Ants are famous in the animal kingdom for their amazing loadcarrying performance. Yet, the mechanisms that allow these insects to maintain their stability when carrying heavy loads have been poorly investigated. Here, we present a study of the kinematics of unloaded and loaded locomotion in the polymorphic seed-harvesting ant Messor barbarus. In this species, large ants have larger heads relative to their size than small ants. Hence, their center of mass is shifted forward, and even more so when they are carrying a load in their mandibles. We tested the hypothesis that this could lead to large ants being less statically stable than small ants, thus explaining their lower load-carrying ability. We found that large ants were indeed less statically stable than small ants when walking unloaded, but they were nonetheless able to adjust their stepping pattern to partly compensate for this instability. When ants were walking loaded on the other hand, there was no evidence of different locomotor behaviors in individuals of different sizes. Loaded ants, whatever their size, move too slowly to maintain their balance through dynamic stability. Rather, they seem to do so by clinging to the ground with their hind legs during part of a stride. We show through a straightforward model that allometric relationships have a minor role in explaining the differences in loadcarrying ability between large ants and small ants, and that a simple scale effect is sufficient to explain these differences.

Ants are famous in the animal kingdom for their amazing load carriage performances. Yet, the mechanisms that allow these insects to maintain their stability when carrying heavy loads have been poorly investigated. Here we present a study of the kinematics of loaded locomotion in the polymorphic seed-harvesting ant Messor barbarus. In this species big ants have larger heads relative to their size than small ants. Hence, their center of mass is shifted forward, and the more so when they are carrying a load in their mandibles. We tested the hypothesis that this could lead to big ants being less statically stable than small ants, thus explaining their lower load carriage performances. When walking unloaded we found that big ants were indeed less statically stable than small ants but that they were nonetheless able to adjust their stepping pattern to partly compensate for this instability. When ants were walking loaded on the other hand, there was no evidence of different locomotor behaviors in individuals of different sizes. Loaded ants, whatever their size, move too slowly to maintain their balance through dynamic stability. Rather, they seem to do so by clinging to the ground with their hind legs during part of a stride. We show through a straightforward model that allometric relationships have a minor role in explaining the differences in load carriage performances between big ants and small ants and that a simple scale effect is sufficient to explain these differences.