Alpine treeline is a sensitive indicator of the responses of terrestrial ecosystems to climate change (Lenoir et al., 2008; Hagedorn et al., 2019) and provides early warnings of global climate change (Alatalo and Ferrarini, 2017; Du et al., 2018). The treeline ecotone is expected to expand to higher latitudes and elevations because of rapid climate warming (Scuderi et al., 1993; Chen et al., 2011; Malanson, 2013). A global study showed that the treelines at only 52% of all 166 global treeline sites have advanced upslope over the past 100 years (Harsch et al., 2009). Recently, Lu et al. (2020) conducted a meta-analysis of the annual shift rate of treelines in the Northern Hemisphere and revealed that the treelines have shifted upward at 127 out of 143 sites (88.8%) since 1901. This shows that the trend of upward treeline movement is becoming increasingly obvious at most sites around the world. However, it has been suggested that the sensitivity of treeline ecotones depends more on changes in tree density than on the position of the treeline (Camarero and Gutierrez, 2004). Sigdel et al. (2020) quantified the changes in density and the spatial patterns of 17 treeline sites in the central Himalayas. The results suggested that younger trees showed clustering near the treeline. The method of historical reconstructions of tree positions has provided numerous examples of forest infilling in some alpine regions (Payette and Filion, 1985; Lyu et al., 2016; Wang et al., 2016). However, regardless of whether the treeline position shifts upward or the tree density increases (La Sorte and Jetz, 2010; Hagedorn et al., 2019; Rees et al., 2020), the area of tundra and meadow ecosystems at higher elevations will be reduced, resulting in the loss of biodiversity and ecosystem services (Thuiller et al., 2005; Feurdean et al., 2016; Pecl et al., 2017). Therefore, it is necessary to pay attention to these two aspects simultaneously while discussing the treeline pattern and its dynamic change.

Treeline exhibits a scale effect (Körner, 2012; Singh et al., 2015), which also determines the diversity of treeline research scales (Holtmeier and Broll, 2019). At the local scale, researchers usually focus on ecological processes, tree density changes, species distribution, and composition of the treeline transition zone (Liang et al., 2016; Sigdel et al., 2020). An increasing number of studies have analyzed position and spatial pattern changes in treeline on relatively large scales (global, regional, or landscape) (Guo et al., 2014; Mohapatra et al., 2019; Wei et al., 2020). However, there are differences in treeline patterns and their dynamics from local to global scales (Holtmeier and Broll, 2012; Marc and Johnson, 2013; Bader et al., 2020). For example, Gaire et al. (2014) found that the treeline in the central Nepal Himalaya is moving upward at a rate of 2.61 m/year at the sample plot scale. Mohapatra et al. (2019) found that the Himalayan treeline is moving upward at a rate of 11.3 m/year at the landscape scale. However, a meta-analysis of annual treeline shift rates in the Northern Hemisphere showed that the mean shift rate was only 0.354 m/year (Lu et al., 2020). This shows that many research results are controversial due to the influence of different scales.

How to choose an appropriate treeline scale to evaluate and predict the relationship between treeline change and climate change more comprehensively and accurately is a key problem. There are great differences among treelines in different climatic zones and biogeographic regions, but the influence of topographic factors on the spatial pattern of treeline is often ignored (Yirdaw et al., 2015; Bader et al., 2020). However, the treeline pattern on a fine scale is usually described as the relative position between trees (Sigdel et al., 2020), but this pattern cannot be applied to other areas. In addition, different scales will also affect the response of treeline to climate. On a fine scale, physical and biological factors may interfere with the response of tree growth to climate (Hellmann et al., 2016), and the position of treeline might be inconsistent with the climate conditions (Marc and Johnson, 2013). On the global scale, local differences will be ignored, and it may be easy to overemphasize coarse drivers such as temperature (Harsch et al., 2009; Moyes et al., 2015). The landscape scale may better reflect the relationship between treeline dynamics and climate (Batllori and Gutiérrez, 2008; Case and Duncan, 2014; Singh et al., 2015). However, how to obtain large-scale and high-precision treeline data is also a problem to be solved in the current research.

