If you are writing two-voice species counterpoint according to Fux's rules,
and the cantus firmus is in the bottom voice
and that bottom voice descends by step to the final in the Phyrgian mode--that is, the bass moves down a minor second from F to E in an untransposed mode;
then the upper voice in two-voice counterpoint sings D--E, because the final octave consonance must be preceded by a major sixth.
A D# in the upper voice would create an augmented sixth, which Palestrina did not use.
These are Fux's rules, based on his analysis of Palestrina. They are akin to a description of Latin grammar based on Cicero. They describe idioms that were used in Renaissance polyphony, but there is not necessarily any logic to them beyond that.
Further, Fux's counterpoint has nothing to do with scales in jazz, or with the harmony of post-1800 tonal music. Fux is not describing "vertical" harmonic progressions but "horizontal" contrapuntal movements.