I'm skeptical that this question can be asked and answered in a meaningful manner. Do we really want to know which area of mathematics produces, say, the most papers? Or even the most citations? What might be more meaningful (but maybe not) is which fields get the most funding from NSF.