How to interpret the changes in the decision tree for subgroups?

[epark21]

I was examining if gender predicts the target using decision tree in SAS VA. Gender did not appear in the resulted tree and in the variable importance chart. However, if I run the same tree for each gender subgroup (i.e., one for male and one for female), the predictors in the resulted tree  were different for male and female. How do we interpret this? Does gender matter or not?

Thanks,