Fig 1.
Fittings to the step length distribution for 8 individual Me’Phaa farmers searching for firewood.
Rank frequency plots of the step lengths are shown (o) together with fits to power-law (red-lines), exponential (blue lines) and bi-exponentials (green lines). The Akaike weights, wp and wbi, for the power-law and bi-exponential fits are shown together with the maximum likelihood estimates for the power-law (Lévy) exponents, μ. The numbers of steps before application of the method of Tromer et al. [21] are 247, 92, 455, 183, 258, 235, 77 and 239. After application of the method of Tromer et al. [21] which combines shorter steps into larger steps the numbers of steps are 33, 13, 138, 36, 100, 52, 32 and 99.
Fig 2.
Fittings to the step length distribution for 6 Cariri Brazilian farmers searching for fruit.
Rank frequency plots of the step lengths are shown (o) together with fits to power-law (red-lines), exponential (blue lines) and bi-exponentials (green lines). The Akaike weights, wp and wbi, for the power-law and bi-exponential fits are shown together with the maximum likelihood estimates for the power-law (Lévy) exponents, μ. Outlying displacements due to initial bouts of cycle riding have not been removed prior to the analysis. The numbers of steps are 237, 357, 422, 380, 487 and 619.
Fig 3.
Fittings to the step length distribution for 6 Amazonian farmers searching for Brazilian nuts.
Rank frequency plots of the step lengths are shown (o) together with fits to power-law (red-lines), exponential (blue lines) and bi-exponentials (green lines). The Akaike weights, wp and wbi, for the power-law and bi-exponential fits are shown together with the maximum likelihood estimates for the power-law (Lévy) exponents, μ. The numbers of steps are 368, 317, 214, 241, 444 and 871.
Fig 4.
Trail following by the Me’Phaa out of Agua Tordillo small village.
This is a typical example of a GPS recorded trajectory followed by an individual while executing a firewood searching exploration in the Guerrero Mountains in Mexico. The full trajectory is depicted overlaid in the map showing the main dirt roads used for walking. A is the start of the trajectory while B marks its end and the site where most of the wood search and collection has taken place. Once the individual goes off the dirty road (R) near the end of the trajectory, he goes into the woods following a network of footpaths (F). In the inset, a section of the outward and inward walking trajectory, as registered by the GPS, is shown overlaid over the almost straight line geometry of the dirt road segment. Notice the walker meandering across this bounded terrain feature. Map elaborated from our own GPS records. See S1 File for more details about this figure.
Fig 5.
Trail following by the Brazilian Cariri gatherers.
Typical example of a GPS recorded trajectory followed by an individual while executing a fruit gathering exploration in the Chapada do Araripe, Brazil. The trajectory starts at the end of a dirt road/trail and most of it follows the contours of a cliff edge (F), marked as the boundary of the two green colors. The foraging was made in the top of the mountain were the natural park is located. A is the start of the walking trajectory while B marks its end. R are dirt roads. There is also an extensive network of a footpaths. Location coordinates: 7°23'11.60"SS, 39°24'35.82"W. Map elaborated from our own GPS records. See S1 File for more details about this figure.
Fig 6.
Typical example of a GPS recorded trajectory followed by an individual while executing a nut gathering excursion in the Amazonian rainforest in Rondônia, Brazil exhibiting the local terrain features as well.
In some cases forest features (i.e. changes in canopy density) other than footpaths are used when walking, these may include the borders of forest gaps where vegetation is secondary or primary (F) and is somewhat more easy to walk. A is the start and B the end of the trajectory, R is a dirt road. Location coordinates: 10°53'40.68" S, 65°03'44.68"W. Map elaborated from our own GPS records. See S1 File for more details about this figure.
Fig 7.
Feature following by the Hadza.
A GPS recorded trajectory followed by an individual while executing a foraging trip, south of lake Eyasi in the Arusha region of central Tanzania, and overlaid on the local landscape (trajectory in blue from graph 2a in Raichlen et al, 2014). Location coordinates: 3°43'32.70"S, 35°11'7.00"E. In the image, the individual follows terrain features such a dry river basin (D). A network of footpaths is also evident (F). It remains to be seen whether or not feature and/or trailing following is evident in any other of 341 Hadza recorded trajectories. Image source is DigitalGlobe Inc. under a CC BY 4.0 license.
Fig 8.
Fittings to the step length distribution for 4 simulated individuals.
Rank frequency plots of the step lengths are shown (o) together with fits to power-law (red-lines), exponential (blue lines) and bi-exponentials (green lines). The Akaike weights, wp and wbi, for the power-law and bi-exponential fits are shown together with the maximum likelihood estimates for the power-laws (Lévy) exponent, μ. Individuals perform random walks when making excursions. There is a single, straight trail. The walker moves with constant speed parallel to the trail and moves randomly in the orthogonal direction. Step-lengths in these random walks are exponentially distributed with mean length 1 a.u. An excursion ends when the walker first returns to the trail.