How easily can we obtain optimal trade-offs between conflicting ecosystem services (ES)? We studied this question by crossing metrics of optimality and robustness in a model simulating sheep/cattle mixed-grazing (mixed-grazing is grazing by more than one species). We hypothesized that mixed-grazing processes (complementary use of vegetation and parasitism reduction) would improve ES bundles by increasing meat production with limited additional environmental costs. We assessed bundle optimality with a production possibility frontier and robustness through a management density approach (bundles can be obtained by one or several management decisions). We modeled two provisioning and two regulating ESs and confirmed that mixed-grazing can potentially improve the monetary value of bundles. Optimal bundles were the most robust, because of the shape of the biological function driving animal growth, according to the sheep/cattle ratio. This function is humpshaped with a plateau that buffers small ratio deviations. It makes bundles optimal or quasi-optimal over a wide range of management decisions, which eases their optimization. For this reason, based on the principles of ergonomics and the definition of the adjective operable ('capable of being put into use, operation, or practice'), we considered mixed-grazing a 'Biologically Operable' practice.