Effects of altitude and slope on the climate–radial growth relationships of Larix olgensis A. Henry in the southern Lesser Khingan Mountains, Northeast China

Abstract

Background

The relationship between climate and radial growth of trees exhibits spatial variation due to environmental changes. Therefore, elucidation of how the growth–climate responses of trees vary in space is essential for understanding forest growth dynamics to facilitate scientific management with the ongoing global climate warming. To explore the altitudinal and slope variations of these interactions, tree-ring width chronologies of Larix olgensis A. Henry were analyzed in the southern Lesser Khingan Mountains, Northeast China.

Results

The radial growth of L. olgensis exhibited significant 5- to 10-year periodic changes at three altitudes and two slopes, and the frequency change occurred mainly during the early growth stage and after 2000. The radial growth of L. olgensis was significantly negatively correlated with September precipitation only at low altitudes, but also with the mean temperature in July–August and the mean maximum temperature in June–August at high altitudes. The radial growth of L. olgensis at low and middle altitudes as well as on the sunny slope led to a higher demand for moisture, while temperature was the key limiting factor at high altitudes and on the shady slope.

Conclusions

The climate–radial growth relationship of L. olgensis exhibits altitudinal and slope variability. This study quantitatively describes the spatially varying growth–climate responses of L. olgensis in the southern Lesser Khingan Mountains, which provides basic data for the management of L. olgensis forests and the prediction of future climate impacts on forest ecosystems.

Introduction

Global warming has observably affected the environment (IPCC 2014; NASA 2021) and has changed the structure and function of forest ecosystems (Lenoir et al. 2008; Lindner et al. 2010), especially in high latitudes and altitudes in the Northern Hemisphere (Serreze et al. 2000; Andreu et al. 2007; Zhang et al. 2016; Wang et al. 2017). Whether there are stable climate–radial growth relationships of specific tree species has become a scientific issue of wide concern (Babst et al. 2018). Tree ring represents the footprint of tree growth (Ogden 1981; Babst et al. 2018; Silva et al. 2019) and is associated with advantages, such as high annual resolution, accurate age-dating, and wide sample distribution (Douglass 1941; Fritts 1976; He et al. 2019; Silva et al. 2019). Tree rings are one of the essential means to study climate–growth relationships and calculate and predict the variation of forest growth, forest biomass, forest stock volume, and forest carbon storage at various time scales (Babst et al. 2014, 2018; He et al. 2019; Yu and Liu 2020).

Numerous studies investigating climate–radial growth relationships involve dominant tree species in alpine mountains, arid areas, and other monsoon climate regions (Yu et al. 2006; Shen et al. 2016; Babst et al. 2018; Panthi et al. 2018; He et al. 2019; Jiao et al. 2019). The findings consistently show that the radial growth of trees is significantly affected not only by weather-related environmental factors, but also by regional physiognomic characteristics, especially altitude and slope (Yu et al. 2013; Zhang et al. 2017; Zhu et al. 2018; Yu and Liu 2020). Specifically, growth is mainly affected by precipitation at low altitudes and temperature at high altitudes in Laobai Mountains, Lesser Khingan Mountains, Changbai Mountains, and Hengduan Mountains (Zhu et al. 2018; Sun et al. 2020; Yu and Liu 2020). Temperature was the key limiting factor for tree distribution on different slopes of Changbai Mountains (Chi et al. 1982; Yu et al. 2011, 2013; Yu and Liu 2020). However, these key limiting climatic factors of tree growth are not necessarily applicable for other tree species at various altitudes and slopes in other regions (Yu et al. 2011; Babst et al. 2018; He et al. 2019; Li et al. 2020a, b, c; Zhang et al. 2020a, b, c). Therefore, extensive studies are needed to investigate the spatially varying growth–climate responses of widely distributed forests.

