The files hold the data for the paper entitled "Density-dependent reproductive success among sympatric dipterocarps during a major mast fruiting event", which includes (i) the weight and status of seeds collected with seed traps in a Bornean tropical forest during a mast fruiting event, and (ii) the results of a seed sowing experiment in primary and secondary forests. Study site was the Deramakot and Tangkulap Forest Reserves, Sabah, Malaysia (5°14– 30’N, 117°11–36’E).

Description for the data (i)

A 2-ha vegetation plot (100 m × 200 m) was established in 2006–2008 in a primary forest in the Deramakot Forest Reserve. Within the 2-ha plot, ten seed traps were arranged at 20-m intervals along each of the five parallel 200-m lines. Five additional traps were installed at the center of the plot (i.e., 15 traps at the center line, see Fig S1 of Aoyagi et al., 2018). The seeds captured by the traps was collected biweekly (from April 25, 2010 to December 11, 2012). Collected samples were oven-dried at 60 ℃. For data from June to September, samples were identified to species level, and the status of the seeds was recorded. Seed fate was classified as mature seeds, aborted seeds (apparently smaller than mature seeds), seeds predated by insects (a seed with an insect hole or a seed with dead insect larvae within it), and broken seeds predated by unknown predators (possibly arboreal vertebrates). Seeds of different species and statuses were weighed separately. From July 13 to 27, 2010, data on seed production of each species were lost.

Description for the data (ii)

We conducted a seed-sowing experiment with and without vertebrate exclusion. Dipterocarp seeds, including S. multiflora, P. malaanonan, S. atrinervosa, S. macroptera, Dryobalanops lanceolata, S. mecistopteryx, and S. parvifolia were collected from forests near the primary forest plot three days before starting the experiment (September 7–12, 2010). Seeds with fungal and insect attacks or those with apparently smaller sizes than other conspecific seeds were discarded before the experiment. In addition, the seed wings were cut off before the experiment because wings of some species are too large to put in cages to exclude vertebrates.

Because community-level seed density significantly influences dipterocarp seed survival, we conducted the experiment at three sites with contrasting seed densities: (1) the primary forest plot (high seed density; Deramakot), (2) forests within 500-m distance from the edge of the primary forest plot, but with no dipterocarp seed production at least 50 m around the site (low seed density), and (3) the secondary forest plot (low seed density; Tangkulap). Three replicates with no canopy damage, and no understory grass or shrubs were selected within each site and were at least 50-m apart from each other. Naturally growing small seedlings were removed. Cages made of iron wire (60 × 60 × 60 cm) with a mesh size of 1×1 cm were used. These cages can exclude small-to-medium-sized predators, such as deer and wild boars, but insects and pathogens can attack the seeds inside the cages. For each replicate, three groups of seeds were sown for each species. Group 1 was protected with cages from vertebrate predators throughout the experimental period (denoted as Protected)., while group 2 was not protected (denoted as Out). Group 3 was protected first, and the cage was removed after 90% of the seeds germinated and stood alone (denoted as Cotyledon_out). Thus, the impact of small-to-medium-sized predators on the survival rate of seed-stage or early cotyledon stages could be quantified by comparing the survival rates of groups 1 and 2 or between the survival rates of groups 1 and 3, respectively. All surviving seeds germinated within 12 days, and the established seedlings were re-censused after five (February 14–20, 2011) and ten months (July 5–16, 2011).