In this thesis, I investigate the mechanical and energetic constraints of leaf cutting—a key aspect of insect herbivore foraging behaviour—using highly polymorphic leaf-cutter ants Atta cephalotes as a model system. I begin by exploring the factors that influence insect herbivory through the lens of fracture mechanics. I examine how the force required to cut substrates, which vary widely in mechanical properties, changes with both substrate type and worker size. Using a mechanical model, I test predictions on how substrate thickness, toughness, and cutting-edge radius affect cutting force. Cutting forces are found to be size-invariant across worker sizes, as focused ion beam milling (FIB) and scanning electron microscopy (SEM) scans reveal no difference in cutting edge radius. Hence, my results indicate that for pristine tools, cutting force scales directly with leaf toughness and thickness. Next, I assess the variation in metabolic cost of cutting across worker size and substrate. Utilising flow through respirometry, I find that the metabolic power output during cutting remains constant at 30 times the resting rate, regardless of mechanical demand, suggesting that metabolic power output is invariant of substrate properties. Finally, having established the scaling of metabolic power and its independence from substrate mechanical properties, I turn to the scaling of mechanical power. I develop a predictive mechanical model to describe the non-linear variation in cutting speed, which I validate through experimental data. To conclude, I integrate findings from all three experimental chapters of my thesis, deriving a model to estimate changes in cutting efficiency. The result of this work adds a novel perspective to the age-old question - do leaf-cutter ants assign workers to cutting according to ergonomic criteria?