genitive?

If one take a look at the maps with Dacian "dava"-s one will observe
imediately that the names are compounds of a certain word plus the
well-known "dawa".
The curiosity appears to be that all the names are ending in "i" and "o"
before the "dava". Examples with "i"
Singidava, Ziridava, Buridava, Piroboridava, Sacidava, Arcidava,
Utidava, Cumidava, Carsidava, Capidava,Buteridava, etc.

Expections with "o":
Petrodava, Marcodava

In the properly Thracian Space (South of Balkans) there appears to be
the same "i" and "o" where predmoninantely should be these which ends in
"o" comparative with the Dacian where predominant are these which ends
in "i". Even if in the Thracia propria dictam there is not "dava" the
widest compound part but mostly "para" the way they are "bound" appears
,how I said prior, to be the same:

Examples with "i":
Bealidipara, Bendipara, Beripara, Hypara, Drusipara,Autipara, etc.

Examples with "o":
Spinopara, Skaptopara, Berentopara, Longinopara, Stratopara, etc.

Studying the map one will observe that sometimes the "o" alternate with
"u":
Desudaba, Subzupara, Pulpudeva, Gelupara, Subzupara.

OK, for sure can be considered we have once endings in "i" and once in
"o/u"
Can these endings, "-i" and "-o, -u" considered to be genitival markers
making compounds like "dava of Marco"=Markodava and "para of Bendia" =
Bendipara?
If yes, then, .. can one consider the "-i" is a steme for feminine (
because Bendis was a feminine diety) and "-o/-u" can be considered it
has been a stem for genitival masculine ?

Alex