forces exceeding the shear forces found in climbing ants. Adhesion forces were small but not negligible, and higher on the smooth substrate. Our results indicate that the dense tarsal hair arrays produce friction forces when pressed against the substrate, and help the ants to push outwards during horizontal and vertical walking.

Conflict of interest statement

Competing Interests: The authors have declared that no competing interests exist.


Fig 1
Fig 1. Set-up for ground reaction force recordings.
(A) Weaver ants climbing on a vertical surface stepped onto a small, transparent glass plate attached to a sensitive force transducer. (B) Forces in the normal (Z) and vertical (X) directions, and contact area of the arolium for the step of a front leg above the CoM.
Fig 2
Fig 2. Tarsomere angle of freely walking weaver ants.
(A) Level walking. Here, the ants stood mainly on their 3rd and 4th tarsomeres and held the 5th tarsomeres away from the substrate (5th tarsomere angle <0°; see photo). (B) Inverted climbing. Here, the 5th tarsomere angle of was mostly positive, allowing the adhesive pad (arolium) to make contact (see also photo).
Fig 3
Fig 3. Ground reaction forces in vertically climbing weaver ants.
(A) normal forces. (B) vertical forces (number of legs above (below) CoM: N front = 9(3);N middle = 10(8);N hind = 8(18)).
Fig 4
Fig 4. Single-leg vertical forces in weaver ants climbing on a vertical glass surface, as a function of leg orientation.
Leg orientation is given as an angle ranging from -90° (pointing downward; N front = 3, N middle = 8, N hind = 18) to 0° (horizontal) to +90° (pointing upward; N front = 9, N middle = 10, N hind = 8). For all legs with negative angles, the foot position was below the CoM, and positive angles corresponded to foot positions above the CoM.
Fig 5
Fig 5. Relationship between peak normal force and corresponding shear force along the (projected) axis of the leg for individual steps.
Filled circles denote steps in which the arolium made visible surface contact, whereas steps with detached arolium are marked with open circles. Lines show standardised major axis regressions on both types of steps. In most steps where the adhesive pad made contact, the legs were pulling, whereas steps without visible arolium contact occurred mostly when legs pushed. The drawings illustrate the direction of the normal forces (F N) and shear forces (F S) and the presence or absence of adhesive pad contact (black or white filling of the pad).
Fig 6
Fig 6. Morphology of tarsal hairs in O. smaragdina ants and generated friction forces.
(A) Morphology of hairs on the underside of the tarsus. Ta4 & Ta5: 4th & 5th tarsomere, respectively; Cl: Claws; Ar: Arolium; (B) Friction forces of tarsal hairs on a smooth and a rough substrate. Slides were performed in the pulling and pushing direction with a constant normal load of 30, 60 and 100 μN. The dashed line indicates the ants’ body weight (82 μN).
Fig 7
Fig 7. Friction forces of tarsal hairs in the pushing direction.
(A) on a smooth and (B) on a rough surface for varying normal preloads (30, 60 and 100 μN). The dashed lines show the results of a linear regression on the median for each normal force level.
Fig 8
Fig 8. Adhesive forces of tarsal hairs on a smooth and rough surface (1 μm asperity size) after pushing slides under different normal loads (30, 60 and 100 μN).
Fig 9
Fig 9. Typical friction force trace from tarsal hair fields against a smooth and rough surface.
(A) SEM image showing the tarsal hairs of the 3rd and 4th tarsomeres (the part in contact with the surface is outlined with a dashed line). (B) Contact area of the hairs in pulling orientation using coaxial illumination on a stereo-microscope. (C) Hairs at the point of buckling. (D) Hairs re-orientated under pushing shear forces. (E) Force trace of tarsal hairs under different preloads and shearing directions on smooth and rough surfaces. Images in sub-figures (B-D) were taken at the points marked with squares in the raw trace curve. Circles mark characteristic ‘kinks’ in the force curve indicating the buckling of the hairs. Darker shaded regions indicate the time period during which the motorised stage moved the sample in the pushing direction.
Fig 10
Fig 10. Reflected-light microscopy images of the tarsal hairs in contact with the smooth surface.
(A) side contact of the hairs (B) tiny droplets left on the surface after pull-off.

Similar articles

See all similar articles

Cited by 9 articles

See all "Cited by" articles


    1. Niederegger S, Gorb S. Tarsal movements in flies during leg attachment and detachment on a smooth substrate. J. Insect Physiol. 2003; 49; 611–620. 10.1016/S0022-1910(03)00048-9 - DOI - PubMed
    1. Gorb S, Jiao Y, Scherge M. Ultrastructural architecture and mechanical properties of attachment pads in Tettigonia viridissima (Orthoptera: Tettigoniidae). J. Comp. Physiol. A. 2000; 186; 821–831. 10.1007/s003590000135 - DOI - PubMed
    1. Federle W, Brainerd EL, McMahon TA, Hölldobler B. Biomechanics of the movable pretarsal adhesive organ in ants and bees. P. Natl. Acad. Sci., USA. 2001; 98; 6215–6220. 10.1073/pnas.111139298 - DOI - PMC - PubMed
    1. Federle W, Endlein T. Locomotion and adhesion: dynamic control of adhesive surface contact in ants. Arthropod Struct. Dev. 2004; 33; 67–75. 10.1016/j.asd.2003.11.001 - DOI - PubMed
    1. Clemente CJ, Federle W. Pushing versus pulling: division of labour between tarsal attachment pads in cockroaches. P. Roy. Soc. B-Biol. Sci. 2008; 275; 1329–1336. 10.1098/rspb.2007.1660 - DOI - PMC - PubMed

Publication types

LinkOut - more resources