Rapid climate warming has been reported in the middle to high latitudes and high-elevation mountainous regions (Muhlfeld et al. 2011). The Lesser Khingan Mountains are a typical region of climate warming due to its high latitude, where the climate–radial growth relationship of trees exhibits obvious regional characteristics (Shen et al. 2015; Lei et al. 2016). Numerous studies focused on the mixed broad-leaved Korean pine (Pinus koraiensis Siebold & Zucc.) forest have shown that the tree growth in this region is affected by temperature and precipitation, especially high temperatures (Yin et al. 2009; Lei et al. 2016; Yu et al. 2017; Li et al. 2020a, b, c). Larix olgensis A. Henry is the main afforestation tree species in the region, with a considerable distribution area (Wang et al. 2011; Yu et al. 2017; Li et al. 2020a, b, c). With its substantial forest stock volume and carbon sequestration capacity, this tree species plays a pivotal role in regional and even global carbon and nitrogen cycles and sustainable forest management (Wang et al. 2011; Lei et al. 2016). However, comparatively fewer studies explored its climate–radial growth relationships, with inconsistent results (Andreu et al. 2007; Yin et al. 2009; Lin et al. 2013; Yu et al. 2017; Yu and Liu 2020). Climate warming promoted the radial growth of L. olgensis in the Changbai Mountains (Lin et al. 2013). Its radial growth is significantly limited by climate change in Iberian, which showed an upward abrupt at the end of the first half of the twentieth century and a downward shift during the mid-twentieth century (Andreu et al. 2007). Climate warming exerted an inhibitory effect on its radial growth in other areas of Northeast China (Yu and Liu 2020). Therefore, how the radial growth of L. olgensis responds to climate change in the Lesser Khingan Mountains, Northeast China requires further exploration.

We hypothesized that the climate–radial growth relationship of L. olgensis is strongly regulated by the local climate in our study area. To investigate this hypothesis, we developed a widely distributed tree-ring width chronology to analyze the diverse growth–climate relationships over different altitudes and slopes. Our aims were (a) to elucidate the periodicity patterns of the radial growth of L. olgensis at different altitudes and slopes, (b) to detect the growth–climate relationships, and (c) to identify and quantify the climatic factors driving its radial growth. The findings provide first-hand data underlying the dynamic growth of L. olgensis and the prediction of forest growth, biomass, stock volume, and carbon storage, which can guide sustainable regional forest management.

Materials and methods

Study area

The study area is located in Dongzhelenghe Nature Reserve in the southern Lesser Khingan Mountains, Northeast China (46° 29ʹ to 47° 06ʹ N, 128° 30ʹ to 129° 24ʹ E, 200–970 m, Fig. 1a and b). It has a temperate continental monsoon climate with warm rainy summers and long cold winters. The mean annual temperature recorded for the 1966–2018 period at the Yilan Meteorological Station (46° 18ʹ N, 129° 35ʹ E, 100.1 m, Fig. 1b) is 3.7 °C, with January being the coldest month (− 22.4 ℃) and July being the warmest (27.5 ℃). The mean annual precipitation is 546.7 mm, approximately 82% of which is deposited in the warm season (May–September). Frost occurs often, and the frost-free period lasts only 90 to 110 days (end of May–beginning of September). Its main forest community type is natural Korean pine and broad-leaved mixed forest. L. olgensis and P. koraiensis are distributed widely as the dominant coniferous tree species.

Fig. 1

Location and distribution of the sampling sites in the southern Lesser Khingan Mountains, Northeast China. a Location of the study area; b Yilan Meteorological Station; c Distribution of the sampling sites. Sampling sites are abbreviated as LA (Low altitude), MA (Middle altitude), HA (High altitude), SUN (Sunny slope), and SHA (Shady slope). Detailed information is available in Table 1

Sampling and chronology establishment

The sample cores of L. olgensis were collected at five sampling sites in July 2019 (Fig. 1c and Table 1), spanning an elevation of 286–600 m. Sampling sites of different altitudes were set on a sunny slope, and those of different slopes were set at an altitude of approximately 300 m (Table 1). All sample trees were visually assessed before sampling to ensure that only healthy trees in the upper part of the canopy were sampled. To minimize damage to the trees, only one core was extracted from each tree at diameter at breast height (DBH, 1.3 m above ground). In total, 390 cores were extracted.

