In the study area, a set of 17 permanent plots were established in 2003 to represent a chronosequence, as the times of abandonment (successional ages) in them differed, ranging in that year between 0 and more than 50 yr (Brienen et al. 2009); thereafter, the permanent plots have been used to monitor community development through annual censuses of the vegetation. Each year, individual height (m) and survival are recorded following the sampling protocol described by Lebrija-Trejos et al. (2008) and Muñoz et al. (2021); briefly, data are gathered in four 20 × 5 m transects, separated by 2 m wide corridors, each divided into four 5 × 5 m (25 m²) quadrats. At each plot, all M. acantholoba ²²²² subquadrats randomly placed in the transects (32 m²). Data of the different subquadrats were re-scaled in order to express all densities in the same area (see Muñoz et al. 2021).

We recorded reproduction data in November 2017 in 11 of the 17 permanent plots. For each individual, we recorded reproductive status and fruitset in one of the main stems. We extrapolated the fruitset per reproductive individual by multiplying the latter figure by its number of main stems. We collected a sample of 98 fruits from 13 trees, covering tree size variation, and counted the number of seeds per fruit.