Data from: Does functional trait diversity predict aboveground biomass and productivity of tropical forests? Testing three alternative hypotheses

1. Tropical forests are globally important, but it is not clear whether biodiversity enhances carbon storage and sequestration in them. We tested this relationship focusing on components of functional trait biodiversity as predictors. 2. Data are presented for three rain forests in Bolivia, Brazil and Costa Rica. Initial above-ground biomass and biomass increments of survivors, recruits and survivors + recruits (total) were estimated for trees ≥10 cm d.b.h. in 62 and 21 1.0-ha plots, respectively. We determined relationships of biomass increments to initial standing biomass (AGBi), biomass-weighted community mean values (CWM) of eight functional traits and four functional trait variety indices (functional richness, functional evenness, functional diversity and functional dispersion). 3. The forest continuum sampled ranged from ‘slow’ stands dominated by trees with tough tissues and high AGBi, to ‘fast’ stands dominated by trees with soft, nutrient-rich leaves, lighter woods and lower AGBi. 4. We tested whether AGBi and biomass increments were related to the CWM trait values of the dominant species in the system (the biomass ratio hypothesis), to the variety of functional trait values (the niche complementarity hypothesis), or in the case of biomass increments, simply to initial standing biomass (the green soup hypothesis). 5. CWMs were reasonable bivariate predictors of AGBi and biomass increments, with CWM specific leaf area SLA, CWM leaf nitrogen content, CWM force to tear the leaf, CWM maximum adult height Hmax and CWM wood specific gravity the most important. AGBi was also a reasonable predictor of the three measures of biomass increment. In best-fit multiple regression models, CWMHmax was the most important predictor of initial standing biomass AGBi. Only leaf traits were selected in the best models for biomass increment; CWM SLA was the most important predictor, with the expected positive relationship. There were no relationships of functional variety indices to biomass increments, and AGBi was the only predictor for biomass increments from recruits. 6. Synthesis. We found no support for the niche complementarity hypothesis and support for the green soup hypothesis only for biomass increments of recruits. We have strong support for the biomass ratio hypothesis. CWMHmax is a strong driver of ecosystem biomass and carbon storage and CWM SLA, and other CWM leaf traits are especially important for biomass increments and carbon sequestration.

1. 热带森林在全球尺度上具有重要生态价值，但目前尚不清楚生物多样性是否能够提升其碳储存与固碳能力。本研究以功能性状生物多样性的组成要素作为预测变量，检验了二者之间的关联。
2. 本数据集涵盖玻利维亚、巴西与哥斯达黎加的三处热带雨林。研究针对胸径（diameter at breast height, d.b.h.）≥10厘米的林木，分别在62个与21个1.0公顷样地中，估算了其初始地上生物量，以及存活木、补充木与存活木+补充木（总计）的生物量增量。我们分析了生物量增量与初始立地地上生物量（AGBi）、8种功能性状的生物量加权群落均值（community weighted mean, CWM），以及4种功能性状多样性指数（功能丰富度、功能均匀度、功能多样性与功能离散度）之间的关联。
3. 所采样的森林群落连续体涵盖两类林分：一类是以组织坚韧、初始立地地上生物量（AGBi）较高的林木为优势的“慢生”林分，另一类是以叶片柔软、养分丰富、木材密度较低且初始立地地上生物量（AGBi）更低的林木为优势的“速生”林分。
4. 本研究检验了三种假说：一是初始立地地上生物量与生物量增量与系统中优势物种的功能性状群落均值相关，即生物比假说（biomass ratio hypothesis）；二是与功能性状值的多样性相关，即生态位互补假说（niche complementarity hypothesis）；三是就生物量增量而言，仅与初始立地地上生物量相关，即“绿色浓汤假说”（green soup hypothesis）。
5. 功能性状群落均值（CWM）可较好地作为初始地上生物量（AGBi）与生物量增量的预测因子，其中比叶面积（specific leaf area, SLA）的CWM、叶片氮含量的CWM、叶片撕裂力的CWM、最大成年树高（maximum adult height, Hmax）的CWM以及木材比重的CWM为最重要的预测因子。初始立地地上生物量（AGBi）同样可较好地预测3种生物量增量指标。在最优多元回归模型中，最大成年树高（Hmax）的CWM是预测初始立地地上生物量（AGBi）的最关键因子。针对生物量增量的最优模型仅选取了叶片性状作为预测因子，其中比叶面积（SLA）的CWM为最重要的预测因子，且二者呈现预期的正相关关系。功能性状多样性指数与生物量增量之间未发现显著关联，而初始立地地上生物量（AGBi）是唯一可预测补充木生物量增量的因子。
6. 综合分析结果表明：本研究未发现支持生态位互补假说的证据，仅在补充木的生物量增量方面找到了支持绿色浓汤假说的依据；而对生物比假说则有强有力的支持。最大成年树高（Hmax）的CWM是驱动生态系统生物量与碳储存的关键因子，比叶面积（SLA）的CWM及其他叶片功能性状的CWM则对生物量增量与碳固存尤为重要。