In this paper we solve a decentralized nonlinear control problem where actuator valves and flow rate sensors are collocated in individual branches and do not exchange information. This is in contrast to our previous paper where a centralized controller required measurements from all the branches of the network. We solve both regulation ( constant references) and tracking ( time varying reference signals) problems. To eliminate conservativeness in choosing the gains of the controllers, we employ adaptation. We illustrate the results with an example reminiscent of a blood flow network.