Using Phenology to Forecast Species Distributions across the Eastern United States in the 2070s

Studies that use species distribution models (SDMs) to document the relationship between species’ geographic range and environmental conditions rarely consider functional traits, such as phenology, that strongly affect species’ demography and fitness. Using more than 120,000 herbarium specimens representing 360 plant species across the eastern United States, we created a novel “phenology-informed” SDM that integrates dynamic phenological responses to changing climates. Compared to standard SDMs based only on abiotic variables, our phenology-informed SDMs forecast significantly lower species habitat loss, and less species turnover within communities under climate change. These results suggest that phenotypic plasticity and/or local adaptation in phenology may help many species adjust their ecological niches and persist in their habitats during periods of rapid environmental change. By modeling historical data that link phenology, climate and species distributions, our findings reveal how species’ reproductive phenology mediates their geographic distributions along environmental gradients and affect regional biodiversity patterns under future climate changes. More importantly, our newly developed model also circumvents the need for mechanistic models that explicitly link traits to occurrences for each species, and could thus facilitate the deployment of trait-based SDMs across unprecedented spatial and taxonomic scales.

现有利用物种分布模型（Species Distribution Models, SDMs）探究物种地理分布范围与环境条件间关联的研究，极少会考量诸如物候学这类对物种种群动态与适合度具有显著调控作用的功能性状。本研究依托采集自美国东部、涵盖360种植物的逾12万份植物标本馆标本，构建了一种全新的纳入物候信息的物种分布模型（phenology-informed SDM），该模型整合了物种针对气候变化的动态物候响应。相较于仅以非生物变量为基础构建的传统物种分布模型，本研究提出的纳入物候信息的物种分布模型，在气候变化情景下预测得到的物种栖息地丧失程度显著更低，群落内部的物种周转速率也更为平缓。上述结果表明，物候相关的表型可塑性与/或本地适应，或许能够帮助众多物种在环境快速变化的时期调整其生态位，并在原有栖息地中持续存续。本研究通过对关联物候、气候与物种分布的历史数据开展建模分析，揭示了物种繁殖物候如何介导其沿环境梯度的地理分布格局，并阐明了其在未来气候变化情景下对区域生物多样性模式的调控机制。更为关键的是，本研究新构建的模型无需借助需针对每个物种明确建立性状与出现记录关联的机理模型，因此能够推动基于功能性状的物种分布模型在前所未有的空间尺度与分类学尺度上推广应用。