Concerns about habitat difference interpretation

[Aariq]

R2:

• [x] 7) Delayed effects: Since the authors fit separate models for plants in fragments and in continuous forest to allow for the crossbasis function to differ between habitats we don’t know if any of the small differences reported in Fig 4c, 5c and 6c are statistically significant. Hence, I would recommend interpreting the results concerning the difference between fragments and CF with caution. I would like to see at least a justification for why representing the difference figures are meaningful.

• [x] 9) Line 323-325: I don’t know why the growth of plants in fragments would affect the size of plant in CF.

• [x] 10) Lines 355-359: The analysis did not consider testing a statistical difference between habitat type (see comment 7), so I think this sentence is misleading (overinterpreting the results).

• [x] 11) Line 370: There are no confidence intervals in Fig 2, so we don’t know if the difference is significant.

• [x] 12) Lines 383 ff: see comment 7

• [x] 14) 424-427: I think this is exaggerating since the differences between habitat types is not significant.

• [x] 22) Lines 484-486: I don’t think this study strongly support that plants do worse in fragments than CF.

[Aariq]

Figure 2 (survival curve of 1998 "cohort") is not something that there can be confidence intervals around. Also, line 370 makes no claims about statistical significance.

[Aariq]

change declarative statements ("are different") to less declarative ("strongly suggests differences", "is consistent with differences").