We investigate distributed combining techniques for distributed detection in wireless sensor networks (WSNs) over Rayleigh fading multiple access channel (MAC). The MAC also suffers from with path loss and additive noise. The WSN is modelled as a Poisson point process (PPP). Two distributed transmit combining techniques are proposed to mitigate fading; distributed equal gain transmit combining (ddEGTC) and distributed maximum ratio transmit combining (dMRTC). The performance of the previous methods is analysed using stochastic geometry tools, where the mean and variance of the detector's test statistic are found thus enabling the fitting of the received signal distribution by a log-normal distribution. Surprisingly, simulation results show a that ddEGTC outperforms dMRTC.