Symmetry in variance map?



I calculated a variance map from 1000 bootstrap volumes, my dataset was downsampled by a factor of 2 with resample_mp.exe, and the resolution of the reconstructed bootstrap volumes was limited to 8 A. And I imposed C4 symmetry. FBEAUT was set to F.

In the resulting variance map I get a strong "column" spanning all slices in the very middle of the box, parallel to the Z-axis. I understand this as a symmetrization artifact (no symmetrization is imposed on the central voxel column?), but not sure. The rest of the variance map seems to make sense.

Top view:
Side view:

Any help in how to interpret this, and how to correct this if necessary, is very much appreciated.

Thanks a lot!