This paper investigates Wireless Sensor Networks (WSNs) with mobile sinks that collect data from sensors by following uncontrolled trajectories. In particular, the paper focuses on proactive data dissemination strategies in which the trajectory of the mobile sink is unknown to the sensors. These strategies attempt to obtain a good trade-off between the number of sensors the mobile sinks has to visit in order to collect representative data of all sensors, and the communication effort required by the sensors. All the investigated techniques also avoid the use of multi-hop routing, due to its high cost. Some of them are based on random walks whereas others are based on a combination of probabilistic forwarding and probabilistic storage. An analysis of the various methods in terms of the performance trade-off between the efficiency of the data gathering and communication overhead is also presented.