Remarkable progress has been made characterising one of nature’s most integrated, hierarchical structures – the fibrillar adhesive system of geckos. Surprisingly, we lack an understanding of how multiple, soft toes coordinate to facilitate gecko’s acrobatic locomotion. Here, we tested the control function of multiple toes of sideways wall running geckos by varying substrate orientation, friction, and available area. Geckos realigned the toes of top feet toward the upward direction to resist gravity. Top front and hind feet functioned like the front feet during vertical climbing. The contact area was not compromised but redistributed by toes. Geckos aligned all toes upward to resist slipping when encountering low friction patches and increased the contribution of particular toes to compensate for adhesion loss when not all toes were available during the negotiation of intermittent slippery strips. Increasing substrate roughness by replacing the flat wall with discrete rods perpendicular to motion resulted in geckos bending and/or rotating toes to conform to and even grasp the rods, with a potential force more than five times body weight. Our findings revealed that geckos increase their effectiveness of manoeuvrability in demanding environments by taking advantage of the distributed control afforded by multiple, adjustable soft toes.

1. Data1.xlsx. The first file contains the frictional adhesion (i.e., shear force) and toes contact area at three moments in the stance phase of geckos climbing a wall upward.  We embedded several test units consisting of (frustrated total internal reflection) FTIR enhanced acrylic and force sensors in the wall. We collected the foot rection force and contact images at the same time (500Hz) through a DAQ module and highspeed camera.  We obtained the contact area of toes through the highspeed video images through image processing (Eason et al. 2015). We assumed toes as vectors with values of their contact area and directions of their own directions. Thus, we obtained the resultant contact area at feet (see the main text for the method). In this file, we show the frictional adhesion and resultant contact area for three moments in stance phase, (i) the moment when the force reached its half peak,(ii) the moment when the force reached its peak, (iii)the moment when the force reduced to its half peak.
2. Data2.xlsx. The second file contains the contact area and orientation angles of all toes of geckos in upward climbing, sideways wall-running without slippery strips, and with slippery strips. We first ran geckos on a vertical, FTIR-enhanced acrylic sideways and recorded their toe contacts through highspeed camera. Then we ran the same individuals upward as control. We also added slippery strips (Teflon, width 5mm, and gaps 10mm) to the rear part of the acrylic and ran the same individuals sideways again.  We obtained the contact are and orientation angles of toes through images processing. The data in this file was selected at mid-stance.
3. Data3.xlsx. The third file contains the relative force of geckos’ feet while being pulled along and across acrylic rods with diameters ranging from 6.4mm to 38mm. We connected the rods to force sensors and measured the shear force while dragging geckos feet along the rods and across the rods, respectively. We also pulled the geckos feet on a flat acrylic sheet before and after a test on the rod as its control. For each test, we divided the maximum shear force on the rod with the corresponding maximum shear force on the flat control surface to calculated the relative force.

There are many blanks in the second file. These are not missing data. The blanks indicate the toes that show no contact at the mid-stance of corresponding feet. This also indicates that some toes contribute to the whole feet adhesion occasionally.

There is also some negative contact area in the dataset. This is not a n erroir, just because the direction of the resultant contact area calculated by defining toes as vectors is not in the positive direction we defined.

These notes were also included in a ReadMe.txt.