African savannas are characterised by temporal and spatial fluxes that are linked to fluxes in herbivore populations and vegetation structure and composition. We need to be concerned about these fluxes only when management actions cause the system to shift towards a less desired state. Large herbivores are a key attribute of African savannas and are important for tourism and biodiversity. Large protected areas such as the Kruger National Park (KNP) manage for high biodiversity as the desired state, whilst private protected areas, such as those adjacent to the KNP, generally manage for high income. Biodiversity, sustainability and economic indicators are thus required to flag thresholds of potential concern (TPCs) that may result in a particular set of objectives not being achieved. In large conservation areas such as the KNP, vegetation changes that result from herbivore impact, or lack thereof, affect biodiversity and TPCs are used to indicate unacceptable change leading to a possible loss of biodiversity; in private protected areas the loss of large herbivores is seen as an important indicator of economic loss. Therefore, the first-level indicators aim to evaluate the forage available to sustain grazers without deleteriously affecting the vegetation composition, structure and basal cover. Various approaches to monitoring for these indicators were considered and the importance of the selection of sites that are representative of the intensity of herbivore use is emphasised. The most crucial step in the adaptive management process is the feedback of information to inform management decisions and enable learning. Feedback loops tend to be more efficient where the organisation’s vision is focused on, for example, economic gain, than in larger protected areas, such as the KNP, where the vision to conserve biodiversity is broader and more complex. Conservation implications: In rangeland, optimising herbivore numbers to achieve the management objectives without causing unacceptable or irreversible change in the vegetation is challenging. This manuscript explores different avenues to evaluate herbivore impact and the outcomes of management approaches that may affect vegetation.