The running kinematics of free-roaming giraffes, measured using a low cost unmanned aerial vehicle (UAV)

Abstract

The study of animal locomotion can be logistically challenging, especially in the case of large or unhandleable animals in uncontrolled environments. Here we demonstrate the utility of a low cost unmanned aerial vehicle (UAV) in measuring two-dimensional running kinematics from free-roaming giraffes (Giraffa camelopardalis giraffa) in the Free State Province, South Africa. We collected 120 Hz video of running giraffes, and calibrated each video frame using metatarsal length as a constant object of scale. We tested a number of methods to measure metatarsal length. The method with the least variation used close range photography and a trigonometric equation to spatially calibrate the still image, and derive metatarsal length. In the absence of this option, a spatially calibrated surface model of the study terrain was used to estimate topographical dimensions in video footage of interest. Data for the terrain models were collected using the same equipment, during the same study period. We subsequently validated the accuracy of the UAV method by comparing similar speed measurements of a human subject running on a treadmill, with treadmill speed. At 8 m focal distance we observed an error of 8% between the two measures of speed. This error was greater at a shorter focal distance, and when the subject was not in the central field of view. We recommend that future users maximise the camera focal distance, and keep the subject in the central field of view. The studied giraffes used a grounded rotary gallop with a speed range of 3.4–6.9 ms−1 (never cantering, trotting or pacing), and lower duty factors when compared with other cursorial quadrupeds. As this pattern might result in adverse increases in peak vertical limb forces with speed, it was notable to find that contralateral limbs became more in-phase with speed. Considering the latter pattern and the modest maximal speed of giraffes, we speculate that tissue safety factors are maintained within tolerable bounds this way. Furthermore, the angular kinematics of the neck were frequently isolated from the pitching of the body during running; this may be a result of the large mass of the head and neck. Further field experiments and biomechanical models are needed to robustly test these speculations.

Giraffes and hominins: reductionist model predictions of compressive loads at the spine base for erect exponents of the animal kingdom

“…Both in walking ( Basu et al, 2019b ) and galloping ( Basu et al, 2019a ), giraffes hold their necks at about φ =35°. Values of the compressive stress on the giraffe's neck base IVD in their everyday standing postures ( Fig.…”

The evolutionary biomechanics of locomotor function in giant land animals

“…At best, these methods might only coarsely apply for comparisons with land giants (e.g. Basu et al, 2019a , b ). Indeed, some aspects of dynamic similarity are not maintained even across smaller animals – peak joint and other maximal forces scale with strong negative allometry ( Alexander, 1980 , 1985a , b ) – raising the question, how is such scaling explained in terrestrial vertebrates and what does it mean for land giants?…”

“…Furthermore, some other maximal speed estimates used for giant mammals seem excessive (e.g. 16.7 m s −1 for 1000 kg giraffes versus ≤11 m s −1 in Alexander et al, 1977 ; Basu et al, 2019a ). Regardless, such amendments would only strengthen the conclusion that maximal speed declines steeply with mass in giant land mammals.…”

“…Giraffes, like large bovids, straddle the boundary between large and giant megafauna, and their locomotor abilities reflect that status. Their locomotor kinematics and kinematics generally maintain some dynamic similarity ( Box 1 ) with smaller quadrupeds, except that they adopt lower stride frequencies, related to their apomorphically elongate limbs ( Basu et al, 2019a ). Those long ‘cursorial’ limbs might incur penalties to EMA ( Basu, 2019 ; Basu and Hutchinson, 2021 preprint) and further trade-offs with maximal performance, since giraffes are slower than horse-sized mammals (maximal speed ∼11 m s −1 ; Dagg and Vos, 1968 ; Alexander et al, 1977 ; Basu et al, 2019a ).…”

“…Sensorimotor responsiveness is also slower due to the elongate limbs ( More et al, 2013 ). Although large giraffes can still gallop (but do not trot; Dagg, 1973 ; Dagg and Vos, 1968 ), they do so more sedately than smaller individuals, shifting limb kinematics in ways that reduce peak forces and enable this athleticism ( Basu et al, 2019a , b ).…”