Parallel seeding effect

[Q&A-Freak]

Hello!

The following scenario:
One seeds the same torrent twice from the same computer. The underlying files are identical.

First seed with i2psnark.
Second seed with BiglyBT.

I made the typical observation that the (upstream) bandwidth of first seed (i2psnark) is systematically poorer compared to that of the second seed (BiglyBT).

What might be the cause?

Best regards