Table 1 Information of sampling sites

The collected core samples were taken back to the laboratory for subsequent processing. The mounted cores were first polished with increasingly finer sandpaper from 100 to 1000 girt until the tree-ring boundaries were clearly visible. They were then visually cross-dated with the skeleton plot method under a Leica-S4E stereo microscope with LINTAB™ 6.0. Forty-five cores with abnormal ring features were excluded from the analysis, such as those with missing rings or indistinct boundaries that made cross-dating difficult (Douglass 1941; Silva et al. 2019). The tree-ring width was measured using a LINTAB™ 6.0 measuring system with an accuracy of 0.01 mm. The quality of measurement and cross-dating was checked using the COFECHA program (Holmes 1983). To remove the effect of non-climatic factors, tree-ring width series were detrended and standardized using the ARSTAN program (Cook and Holmes 1986) with a 67% cubic smoothing spline function at a 50% cutoff frequency. The standard chronologies (STD), the residual chronologies (RES), and the autoregressive chronologies (ARS) in each site were established. Having taken all chronological statistical parameters (Fritts 1976) into account, we focused on STD (Fig. 2) and RES (Fig. 3) for the following analysis at different altitudes and slopes, respectively.

Fig. 2

Tree-ring width standard chronologies (solid black line) with their sample depths (grey-shaded area) along the altitudinal gradient

Fig. 3

Tree-ring width residual chronologies (solid black line) and the sample depths (grey-shaded area) at different slopes

Climate data

As the local meteorological stations were far from our sampling sites, the monthly temperature and precipitation of the Climate Research Unit (CRU) TS 4.04 (land) gridded dataset from 1966 to 2018 were utilized for this study. The Palmer Drought Severity Index (PDSI) was used to represent the soil water balance and reflect drought conditions in this study (Mika et al. 2005), which was downloaded from the self-calibrating PDSI Global dataset of the CRU spanning 1966–2017. The above datasets were collected from the Royal Netherlands Meteorological Institute (KNMI) climate explorer (http://climexp.knmi.nl) with a spatial resolution of 0.5° × 0.5°. Considering the growth characteristics of L. olgensis and the possible lagged effects of weather conditions on it, the climatic factors from May of the past year to September of the current year were selected for subsequent analysis (Fig. 4).

Fig. 4

Interannual (left) and monthly (right) variations of temperatures (mean, maximum, minimum) and precipitation in Dongzhelenghe Nature Reserve from 1966 to 2018

Statistical analysis

The Mann–Kendall method (Kendall and Gibbons 1992) and wavelet analysis (Addison 2002) were used with MATLAB to analyze the trends and phase mutation of radial growth and its periodic change patterns on multiple time scales, respectively. Radial growth–climate relationships among different altitudes and slopes were determined by response and correlation function analysis with the DendroClim2002 program (Biondi and Waikul 2004). Then, they were further tested by redundancy analysis (RDA) with CANOCO 5.0 software (Ter Braak and Smilauer 2012). The impact of each climatic factor on the radial growth of L. olgensis at different altitudes and slopes was quantitatively described by simplified regression equations with R software. Figures were drawn with Origin 16.0.

Results and discussion

Interannual characteristics of climate change

The mean, mean minimum, and mean maximum annual temperature has been rising since 1966 (Fig. 4a–c). The mean minimum temperature has risen almost three times as much as the mean maximum temperature. The precipitation and air relative humidity fluctuated in a small range, with the overall trend decreasing first and increasing slightly subsequently (Fig. 5d and e). The PDSI ranged from − 3.36 to 2.67, much of which was less than − 2, and its rate of decrease was 0.024 year⁻¹ (Fig. 5f). The region showed obvious climate warming and drying, consistent with the overall climate change in Northeast China (Sun et al. 2005; Ye et al. 2019a, b).

Fig. 5

Climatic interannual changing trend in the study area during the period of 1966–2018. a Mean temperature, b mean minimum temperature, c mean maximum temperature, d precipitation, e PDSI, and f air relative humidity. The dashed lines are indicative of the significant linear regression trends with the equations, p values, and R² values highlighted

With this trend enhanced and its affected areas expanded, more extensive and severe droughts will occur in the land area in the next 30–90 years (Huang et al. 2016; Ye et al. 2019a, b). The adverse effects of drought, such as tree mortality and forest degradation, have been confirmed by some studies (Tang et al. 2015; Jiao et al. 2019). Barber et al. (2000) found reduced growth of Alaskan white spruce from temperature-induced drought stress in most areas of the northern United States. The widespread decline and mortality of trees have also been confirmed in the shelterbelt forests of northern China in recent decades and may become more severe (Li et al. 2020a, b, c). At the same time, the spruce–fir–Korean pine forest would replace Pinus sylvestris var. sylvestriformis in the community ecotone of the Changbai Mountains under this continuous trend (Yu et al. 2006). Similarly, the current regional climate warming and drying have affected the growth of L. olgensis, which can hamper the maintenance of its existing dominant ecological niche.

