Your effect is most likely currently setup to be applied on the voice level (in the actor mixer hierarchy, on your sequence container or on one of its direct parent), meaning the tail of the effect is part of the source length.
When mixing many sounds into one environment, you should try to put your effect on a bus and route your container into it. All the voices of your container would be mixed into one bus, that would make things work as you expect them to and as a bonus, if the effect was the same for all your sounds in the container, they would be mixed before applying the effect, leaving one effect being processed instead of multiple at one time, saving some Reverberation CPU and memory.