I like to document some simulation results done in BRAHMS with response at high rapidity using HIJING simulations
The data are from Hiro  Ito's thesis (from U Kansas, now at RACF). You can see at eta~3 the observed yield is about a factor of 2 larger than the input from charged particles.
In addition the EPD has lost of material from the TPC support structures, and a larger value can be expected.