Statistical characteristics of tree-ring width chronologies

There were no significant differences in growth rate among the five chronologies, with a mean growth rate of 0.9894 mm year⁻¹ (Table 2). The standard deviations and correlations between trees ranged from 0.12 to 0.22 and 0.30 to 0.71, respectively, which showed that the established chronologies had high regional consistency and could reflect the growth status of L. olgensis at different altitudes and slopes. The mean sensitivity varied from 0.15 to 0.23, and the signal-to-noise rate varied from 4.69 to 12.14, indicating that all chronologies retained abundant climatic information. The variation in first eigenvector ranged from 33.5 to 78.1%. The expressed population signal ranged from 0.966 to 0.995, which all exceeded a threshold of 0.85. The chronology statistics are similar to those reported in previous studies for this (Lin et al. 2013; Shen et al. 2016; Yu and Liu, 2020) and other species (Yin et al. 2009; Yu et al. 2017; Li et al. 2020a, b, c). The established chronologies are suitable for climate–growth analyses.

Table 2 Statistics of tree-ring width chronologies and common interval analysis

Periodicity of the tree-ring width index

The overall change trends of the tree-ring width index of L. olgensis were similar at different altitudes and slopes, with large fluctuations from 2007 to 2012 (Figs. 2 and 3). The minimum value of the tree-ring width index was detected at high altitudes and on shady slope in 2007 and at other altitudes and sunny slope in 2012, while the maximum value was found in 2009. The severe autumn drought of 2007, the low temperature and heavy summer rainfall of 2012, and the Dendrolimus superans infestation in some areas during this time adversely affected the growth of L. olgensis (Li et al. 2016; Liang et al. 2018), which was reflected in periodic change in its radial growth.

A statistically significant decrease or increase in tree-ring width index was observed in the early growth stage and after 2000, despite the absence of precise mutation year (Fig. 6). The rapid warming in Northeast China after 2000 may have contributed to these results (Zhou et al. 2020). In addition, during the early growth stage, the roots of trees grow relatively slowly and exhibit poor water use efficiency, making trees more sensitive to hydrothermal conditions, resulting in significant fluctuation in radial growth (Schenk and Jackson 2002; Rozas et al. 2009; Brunner et al. 2015). Relevant studies also proved that trees in the early growth stage are more sensitive to climatic factors, such as Quercus rubra L. in the northern USA, Smith fir (Abies forrestii var. smithii R. Vig. & Gaussen) in the northeast Tibetan Plateau, black spruce (Picea mariana [Mill.] Britton, Sterns & Poggenb.) in the semi-humid climate region of Manitoba, and Juniperus thurifera L. in the semi-humid climate region of north-central Spain (McMillan et al. 2008; Rozas et al. 2009; Haavik et al. 2011; Li et al. 2013).

Fig. 6

Mann–Kendall mutation test curves of tree-ring width index with forward statistic UF (solid line) and backward statistic UB (dashed line) at the 0.05 significance level (dotted horizontal lines). The intersection of the UF and UB curves is located between the critical lines, corresponding to the time the mutation begins

The tree-ring width index of L. olgensis showed significant 5- to 10-year periodic variations at various altitudes and slopes during different periods. It was 2003–2018, 2004–2017, and 1996–2018 with the altitude increased and 2002–2018 and 2009–2018 on the sunny slope and shady slope, respectively (Fig. 7). These may be related to large-scale climate change and non-climatic pressures, such as global climatic oscillation and land–sea thermal differences (Piraino and Roig 2013; Venegas-González et al. 2015; Zhu et al. 2017; Yu et al. 2021). Similar results were reported in the Changbai Mountains, Qinling Mountains, and Tianshan Mountains, indicating a universal impact of large-scale climate on tree growth (Yu et al. 2018, 2021; Jiang et al. 2019; Jiao et al. 2019).

Fig. 7

Real part wavelet coefficient contour map of tree-ring width index at a low altitude, b middle altitude, c high altitude, d sunny slope, and e shady slope. Y-axis is the scale in year, and X-axis is the time in year

Growth–climate relationships

Altitudinal variability of the growth–climate relationships

The radial growth of L. olgensis at various altitudes was mainly affected by the climatic factors of the current year (Fig. 8). The impact of temperature increased, while precipitation decreased with increasing altitude (Figs. 8, 9a). Similar studies conducted in Northeast China (including Changbai Mountains), Qilian Mountains, and Hengduan Mountains reported the same conclusions (Zhang et al. 2017; Zhu et al. 2018; Sun et al. 2020; Yu and Liu, 2020).

