In the Arid Chaco savannahs, the presence of permanent water sources for livestock affects the spatial patterns of grazing, generating a gradient of grazing intensity that affects the composition of plant communities and the functioning of the ecosystem. The characteristics of plant species give rise to species-specific responses to environmental stimuli. These characteristics can group species into different functional plant groups that determine the effect of various disturbances on the functioning of the ecosystem. In this context, our study proposes the hypothesis that continuous grazing associated with permanent water sources generates a cumulative livestock pressure that determines different ecological conditions for plant communities in different geomorphological units. Under this hypothesis, we make the following predictions: 1) The distribution of water sources directly and indirectly influences the spatial distribution of functional plant groups, and 2) Variation in the quality of forage resources is associated with the distribution of some functional plant groups. Five plant functional groups were defined: annual nonfodder herbaceous plants, perennial nonfodder herbaceous plants, fodder woody plants, fodder herbaceous plants, and nonfodder woody plants. The plant functional groups significantly affected were those related to the production of fodder herbaceous plants and nonfodder woody plants. The relationship between the contribution of the functional groups of plants and the forage quality of the plant communities in the geomorphological units studied shows that forage grasses have a positive relationship and nonforage woody plants have a negative relationship. At distances greater than 1 250 m in the alluvial plain and 1 900 m in the foothills, the growth of forage species increase significantly. This can serve as a reference for planning the distribution of water sources for livestock in semi-arid plains, such as the area studied. (c) 2025 The Society for Range Management. Published by Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.