A large number of WSN applications with bursty traffic have emerged in the recent past such as rare event detection and forest fire monitoring. Bursty traffic in Wireless Sensor Networks (WSNs) offers various challenges in terms of reliability and latency. Unlike periodic monitoring, energy consumption is far more challenging to deal with in WSN applications with bursty traffic due to long periods of inactivity followed by certain bursts. In this paper, we study the influence of using dynamic channel polling intervals for dealing with the bursty traffic. We refer to dynamic polling as the polling process based on probabilistic distribution chosen in correspondence to the incoming traffic. A simple asynchronous MAC protocol has been developed to illustrate the difference in WSN performance when deterministic, exponential and dynamic polling interval distributions are used. The experiments have been conducted using TinyOS source code and Avrora emulator to evaluate the proposed scheme in terms of energy consumption and latency. Performance improvement brought to WSN through dynamic polling has been validated by comparing the results with those of SCP-MAC. It has been found that the MAC with the dynamic polling intervals outperforms deterministic and exponential polling distributions as well as SCP-MAC for the bursty traffic.