Daily minimum temperature and precipitation control on spring phenology in arid-mountain ecosystems in China

摘要

Climate anomaly has caused substantial shifts in phenology of mountain ecosystems, but the underlying mechanism of phenological responses to climate change is still not well understood. In essence, the abundance of vegetation communities increases the complexity of phenology-climate relationships, leading to a certain limitation in predictions of future dynamics among different vegetation types using a unified model. In this study, we focused on the climatic constraints on spring phenology in arid mountains (AMs) of China, and emphasis was laid on accurate representation of mechanisms that control phenology across different vegetation types. We Firstly explored spatio-temporal variations in satellite-derived estimates of starting date of vegetation growing season (SOS) over the period 2000-2015 using moderate-resolution imaging spectro-radiometer (MODIS) normalized difference vegetation index (NDVI). Phenological models in response to climate variability were then established by using mixed-effect models based on satellite observations and an extensive dataset of climatic measurements. Our results showed that the climatic regulation on SOS varied greatly over vegetation types. More climatic factors that regulate phenological development were found in grassland than forest and shrubland. At ecosystem level, two critical climate factors, daily minimum temperature (T-min) and precipitation, explained 74-95% of total variability in predicted SOS. The observed sensitivity to T-min is expected to be closely linked with the risk of frost damage, while preseason precipitation determines water availability in spring. The varying ecosystem sensitivity revealed the different resilience and adaptability to changing climate among vegetation types, which have been linked to their eco-physiological characteristics (e.g., water use efficiency) and environmental conditions (e.g., elevation). Overall, our results indicate a strong dependence of spring phenology on T-min and precipitation, and create an opportunity for a more realistic representation of vegetation phenology and growth of AM plants in China in land surface models.