DB2 requires that a unique index exist on the primary key. If a unique index does not already exist when the PK is defined, then DB2 will create one for you. If one already exists, then DB2 will use it.
Some other databases may just do a table scan making sure there are no duplicates if no unique index exists on the PK, but DB2 always uses the unique index to do make sure there are no duplicates.
You can drop the index if you first remove the PK.
M. A. Feldman
IBM Certified DBA on DB2 for Linux, UNIX, and Windows
IBM Certified DBA on DB2 for z/OS and OS/390