Climate change is known to affect regional weather patterns and phenology; however, we lack understanding of how climate drives phenological change across local spatial gradients. This spatial variation is critical for determining whether subpopulations and metacommunities are changing in unison or diverging in phenology. Divergent responses could reduce synchrony both within species (disrupting gene flow among subpopulations) and among species (disrupting interspecific interactions in communities). We also lack understanding of phenological change in environments where life history events are frequently aseasonal, such as the tropical, arid, and semi-arid ecosystems that cover vast areas. Using a 33-year-long dataset spanning a 1,267-m semi-arid elevational gradient in the southwestern United States, we test whether flowering phenology diverged among subpopulations within species and among five communities comprising 590 species. Applying circular statistics to test for changes in year-round flowering, we show flowering has become earlier for all communities except at the highest elevations. However, flowering times shifted at different rates across elevations likely because of elevation-specific changes in temperature and precipitation, indicating diverging phenologies of neighboring communities. Subpopulations of individual species also diverged at mid-elevation but converged in phenology at high elevation. These changes in flowering phenology among communities and subpopulations are undetectable when data are pooled across the gradient. Furthermore, we show that nonlinear changes in flowering times over the 33-year record are obscured by traditional calculations of long-term trends. These findings reveal greater spatiotemporal complexity in phenological responses than previously recognized and indicate climate is driving phenological reshuffling across local spatial gradients.

中文翻译：

---

气候变化导致海拔升高的开花物候学发生发散性和非线性变化。

众所周知，气候变化会影响区域的天气模式和物候。然而，我们缺乏对气候如何在局部空间梯度上推动物候变化的理解。这种空间变化对于确定亚种群和元社区是一致变化还是物候差异是至关重要的。不同的响应可能会降低物种内部（破坏亚种群之间的基因流）和物种之间（破坏社区中的种间相互作用）的同步性。我们还缺乏对生活历史事件经常发生季节变化的环境中物候变化的理解，例如覆盖广阔区域的热带，干旱和半干旱生态系统。在美国西南部使用长达33年的数据集，该数据集跨越1267米的半干旱海拔梯度，我们测试了开花物候在物种内的亚种群之间以及包括590种物种的五个群落之间是否存在差异。应用循环统计数据测试全年开花的变化，我们发现除海拔最高的地区以外，所有社区的开花时间都早了。但是，开花时间跨海拔的变化速率可能不同，这可能是由于海拔和温度的特定变化所致，表明邻近社区的物候差异很大。各个物种的亚群在中海拔时也出现分歧，但在高海拔时在物候学上趋于一致。当跨梯度收集数据时，无法检测到群落和亚群之间开花物候的这些变化。此外，我们显示，在33年的记录中，开花时间的非线性变化被长期趋势的传统计算所掩盖。这些发现表明，物候响应的时空复杂性比以前认识到的要高，这表明气候正在推动局部空间梯度上的物候改组。