Or, it is also possible, that in some areas voter access to the polls was disrupted/ voters were intimidated. That is something a military/security campaign can do.
But turnout was up, not down, including in Kurdish areas.
And here it would be interesting to see the soldiers. Or, for that matter, intimidation may also express itself in increased turnout. I would expect, for instance, increased turnout by government employees.
It would be interesting to see a more detailed breakdown of votes: by precinct, by locality, etc. It would help to answer some questions.