Twist and chew: three dimensional tongue kinematics during chewing in macaque primates

Three-dimensional (3D) tongue movements are central to performance of feeding functions by mammals and other tetrapods, but 3D tongue kinematics during feeding are poorly understood. Tongue kinematics were recorded during grape chewing by macaque primates using biplanar videoradiography. Complex shape changes in the tongue during chewing are dominated by a combination of flexion in the tongue’s sagittal planes and roll about its long axis. As hypothesized for humans, in macaques during tongue retraction the middle (molar region) of the tongue rolls to the chewing (working) side simultaneous with sagittal flexion, while the tongue tip flexes to the other (balancing) side. Twisting and flexion reach their maxima early in the fast close phase of chewing cycles, positioning the food bolus between the approaching teeth prior to the power stroke. Although 3D tongue kinematics undoubtedly vary with food type, the mechanical role of this movement—placing the food bolus on the post-canine teeth for breakdown—is likely to be a powerful constraint on tongue kinematics during this phase of the chewing cycle. The muscular drivers of these movements are likely to include a combination of intrinsic and extrinsic tongue muscles.

Methods
These data were collected using the University of Chicago XROMM Facility. All procedures were approved by the University of Chicago IACUC. Four adult rhesus macaques (Macaca mulatta) were observed in this study (Table 1). The animals were housed in an AAALAC-accredited animal facility, fed monkey biscuits and given daily enrichment. They were trained to feed while restrained in a radiolucent, acrylic primate chair. Monkeys C and H were trained to feed with their heads restrained, facilitating marker tracking, shortening recording durations, and reducing radiation exposure. The monkeys were fed red grapes (10-20 mm) and biplanar videoradiographic data were collected (technique: 90–100 kVp, 10–16 mA, 200 Hz, and with a 2–4 ms shutter speed). Data were processed using the XROMM workflow and custom R scripts. Analyses were conducted using custom R scripts, included here.