[Dataset:] Estimates of seed production per reproductive basal area per year for 38 Barro Colorado Island tree species for the period 1993-2012

This seed production dataset contains information on total yearly seed captures and estimates of total reproductive basal area and seed production per mm2 basal area for 38 tree species from Barro Colorado Island, Panama. Seed captures are observed in seed traps located in a stratified random manner along trails within the 50-ha Forest Dynamics Plot (FDP) on Barro Colorado Island, Panama (the southwest corner of the FDP is located at 625773 easting and 1011774 northing in UTM Zone 17). The 50-ha FDP is a 1000 by 500 meter rectangle in which all woody stems greater than 1 cm in diameter (at 1.3 meter height) have been identified and mapped (see dataset: DOI: 10.5479/data.bci.20130603). The total observed yearly seed rain (yearly flux of arriving seeds; column Seeds_year) was recorded in 200 0.5 m2 seed traps and contains counts from January 1st to December 31st from 1993 to 2012. Seeds were identified to species and counted weekly (see Wright et al. 1999). The total reproductive basal area (RBA_year) represents the sum of the stem area in that year (at 1.3 meter height) of all reproductive individuals within the FDP plot. RBA_year was calculated from the FDP tree census data (DOI: 10.5479/data.bci.20130603) in combination with the fitted logistic models (GLMMs with binomial error; see Visser et al, in press) that predict a size- dependent probability of reproduction. Models were fit to dataset 5 from Visser et al (in press), see “The Barro Colorado Island Tree Reproduction Dataset” for details (DOI: XXXXX). The logistic models predicted each individual’s reproductive probability as a function of its size. Each individual's basal area is then weighted by its reproductive probability to calculate total reproductive basal area. Total reproductive basal area was interpolated between forest dynamics plot censuses (see dataset: DOI: 10.5479/data.bci.20130603). Species-specific seed production (column sp_est; seeds per year per mm2 of reproductive basal area) was quantified as the mean flux of seeds arriving (seeds per year per m2 of trap area) divided by mean reproductive basal area density (mm2 of reproductive basal area per m2 of plot area). We used seed trap and tree census data from 1993 through 2010.本种子生产数据集涵盖巴拿马巴罗科罗拉多岛（Barro Colorado Island）38个树种的年度总种子捕获量、总繁殖基面积估算值，以及每平方毫米繁殖基面积的种子生产量信息。 种子捕获量数据源自设置于50公顷森林动态样地（Forest Dynamics Plot, FDP）内步道沿线的分层随机布设种子陷阱。该森林动态样地位于巴拿马巴罗科罗拉多岛，其西南角采用UTM 17投影带坐标系，东向分量为625773，北向分量为1011774。 该50公顷森林动态样地为1000米×500米的矩形区域，样地内所有胸高（1.3米高度处）直径大于1厘米的木本茎干均已完成物种鉴定与空间测绘（相关数据集参见DOI: 10.5479/data.bci.20130603）。 年度总观测种子雨（即年度到达种子通量；对应数据列Seeds_year）的记录覆盖200个0.5平方米的种子陷阱，统计时段为1993年1月1日至2012年12月31日。研究人员已对捕获种子完成物种鉴定，并按周进行计数（参见Wright等人1999年的研究）。 总繁殖基面积（RBA_year）指该年度森林动态样地内所有繁殖个体的胸高茎干基面积之和。该指标通过森林动态样地树木普查数据（DOI: 10.5479/data.bci.20130603）结合拟合的逻辑回归模型计算得到：该模型为带有二项分布误差项的广义线性混合模型（Generalized Linear Mixed Models, GLMMs），可预测个体繁殖概率的体型依赖性，相关方法参见Visser等人即将发表的研究。模型拟合自Visser等人（即将发表）的数据集5，详细信息参见《巴罗科罗拉多岛树木繁殖数据集》（DOI: XXXXX）。该逻辑回归模型以个体体型为自变量预测其繁殖概率，随后将个体基面积乘以对应繁殖概率得到加权值，最终求和得到总繁殖基面积。总繁殖基面积通过森林动态样地的历次普查数据进行插值补全（相关数据集参见DOI: 10.5479/data.bci.20130603）。 物种特异性种子生产量（对应数据列sp_est；单位：每年度每平方毫米繁殖基面积的种子数）的量化公式为：种子到达平均通量（每平方米陷阱面积的年度种子数）除以平均繁殖基面积密度（每平方米样地面积对应的繁殖基面积平方毫米数）。本数据集使用了1993年至2010年的种子陷阱与树木普查数据。