This is a well-written, interesting, and useful contribution toward better understanding legacy CDD and Janzen-Connell effects. The empirical analyses appear sound and the modeling insights appears well-justified and well-reasoned. My only minor suggestions for improvement are to: Lines 195 (a period and space between sentences need to be added); 218 (a period needs to be added to end the sentence); and 224 (the parentheses need to be closed).