In this paper, we introduce novel probabilistic transmission protocols for data intensive wireless sensor networks. The protocols combine a randomized transmission scheme with scheduling based on heuristics to alleviate network performance degradation due to excessive collisions and retransmissions. In particular, by utilizing local information about sensor neighbors, our protocols minimize collisions in areas with high channel contention. The simulation results of our hybrid techniques show improved performance of data delivery and reduced energy consumption compared to carrier-sensing based IEEE 802.15.4 CAP mode and time-division based DRAND.