![](/wp-content/uploads/2020/12/abstractheadergraphic2019.gif)
Meeting Abstract
Ants walk long distances to find food and defend territories. Yet, the substrates they experience are typically rugged and uneven, potentially challenging locomotion performance. Because ant walking speed and energetics have significant ecological and evolutionary implications, ants are a useful system to study kinematic modulation on rough ground. Here we combine laboratory studies of walking kinematics with ecologically-relevant field preference experiments to understand how walking patterns and preferences vary with substrate. A custom automated camera system recorded over 8000 high-speed videos of Argentine ants (Linepithema humile) walking on 3D-printed substrates with a checkerboard pattern of varying step width (1, 3 and 5 mm, 1 mm step height). Although ants on rough substrates could occasionally reach the faster speeds observed on flat ground, they demonstrated a strong speed constraint while walking on rough substrates, particularly those with checkerboard widths smaller than their 3-4 mm body length (median speed decreased by 21%, 20%, and 12% on 1, 3, and 5 mm respectively compared to flat ground). An automated, deep-learning method tracked limb kinematics during >210,000 strides, revealing that rough substrates both increased variability in foot placement (by 60, 70, and 44% respectively compared to flat ground) and shifted the proportion of disrupted strides (from 5% on flat ground to 24%, 18%, and 12% on 1, 3, and 5 mm substrates). This study represents a highly-detailed investigation of ant locomotion and limb kinematics on uneven ground. Our findings have the potential to explain observed ecological patterns, inform conservation guidelines, and inspire new robotic control strategies.