This paper presents a self-organizing wireless sensor network that can be used for environmental surveillance tasks, like wildlife behavioural studies. The parameters of interest that are measured and recorded are temperature, humidity, illumination and the motion of living objects inside the span of the network. An accurate time-synchronization algorithm has been implemented and tested successfully. A vital task during the whole development process was to keep the power consumption to a minimum, since the used nodes are battery powered. This paper shows that the TelosB mote combined with the TinyOS operating system provides an easy to implement infrastructure establishment. The TelosB has been extended with additional hardware without disturbing the low power operation. The extension board can be fully controlled via software and has a sleep state where the current consumption drops to less than 0.1 A.