Chronosequence resampling elucidates tree community and forest structure recovery patterns in restored tropical rainforest

We investigated the forest structure, tree diversity and community in an actively restored rainforest in Kibale National Park, East Africa, Uganda. Vegetation surveys were done in 2013 and 2021. The study area consisted of actively restored compartments planted in 1995, 1999, 2000, 2005, 2007, 2008, 2011a and 2017a (a = surveyed only in 2021) and primary forest as a reference site (altogether 55 study sites). 

In the datasets, we present: 
(1) the tree community data (total estimated stem density / ha)
(2) univariable data for each study site (forest structure and diversity variables, and metadata of the study sites)
(3) DBH data for planted trees.

Methods
We used a nested plot design to record:
(1) large trees (20 x 40 m plot, diameter ar breast height (DBH) ≥ 20 cm),
(2) medium trees (20 x 20 m plot, DBH ≥ 10 and < 20 cm)
(3) small trees (20 x 10 m plot, DBH ≥ 5 and < 10 cm), and
(4) saplings (10 x 10 m plot, DBH < 5 cm; including seedlings).
Plots shared one corner. We identified each stem to the species or genus level (if possible) and measured the DBH of small, medium and large trees. We calculated the estimated total basal area (m2/ha), the estimated total stem density (stems/ha), the taxa richness (count of the total tree taxa per study site), and the community similarity of the study site to the mean community similarity of primary forests (%). In 2021, we also recorded the mean tree height (m), canopy cover (%), and understorey vegetation covers (%) of Acanthus pubescens, Lantana camara, Cenchurus purpureus, and Marantochloa spp. in the 20 x 40 m plots.