The experimental design in this study is based on the integration of two databases, which needs some consideration in interpreting the results of this work. We censused the plots from which we obtained the data used to model survival and growth from 2003 to 2016. In addition, for this study, it was necessary to gather data on the presence and number of reproductive structures in order to model fertility. These data had not been previously recorded, thus in 2017 we sampled these variables in the field. In that year, rains were recorded after several relatively dry years, possibly associated with ENSO (Muñoz et al., in revision). Therefore, it is important to note that the environmental conditions may have been very different between the two sets of data that fed the different models.

We carried out fruit counting and sampling in November, based on the phenological behavior of the species at the study site (Maldonado, 2014). During fieldwork, we observed ripe fruits and some fruits on the ground. Therefore, it is possible that the number of fruits counted may be an underestimation of the actual number produced, and thus underestimate f ₂. No laboratory germination tests were done for this species and no studies are known to have done so, thus the conversion factor from seeds to seedlings was calculated using inverse estimation (González, Martorell, & Bolker, 2016). Very likely, the inclusion of an in situ germination study would produce better germination estimates; however, inverse estimation is a good approach in the absence of such studies.

Similarly, it is important to keep in mind that the database used to model growth and survival is much larger than that for fecundity, making it more difficult to identify outliers in fecundity. Therefore, we cannot rule out that the low values observed in fecundity may be related to all of the above. In order to better understand reproduction, it is important to have a larger database and to consider different years of sampling in order to have greater control over interannual variation.