In this paper, simultaneous lower bounds on space and input head reversals for deterministic, nondeterministic and alternating Turing machines accepting nonregular languages are studied. Three notions of space complexity, namely strong, middle, and weak, are considered; moreover, another notion called accept, is introduced. For all cases we obtain tight lower bounds. In particular, we prove that while in the deterministic and nondeterministic case these bounds are \u201cstrongly\u201d optimal\u2014in the sense that we exhibit a nonregular language over a unary alphabet exactly fitting them\u2014in the alternating case optimal lower bounds for tally languages turn out to be higher than those for arbitrary languages