**1. Introduction**

The global climate has been warming gradually over the past several decades, which has important impacts on vegetation phenology in ecological systems [1–3]. Vegetation phenology acts as a sensitive and precise indicator that responds to climate warming and has become an important topic in the fields of climate and ecology [4,5]. Studies have shown that changes in spring and autumn plant phenology caused by climate change can differentially alter the length of the growing season and affect water, carbon, and energy fluxes between the atmosphere and the terrestrial biosphere [6]. Increased carbon uptake stimulated by an extended growing season has the potential to mitigate climate change [7]. Therefore, elucidating the trends in plant phenology can improve our understanding of the influence of climate change on ecosystem productivity, carbon cycling, and energy flow.

Although many studies have investigated plant phenology, little attention has been paid to alpine wetland ecosystems [8,9]. As the largest alpine wetland ecosystem in the world, the Three-River Headwaters region (TRHR) is considered the premonitory region of global climate change. It is worth noting that increasing human activities and global climate warming have led to severe ecological degradation in the TRHR, such as vegetation degradation, soil erosion, desertification, lake and wetland decline, and glacial retreat [10,11]. Because of the unique geographical location and climate of TRHR, a large number of researchers have studied this area. For example, Han et al. studied the relationship between plant greening and climate factors based on plant phenological site data, and the results showed that the trend for the time of plant greening was ahead–postpone–ahead–postpone [12]. Li explored the phenology response of plant to hydrothermal conditions from 1999 to 2010 based on SPOT NDVI, and the results indicated that the increase of cumulative precipitation and temperature of response time make SOS delayed [13]. Chen et al. used SPOT NDVI to explore the spatiotemporal patterns of plant phenology during 2000–2013, and the results showed that the SOS advanced, EOS delayed, and LOS extended [14]. Hence, it is a good idea to select the TRHR as a study area to explore the changes in plant phenology under climate change, which will improve our understanding of changes in plant phenology in alpine wetland ecosystems. Increased warming trends and frequent extreme events caused by climate change have produced significant impacts on many ecosystems, such as changes in vegetation phenology, grassland degradation, wetland shrinkage, and encroachment upon farmlands [15]. Currently, many research studies have focused on the response of vegetation phenology to specific climate factors, including temperature, precipitation, and shortwave radiation. The results indicated that interaction between temperature, shortwave radiation, and water has caused various impacts on vegetation activities in different regions [16–18]. For example, the SOS arrived 2.5 days earlier, and the EOS was delayed by 1 day for every 1 ◦C increase in the temperature across 19 European countries [19]. The onset time of 70.1% of vegetation in the growth season was delayed by 2.7 days because of winter precipitation in boreal forests [20]. Shortwave radiation plays a potential role in regulating vegetation growth in humid tropical or subtropical regions [21]. However, many factors can affect vegetation growth. Some changes in vegetation growth are caused by changes in climatic factors, but the soil factor (i.e., total soil C, N, and K) also affects vegetation dynamics because of the effects of soil conditions on the production of new cells that control plant photosynthesis [22]. For example, increasing the N input to land terrestrial ecosystems can promote vegetative growth and accelerate respiration in plants and soil microorganisms [23]. In fact, plants are very sensitive to resource conditions and tend to adjust their growth rates according to changing environments at different time scales [24]. A change in soil nutrient availability and mobility can change the photosynthetic rate of vegetation, which ultimately determines the difference in vegetative growth [25]. Therefore, it is important to understand the underlying mechanisms of how soil resources affect vegetative growth, especially under global climate change. Furthermore, traditional multivariate analysis ignored the total effects associated with the interaction between variables and only focused on the direct effects of predictors on the response variables [24]. Simultaneously, the interaction between

variables often has a greater impact on response variables. Hence, it is necessary to analyze the direct or indirect effects of a particular variable on another variable to study the factors that influence plant phenology.

In this study, we extracted plant phenological information based on MOD09A1 datasets with Google Earth Engine; the accuracy of the extracted plant phenology results was verified by using the station data of plant phenology; our main aims are: (1) investigating spatiotemporal characteristics of plant phenology (2) and analyzing the potential influence mechanism of climate and soil factors on the plant phenology of TRHR.

#### **2. Materials and Methods**