The Hengduan Mountains provide an extensive longitudinal, latitudinal, and topographical framework for studying the landscape pattern and dynamics of treeline. The objective of this study was to characterize the responses of treeline dynamics (treeline position and vegetation cover) to changes in climate at the landscape scale. Specifically, high-resolution Google Earth images were used to extract the alpine treeline in the Hengduan Mountains area in different periods, and then the changes in treeline position and vegetation coverage, and also their relationships with climate, were assessed. The results of this study provide new insights into the factors driving treeline dynamics at the landscape scale, and these insights could aid in the evaluation of the risks faced by alpine ecosystems and species in the alpine regions under future climate change.

The treeline elevation in the Hengduan Mountains area is high in the middle but lower in the surrounding areas. This may be caused by the mass elevation effect and the unique monsoon climate. Compared with the edges of the mountain at similar elevations, the interior part of the mountain range experiences higher solar radiation input and higher temperatures (Holtmeier, 2009; Han et al., 2020). This temperature difference is consistent with the difference in the elevation of the treeline between the inner and outer parts of the mountain range (Wang et al., 2017). The study area is also affected by the southwestern monsoon and the southeastern monsoon; precipitation decreases from southwest to northeast and from southeast to northwest (Li et al., 2010; Yao et al., 2012). The elevation of the treeline also increases along the direction of seasonal winds (Wang et al., 2013; Yao and Zhang, 2015).

In general, the treeline can reach greater heights on warm southern slopes compared with cold northern slopes (Malyshev and Nimis, 1997). Our study found that the elevation of the treeline is generally lower on the southern slope of the Hengduan Mountains than on the northern slope, but there is no difference in the mean temperature during the growing season. This shows that the growing season temperature at the treeline is consistent, while the isotherm distribution of the treeline on the southern slope is lower than that on the slopes with other aspects. To further support our conclusion, we calculated statistics for the mean growing season temperature of different aspects at different elevations and found that the mean growing season temperature of the southern slope at a given elevation is lower than that of the northern slope (Figure 9). These findings are consistent with a study in the Mediterranean region showing that treeline elevation on southern slopes may be limited by summer drought, which reduces the elevation of the treeline (Piper et al., 2016; Bonanomi et al., 2018). Different from the previous conclusions, the low elevation of the treeline on the southern slope of the Hengduan Mountains is not caused by drought but by the low temperatures induced by the special regional monsoon climate. This stems from the formation of a large number of clouds on the southern and western slopes of the Hengduan Mountains after noon, and the blocking of solar radiation and the low temperatures caused by rainfall decrease the elevation of the treeline (Poveda et al., 2005; Körner, 2012; Wang et al., 2013). In addition, the south-facing slopes are steeper and barren, making it very difficult for seedlings to colonize and grow.

Treeline location has a time lag in terms of reflecting climate (Holtmeier and Broll, 2019). In this study, we calculated the multiyear mean seasonal temperature to analyze the relationships between treeline elevation and various climate factors. Since there are obvious differences in the treeline elevation among each aspect, we use CCM to test the causal relationships between the treeline elevation of each aspect and climate variables and topographic variables. Compared with topography and precipitation, temperature has a stronger driving effect on treeline elevation. Although the range of the temperature threshold for tree growth is not clearly defined, it is widely accepted that the treeline position is closely related to temperature (Case and Duncan, 2014). The upper limit of the treeline is thought to be determined by the effect of temperature on the ability of trees to form new tissue rather than by a lack of photosynthesis (Körner, 1998). Both biogeographic and physiological evidence strongly suggests that low temperatures are the most important factor limiting the upslope advance of the treeline (Cieraad, 2011; Case and Duncan, 2014). Temperature is a critical factor limiting the production and differentiation of xylem cells in cold climates, and thermal conditions below a certain temperature can inhibit xylem growth (Rossi et al., 2008). Immature trees are unable to survive in alpine areas with low air temperatures (Li and Yang, 2004). Thus, under undisturbed natural conditions, as long as there is soil and sufficient moisture, trees can grow until a low temperature limit is reached, forming a natural climatic treeline (Körner, 2020).

