植物生态学报 ›› 2018, Vol. 42 ›› Issue (5): 526-538.DOI: 10.17521/cjpe.2017.0305

• 研究论文 • 上一篇    下一篇

中国草地和欧洲木本植物返青期对气温和降水变化的响应: 基于生存分析的研究

周彤1,曹入尹2,王少鹏1,陈晋3,唐艳鸿1,*()   

  1. 1 北京大学城市与环境学院, 北京 100871
    2 电子科技大学资源与环境学院, 成都 611731
    3 北京师范大学地理科学学部, 地表过程与资源生态国家重点实验室, 北京 100875
  • 收稿日期:2017-11-21 修回日期:2018-02-11 出版日期:2018-05-20 发布日期:2018-07-20
  • 通讯作者: 唐艳鸿
  • 基金资助:
    国家自然科学基金(41601381)

Responses of green-up dates of grasslands in China and woody plants in Europe to air temperature and precipitation: Empirical evidences based on survival analysis

ZHOU Tong1,CAO Ru-Yin2,WANG Shao-Peng1,CHEN Jin3,TANG Yan-Hong1,*()   

  1. 1 College of Urban and Environmental Sciences, Peking University, Beijing 100871, China
    2 School of Resources and Environment, University of Electronic Science and Technology of China, Chengdu 611731, China
    3 State Key Laboratory of Earth Surface Processes and Resource Ecology, Institute of Remote Sensing Science and Engineering, Faculty of Geographical Science, Beijing Normal University, Beijing 100875, China
  • Received:2017-11-21 Revised:2018-02-11 Online:2018-05-20 Published:2018-07-20
  • Contact: Yan-Hong TANG
  • Supported by:
    Supported by the National Natural Science Foundation of China.(41601381)

摘要:

长期以来, 气候与植物物候关系的研究大多基于线性模型, 但植被物候对气候变化的响应可能是非线性的。该文利用非线性模型——生存分析模型来分析时间序列中过去事件(气候因子)对目的变量(物候)的作用: 用生存分析模型分析了春季气温和降水量对内蒙古草地、青藏高原草甸和欧洲地区木本植物返青期的影响。其中, 内蒙古与青藏高原的物候信息来自遥感数据, 欧洲地区物候信息为实测数据。蒙特卡洛方法用于拟合模型参数。结果表明: 生存分析模型适合对上述不同研究对象的物候影响因素进行分析, 并能模拟非线性效应; 在内蒙古草地, 春季气温和降水对春季返青有很大的影响, 而青藏高原草甸和欧洲地区木本植物返青期对春季降水响应较小, 对春季气温变化的响应随Holdridge干燥度指数上升而下降; 在预测返青期时发现: 春季平均气温提高1 ℃会导致上述地区返青期提前1-6天; 而春季气温与降水的增加会导致返青期发生明显的非线性变化, 这种非线性效应无法基于线性模型模拟出来。结果说明生存分析模型既能用于分析不同尺度下植物物候与气候的关系, 也能用于模型预测, 尤其适合探讨大幅度气候变化对物候的非线性影响。

关键词: 气候变化, 草地, 物候, 生存分析法, 青藏高原

Abstract:

Aims Linear models have been widely used to examine the impacts of climatic factors on plant phenology, although the relationship between phenology and climate could be nonlinear. Based on survival analysis, robust nonlinear models were empirically developed to examine the phenological changes in relation to air temperature and precipitation for the grasslands in China and individual woody plants in Europe.

Methods Three datasets were used in our survival analysis: two datasets of the remotely-sensed vegetation phenology for grasslands in Nei Mongol grasslands and meadows in Qinghai-Xizang Plateau, and a dataset of the phenological observations of individual woody plants in Europe. Monte Carlo simulations were performed to estimate model parameters in our survival analysis.

Important findings The survival analysis appeared to be a powerful tool in modeling the nonlinear changes in green-up date (GUD) to the climatic factors. The analyses showed that both spring temperature and precipitation are significantly correlated with the GUD in the semi-arid grasslands in Nei Mongol. For Qinghai-Xizang Plateau and Europe, the spring temperature seemed highly correlated with GUD, while the correlation was weak with the higher Holdridge aridity index. The survival model predicted that the GUD in the three regions would be advanced by 1-6 days with an increase in temperature of 1 °C. A combined increase in spring temperature and precipitation would lead to nonlinear responses, suggesting the need for developing nonlinear models. Our empirical exercise in this study demonstrated that the survival analysis could offer an alternative tool for predicting plant phenology under the changing climate.

Key words: climate change, grassland, phenology, survival analysis, Qinghai-Xizang Plateau