Fig. 8

Response of radial growth of L. olgensis to monthly climatic factors at different altitudes (the upper half) and slopes (the lower half). The capitalized P means months from the past year and C from the current year. The gradual change of color from blue to red indicates a gradual change of correlation from negative to positive. *p < 0.05; **p < 0.01

Fig. 9

Redundancy analysis (RDA) for the chronologies and the monthly climatic factors at different altitudes (a) and slopes (b). Significant (p < 0.05) climatic factors are indicated by solid line arrows and named as “the climatic factor-corresponding month” (e.g., the mean temperature in July of the current year named as Tmean-C7). Arrow (vector) length and the cosine of the angle between two vectors depict the magnitude of variables and their correlation. The longer the vector, the more important the climatic factor. Vectors crossing at sharp angles, obtuse angles, and right angles, respectively, indicate a positive correlation, a negative correlation, and a near-zero correlation

Temperatures and hydrothermal conditions in summer play a critical role in the radial growth of L. olgensis at high altitudes (Fig. 8). Specifically, the radial growth was significantly negatively correlated with the current July–August T_mean (r = − 0.359, r = − 0.313) and the current June–August T_max (r = − 0.317, r = − 0.446, r = − 0.310) and significantly positively correlated with the current July PDSI (r = 0.361) and the current June–July RH (r = 0.434, r = 0.516). Summer is the peak growing season of L. olgensis (Wang et al. 2011). Nevertheless, excessively high temperatures decrease moisture in the atmosphere and soil, adversely affecting its growth by disrupting its basic metabolic balance (Will et al. 2013; Zhang et al. 2020a, b, c). Similar studies involving Pinus armandii Franch. in Qinling Mountains, Picea abies (L.) H. Karst. in the central part of the Ceskomoravska Upland, and four dominant conifer species in western Labrador, Canada, showed that high temperatures in summer led to the formation of narrow rings (Nishimura and Laroque 2011; Rybnícek et al. 2012; Wang et al. 2016).

The adverse effects of current September precipitation on radial growth were intensified with decreasing altitude, which showed a significant negative correlation at low altitudes (r = − 0.484), a slight negative correlation at middle altitudes (r = − 0.252), and no correlation at high altitudes (r = − 0.054) (Figs. 8, 9a). This is consistent with the results of studies investigating the radial growth response to climate in major conifers on Haba Snow Mountain in Southwest China and areas in southern Europe (Caminero et al. 2018; Zhang et al. 2020a, b, c). Temperature decreases rapidly in September (Fig. 9), and the first snowfall and frost will advance if precipitation continues to increase at this time, which increases the risk of chilling and freezing injury to trees (Horimoto and Araki 1999). At the same time, this weather does not facilitate nutrient accumulation and lignification in trees by regulating the activities of soil microorganisms related to the emission and absorption of carbon dioxide (CO₂) and methane (CH₄) (Bukata and Kyser 2007; Bhattacharyya et al. 2013; Wagner et al. 2016; Praeg et al. 2017), which shortens the growing season and prematurely terminate their radial growth (Babst et al. 2014). Nevertheless, the results of this study differ from those associated with the radial growth response of L. olgensis to climate in the Changbai Mountains and Picea crassifolia Kom. in Qilian Mountains (Yu and Liu, 2020; Zhang et al. 2020a, b, c). The possible reasons are as follows: first, the study areas, located in different climatic provinces, exhibit distinct climate variation; second, regional differences lead to different phenological characteristics and growth rhythm of trees; and third, the differences in the microenvironment of the sampling sites may also contribute to these differences.

