Widespread tree defoliation and mortality rates have increased considerably in recent decades. We investigated the decline pattern of Castanopsis fargesii and examined the effects on conspecific seedling regeneration. This dataset contains the basic characteristics and tree vigor level data of 617 C. fargesii individuals (H >1.3m, identified in nine plots in a national nature reserve in Chongqing in Southwestern China), and the seed traits data of the seeds produced by decline adult trees and non-decline adult trees.

In September 2013, we established a 0.5 ha (50 m × 100 m) core monitoring plot and eight 30 m × 30 m sampling plots that were dominated by C. fargesii to monitor the long-term dynamics of declining forests on Jinyun Mountain (29°49′N, 106°20′E), a national nature reserve in Chongqing in Southwestern China. In September 2013, we examined the decline patterns of C. fargesii adult trees within the core monitoring plot and the eight sampling plots. Within each plot, we tagged and measured diameter-at-breast height (DBH) and tree height (H) of all woody trees ≥ 1.3 m tall. For C. fargesii adult trees (H ≥ 8 m; DBH ≥ 10 cm) and saplings (1.3 - 8 m) in all plots, we determined the vigor levels based on the percentage of crown defoliation, using a semi-quantitative scale. Six tree vigor levels were defined for this study: L0: dead tree; L1: 81-100% crown defoliation; L2: 61-80% crown defoliation; L3: 41-60% crown defoliation; L4: 21-40% crown defoliation; and L5: 0-20% crown defoliation. We classified L0-L1 as a high level of decline, L2-L3 as an intermediate level of decline, and L4-L5 as non-decline. In September 2018, we again determined the vigor levels of all of the adult trees that we had marked in 2013 to evaluate their decline trends.

We investigated seed production and seed traits of the 30 selected C. fargesii adult trees using seed traps. The seed traps were made of a funnel of polyethylene cloth (1 mm mesh), with a receiving area of 0.5 m2 (1 m × 0.5 m) and a height of 1 m above the ground. Considering the terrain, crown shape, and vine coverage, we set up two or three seed traps in each subplot at a distance of 1.5 m from each adult tree (72 seed traps in total, Fig. 1). We fully considered the distance between two selected adult trees to ensure that a certain distance (> 3 m) was present between the seed traps and the adjacent adult tree canopies. In addition, C. fargesii seeds are about 0.5 g, and the horizontal dispersal distance is very close by seed rain. Therefore, the risk that seeds in the seed traps had come from adjacent adult trees was very low.

The seed traps were made of a funnel of polyethylene cloth (1 mm mesh), with a receiving area of 0.5 m² (1 m × 0.5 m) and a height of 1 m above the ground. Considering the terrain, crown shape, and vine coverage, we set up two or three seed traps in each subplot at a distance of 1.5 m from each adult tree (72 seed traps in total). We fully considered the distance between two selected adult trees to ensure that a certain distance (> 3 m) was present between the seed traps and the adjacent adult tree canopies. In addition, C. fargesii seeds are about 0.5 g, and the horizontal dispersal distance is very close by seed rain. Therefore, the risk that seeds in the seed traps had come from adjacent adult trees was very low.

We combined the mature seeds of each adult tree collected every week. Among the combined seeds, we randomly selected 30 intact mature seeds from each of the 11 non-declining adult trees and 10 intermediate declining adult trees to measure their seed traits, including fresh weight (seed mass), nutritional (starch), and defensive (cellulose and tannin) material properties. We did not measure the seed traits of high declining adult trees, because the collected seeds were too few to sample. Starch content was estimated using the optical rotation method following the national standard (NY/T11-1985) for cereals. Cellulose content was estimated using the acid and alkali washing method following the national standard (NY/T13-1986) for cereals. The tannic acid content was determined using the spectrophotometry method following the national standard (GB/T15686-2008) for sorghum.