We examine mechanisms leading to organization of vegetation patterns within the channel network structure of a semi-arid New Mexico river basin under the controlling influence of water stress. We compare the actual pattern of water stress within the basin to patterns resulting from two algorithms of local stress optimization which proceed from an initial fully random vegetation distribution. Here we show that the distribution of vegetation and basin water stress derived from an algorithm that maintains local optimization within the network flow path exhibits considerably better agreement with the actual distribution than one that ignores the network structure of the basin. These results suggest the pattern of actual vegetation observed within the basin corresponds to a condition of feasible optimality in which organization is constrained by the stochastic nature of local interactions mediated by the network configuration. The principles of such organization have important consequences regarding the interaction between land cover change and hydrological dynamics in river basins, as well as the biogeographical evolution of landscapes.