In the context of agroecological transition, the challenge for dairy goat farms is to decrease the use of purchased concentrates and increase the use of grass in diets, with forages or grazing. Such changes potentially expose the system to different constraints: purchasing feedstuff exposes the system to price fluctuations, relying on more grass-based diets exposes to forage yield fluctuations. The system can express different types of robustness in face of these perturbations and trade-offs may appear between them. Objective of this study was to quantify the effects of price and forage yield fluctuations on robustness indicators with a global sensitivity analysis (GSA) of a herd model. Contrary to a one factor at a time approach, GSA allows the exploration of the whole space of parameters and their interactions. With this tool, we explored which management strategies better balance robustness indicators and if they are the same depending on feeding system. Seven parameters were studied: herd milk production potential, breeding season length, reproductive success, replacement rate, culling criteria, extended lactations management and concentrate supplementation level. A simulation plan was defined to optimize parameters space exploration and computation time. Simulations were run for contrasting feeding systems (e.g. maize silage, alfafa hay, pasture grazing). To reduce computational cost and allow GSA, Kriging models were fitted to simulation data to obtain cheap-to-evaluate surrogates of model outputs. Sensitivity indices of outputs, related to robustness to price and forage yield fluctuations, were then computed using the kriging models. They quantified the contribution associated to each input (management parameters) to the variance of robustness indicator. As a complementary tool, a user-friendly interface was developed to allow the visual exploration of robustness indicators depending on feeding systems and management parameters. This study emphasizes the benefits of combining mathematical and computer methods with systemic modelling to deepen our understanding of trade-offs and synergies and quantify them.