Slope variability of the growth–climate relationships

The radial growth of L. olgensis was mainly limited by temperature on the shady slope but by the moisture conditions of soil and air on the sunny slope (Figs. 8, 9b). Specifically, on the shady slope, it was significantly negatively correlated with the current August T_mean (r = − 0.288), the current September T_min (r = − 0.315), the current July–August T_max (r = − 0.300, r = − 0.370), and the current May and previous December RH (r = − 0.343, r = − 0.282). It had a significantly positive correlation with the current May T_max (r = 0.332), the current July–August RH (r = 0.356, r = 0.394), and the current August and previous July–September PDSI (r = 0.274, r = 0.279, r = 0.299, r = 0.305). Because of the shorter duration and weaker intensity of direct solar radiation on the shady slope (Chi et al. 1982), tree growth increases the demand on temperature, particularly at the beginning of and the growing season. Temperature increases in May during the start of L. olgensis growth in this region (Ogden 1981; Wang et al. 2011; Silva et al. 2019), accelerating the metabolic activities of soil microorganisms such as methanotrophs (Praeg et al. 2017) and promoting carbon and nitrogen cycles of the ecosystem (Bukata and Kyser 2007; Bhattacharyya et al. 2013; Babst et al. 2014), which contributes to the accumulation of organic matter in plants (Wagner et al. 2016), thus facilitating the radial growth of trees. Studies on the growth–climate relationships of Larix decidua Mill. in the French Alps, Abies georgei Hand.-Mazz. in Haba Snow Mountain, and the southern part of the Asian boreal forests in Northeast China also corroborated our results (Saulnier et al. 2019; Li et al. 2020a, b, c; Zhang et al. 2020a, b, c).

The PDSI in current June had the strongest impact on the radial growth of L. olgensis on sunny slope, followed by the current March air relative humidity (Figs. 8, 9b). The correlation between the two was positive (r = 0.500) and negative (r = − 0.386), respectively. The results are consistent with those of other relevant studies, in which moisture availability was a major limiting factor for pine forests in Southwest and Northeast China (Zhu et al. 2018; Bi et al. 2020). In March, the mean temperature is still below 0 ℃ in the southern Lesser Khingan Mountains (Fig. 9), when trees are more vulnerable to freezing injury under increased air relative humidity (Horimoto and Araki, 1999). Further, the high temperature in June–August leads to high water evaporation; therefore, trees on the sunny slope benefit from the increased soil moisture (Kim et al. 2011). Studies on the northern and eastern slopes of the Changbai Mountains also established the slope variability of the growth–climate relationships (Yu et al. 2011, 2013; Yu and Liu, 2020).

Simulation of growth–climate relationships

No multicollinearity was found among the explanatory variables, and the statistical characteristics (R² and p-values) of the established models were generally high (Table 3). Temperature had the strongest limiting effect at high altitudes, with an explanation rate of 48.95%. Moisture was the main limiting factor for the radial growth of L. olgensis at low and middle altitudes, accounting for 72.71% and 94.92%, respectively. In addition, temperature had a high limiting effect on the radial growth of L. olgensis on the shady slope (with the explanation rate of 47.60%), while moisture had a high explanation rate of 76.89% on the sunny slope. The established models fit the objective law of the influence of climate factors on the radial growth of L. olgensis and prove the altitude and slope variability in radial growth–climate relationships.

Table 3 Optimal simplified multiple regression model for growth-climate of L. olgensis

Differences in radial growth response to climatic factors cannot be explained merely by geographic location, growth characteristics, microenvironment, and spatial competition. Large-scale climate phenomena, such as global climatic oscillation and land–sea thermal differences, global carbon and nitrogen cycles, and non-climatic pressures also contribute to the differences (Bukata and Kyser 2007; Bhattacharyya et al. 2013; Piraino and Roig 2013; Babst et al. 2014; Venegas-González et al. 2015; Zhu et al. 2017; Yu et al. 2021). Admittedly, our study is only based on L. olgensis in the region and does not cover long-term climate change on a large scale. Therefore, further studies of climate–radial growth relationships with a more extensive and intensive sampling of multiple species are critically important. These efforts will better understand the applicability of current laws and provide a theoretical basis for estimating the carbon stock and sequestration and management of L. olgensis forests, such as afforestation and harvesting.

Conclusions

The radial growth of L. olgensis in the southern Lesser Khingan Mountains shows obvious 5- to 10-year periodicity. The temporal instability mainly occurred in the early growth stage and after 2000. The growth–climate response exhibits distinct altitudinal and slope variability. The radial growth of L. olgensis at low altitudes is mainly affected by precipitation, but also by temperature, especially the high temperature in summer at high altitudes. Temperature is the key climate limiting factor for the distribution of L. olgensis on shady slope. Future climate changes will exacerbate the challenges underlying the adaptive growth of L. olgensis in this region.