Due to the impact of global warming, the growth of trees in most treeline sites has responded positively to the rise in temperature (Harsch et al., 2009; Camarero et al., 2021). A temperature increase gradually alleviates physiological limitations associated with low temperatures in alpine trees (Körner and Paulsen, 2004; Pandey et al., 2020). Most global treeline advances correspond to strong winter warming (Harsch et al., 2009). Due to the obvious differences in the degree of seasonal warming experienced by different regions, the treeline position is greatly affected by the local climatic and topographic conditions (Armbruster et al., 2007; Bonanomi et al., 2020). Since 1960, the temperature in the Hengduan Mountains has increased at a rate of 0.03°C/year, and in recent decades, the temperature has increased by approximately 1.8°C. According to the IPCC (2018) assessment report, the global temperature has increased by nearly 1°C during the past century (1906–2017). Although the speed of the temperature increase in the Hengduan Mountains area is much higher than the global average, the position of the treeline has not obviously moved upward. Based on a previous treeline study in the central Hengduan Mountains, the larch and fir trees required averages of 18 ± 2 (n = 150) and 33 ± 5 (n = 30) years to reach 2 m in height, respectively (Wang et al., 2019). This means that a seedling may require 27–50 years to reach 3 m in height (treeline trees), so 20 years is still too short to recognize a treeline advance.

Although there was no obvious upward trend in treeline elevation in the Hengduan Mountains over the past 20 years, the trend of increasing vegetation coverage was significant. In this study, 80.3% of the regional FVC showed an upward trend, of which 30.17% showed a significant upward trend. The FVC was strongly affected by precipitation in the growing season. A warm-growing season and adequate moisture are favorable for seedling emergence and growth (Körner, 2012; Han et al., 2018). In some drought-prone areas, tree recruitment also showed a gradual decrease in the treeline transition zone (Camarero and Gutierrez, 2004; Liang et al., 2014). However, our study shows that the increase in the FVC is related to a decrease in precipitation. Because precipitation decreases with increasing elevation, considering the impact of water availability (Liang et al., 2014), trees may not become established at higher elevations that are seasonally dry. In this context, trees tend to gather (Moyes et al., 2015; Treml and Chuman, 2018). Sigdel et al. (2020) showed a higher clustering tendency in dry treelines than in wet treelines. In high-elevation areas with high abiotic stress, plants can have a positive impact on each other, reduce the negative impact of adverse environmental conditions and improve the species richness of adjacent areas (Callaway et al., 2002; Smith et al., 2003; Chen et al., 2020). The FVC in 13.37% of the treeline areas showed a decreasing trend, and only 0.69% showed a significant decrease. Although this study carefully removed disturbed treeline areas in the process of treeline extraction in the early stage, the damage to vegetation in the alpine areas caused by natural interference factors (avalanches, landslides, gullies, etc.) and human interference factors (road construction, mining, grazing, fire, etc.) cannot be ignored (Ameztegui et al., 2016; Holtmeier and Broll, 2019; Wang et al., 2019).

The rate of the temperature increase in the Hengduan Mountains is much higher than the global average. Under this background, the treeline position has not changed significantly, but the FVC in the treeline area has markedly increased. The results also revealed a special exposure effect of alpine treeline in the Hengduan Mountains. Because of the lower treeline isotherm position caused by the monsoon climate, the treeline position on south-facing slopes is lower than that on the slopes with other exposures. The aforementioned results are helpful for understanding the dynamic changes in treeline under the background of climate warming at the landscape scale and provide a reference for predicting and evaluating the threats faced by alpine ecosystems in the future.

The original contributions presented in the study are included in the article/supplementary material, further inquiries can be directed to the corresponding author.

WW designed the study. FZ and CY collected and processed data. CT and FZ conducted the statistical analyses. FZ wrote the manuscript. ZZ and DL revised the manuscript. All authors commented upon and contributed to the final version of the manuscript.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.