Department of Biology, Willamette University, 900 State St, Salem, OR, 97301, USA.
Department of Mathematics, University of Utah, 155 S. 1400 E JWB 233, Salt Lake City, UT, 84112, USA.
Int J Biometeorol. 2018 Sep;62(9):1675-1684. doi: 10.1007/s00484-018-1567-6. Epub 2018 Jun 18.
In the Pacific Northwest of the USA, winter and spring temperature vary with the Pacific Decadal Oscillation, making effects of anthropogenic warming difficult to detect. We sought to detect community-level signals of anthropogenic change in a legacy plant phenology dataset. We analyzed both incomplete data from 1959 to 2016 on spring phenology of 115 species and more complete 1996-2016 data on spring and fall events for 607 plant species. We used ordination of the long-term dataset to identify two major axes of community-level change in phenology among years, with the first being a trend toward earlier spring phenology in more recent years. In contrast, for the short-term dataset, variation in spring phenology was mostly PDO-driven and did not reveal a strong trend over time. At both time scales, a second axis of phenological variation reflected summer and fall events, especially earlier appearance of fall color in recent years. In univariate analysis, more than 80% of individual species' leaf out dates and first flower dates occurred earlier over time, for an average advance across all species of 2.5 days per decade from 1959 to 2016. While most events did not advance in the period 1996-2016, fall color advanced by 10.6 days per decade, suggesting that intensification of summer drought has continued regardless of the PDO cycle. While estimates of slope over time depended strongly on the time window chosen for the analysis, estimates of slope versus temperature were consistently negative regardless of time window, averaging 5-7 days per 1 °C for spring events.
在美国太平洋西北地区,冬春两季的温度随太平洋十年涛动而变化,使得人为变暖的影响难以察觉。我们试图在一个植物物候学的历史数据集检测人为变化的群落水平信号。我们分析了 1959 年至 2016 年间不完全的 115 种植物春物候数据和 1996 年至 2016 年间较为完整的 607 种植物春、秋物候数据。我们利用长期数据集的排序来确定多年来物候学中群落水平变化的两个主要轴,第一个是近年来春物候较早的趋势。相比之下,对于短期数据集,春物候的变化主要受 PDO 驱动,并且没有随着时间的推移呈现出强烈的趋势。在两个时间尺度上,物候变化的第二个轴反映了夏季和秋季的事件,尤其是近年来秋季色彩较早出现。在单变量分析中,80%以上的物种的展叶日期和初花期随着时间的推移而提前,在 1959 年至 2016 年期间,所有物种的平均提前速度为每十年 2.5 天。虽然大多数事件在 1996 年至 2016 年期间没有提前,但秋季色彩每十年提前 10.6 天,这表明夏季干旱的加剧仍在继续,无论 PDO 周期如何。虽然随时间的斜率估计强烈依赖于分析中选择的时间窗口,但斜率与温度的关系估计无论时间窗口如何始终为负,春物候的平均斜率为每 1°C 5-7 天。