A Replication-environment is a second database-server having exact same data as the origin in a near-live fashion Usually runs behind it's origin database-server by several minutes. Replication can be useful for taking a load of the production server by having the non-time-critical read actions not done on the production server, but on this replicated server. For instance the (often heavy) queries of BI-systems. Another is having an up-to-date acceptance-environment. There are more use cases for having a replication-environment.
The Fallback-environment is where your application will flee to when the original environment is dead. Usually has to be up-and-running within a contractually set period of time, usually a pretty short period. To confuse you: Replication can also be used to keep a server stand-by to flee to. But that is just one possible implementation for the fallback-environment.
My to cents;
Replication is about copy (replication) of data.
Fallback is about having a fallback database in case of main DB failure. This database of course should be a replication of the main Db. But having a replicated db doesn't mean that it is the fallback.
Fail over is only available on the enterprise edition. not sure if this is limited to production. see; https://www.mendix.com/pricing/