How to avoid intermittent "duplicate entry" error on slave when replicating mysql.time_zone_transition?

Rarely, but intermittently, MariaDB 5.5 fails to replicate the timezone tables onto a completely new replica (instance creation is all automated, so there's no room for missing/changed steps). We never change these internal tables, so how can this happen?

Example result (formatted for readability):

$ mysql -e 'SHOW SLAVE STATUS\G'
[…]
           Master_Log_File: bin-log.000011
       Read_Master_Log_Pos: 5124
            Relay_Log_File: relay-log.000002
             Relay_Log_Pos: 527
     Relay_Master_Log_File: bin-log.000010
          Slave_IO_Running: Yes
         Slave_SQL_Running: No
                Last_Errno: 1062
                Last_Error: Error 'Duplicate entry '339--1956609120' for key 
'PRIMARY'' on query. Default database: 'mysql'. Query: 'INSERT INTO 
time_zone_transition (Time_zone_id, Transition_time, Transition_type_id) […]