Autonomous wireless agents are deployed in difficult to approach environments in order to gather information for example temperature, pressure or pollution distributions. Although the size and cost of agents decrease, their batteries are generally insufficient for the desired measurement duration. In this study, adaptive sampling techniques are considered to prolong the agents operation time. This is achieved via adaptations of the agents' sampling periods while minimizing the information loss. Accordingly, exponential double smoothing adaptive sampling (EDSAS) is operated and Wiener filter based adaptive sampling (WFAS) is proposed considering the limited energy capacity of the agents. This study demonstrates the performance of deterministic and stochastic adaptive sampling strategies for real data in terms of reconstruction error, achieved data reduction rate and minimum sampling rate. Simulation results show that agents are able to decrease their sampling rate up to 10 which is equivalent to achieve a data reduction around 80 percent. In addition, it is also demonstrated that the proposed WFAS algorithm is superior to the EDSAS algorithm in terms of improved data reduction rate, and correspondingly, energy efficiency through relaxed error tolerances while maintaining a reasonable reconstruction performance.