[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Next Article in Journal
Deforestation in the Colombian Amazon: Perception of Its Causes and Actors in Puerto Guzmán, Putumayo
Previous Article in Journal
Long-Term Effects of Biochar Application on Soil Heterotrophic Respiration in a Warm–Temperate Oak Forest
You seem to have javascript disabled. Please note that many of the page functionalities won't work as expected without javascript enabled.
 
 
Search for Articles:
Title / Keyword
Author / Affiliation / Email
Journal
Article Type
 
 
Advanced Search
 
Section
Special Issue
Volume
Issue
Number
Page
 
Journals
Forests
Volume 16
Issue 3
10.3390/f16030490
forests-logo
Submit to this Journal Review for this Journal Propose a Special Issue
► Article Menu

Article Menu

share Share announcement Help format_quote Cite
first_page
settings
Order Article Reprints
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Phenological Spatial Divergences Promoted by Climate, Terrain, and Forest Height in a Cold Temperate Forest Landscape: A Case Study of the Greater Khingan Mountain in Hulun Buir, China

by
Yu Tian
1,2,
Lei Wang
1,2,*,
Bingxi Liu
1,2,
Yunlong Yao
3 and
Dawei Xu
1,2
1
College of Landscape Architecture, Northeast Forestry University, Harbin 150040, China
2
Key Lab for Garden Plant Germplasm Development & Landscape Eco-Restoration in Cold Regions of Heilongjiang Province, Harbin 150040, China
3
College of Wildlife and Protected Area, Northeast Forestry University, Harbin 150040, China
*
Author to whom correspondence should be addressed.
Forests 2025, 16(3), 490; https://doi.org/10.3390/f16030490 (registering DOI)
Submission received: 7 February 2025 / Revised: 2 March 2025 / Accepted: 10 March 2025 / Published: 11 March 2025
(This article belongs to the Section Forest Meteorology and Climate Change)
Browse Figures
Figure 1
<p>The study area, (<b>a</b>) location and forest height of Hulun Buir in China; (<b>b</b>) the elevation of Hulun Buir; (<b>c</b>) the slope of Hulun Buir; and (<b>d</b>) the aspect of Hulun Buir.</p> "> Figure 2
<p>Scatter plots of SOS with pre-season temperature (<b>a</b>) and precipitation (<b>c</b>). LISA maps of SOS with pre-season temperature (<b>b</b>) and precipitation (<b>d</b>).</p> "> Figure 3
<p>Scatter plots of EOS with autumn temperature (<b>a</b>) and precipitation (<b>c</b>). LISA maps of EOS with autumn temperature (<b>b</b>) and precipitation (<b>d</b>).</p> "> Figure 4
<p>Scatter plots of SOS with elevation (<b>a</b>) and slope (<b>c</b>). LISA maps of SOS with elevation (<b>b</b>) and slope (<b>d</b>).</p> "> Figure 5
<p>Scatter plots of EOS with elevation (<b>a</b>) and slope (<b>c</b>). LISA maps of EOS with elevation (<b>b</b>) and slope (<b>d</b>).</p> "> Figure 6
<p>Box plots of SOS (<b>a</b>) and EOS (<b>b</b>) in different aspects. The green dashed lines and numbers in each box represent mean values, the orange lines represent median values, and the top and bottom edges of each box represent 75th and 25th percentiles, respectively.</p> "> Figure 7
<p>Scatter plot of SOS with forest height (<b>a</b>). LISA map of SOS with forest height (<b>b</b>).</p> "> Figure 8
<p>Scatter plot of EOS with forest height (<b>a</b>). LISA map of EOS with forest height (<b>b</b>).</p> "> Figure 9
<p>Scatter plots of elevation and slop with pre-season temperature (<b>a</b>,<b>b</b>) and pre-season precipitation (<b>c</b>,<b>d</b>). Scatter plots without fitted curves indicated a non-significant relationship.</p> "> Figure 10
<p>Scatter plots of elevation and slope with autumn temperature (<b>a</b>,<b>b</b>) and autumn precipitation (<b>c</b>,<b>d</b>). Scatter plots without fitted curves indicated a non-significant relationship.</p> ">
Versions Notes

Abstract

:
Vegetation phenology has attracted considerable attention as one of the most sensitive indicators of global climate change. Remote sensing has significantly expanded our understanding of the spatial divergences of vegetation phenology. However, the current understanding of the reasons behind spatial divergences of vegetation phenology is not yet complete, and there is an urgent need to unravel the landscape processes driving spatial divergences of vegetation phenology. In light of this, the present study focused on montane forests of the cold temperate zone as its study area, collecting datasets such as the MCD12Q2 land surface phenology product, climate, topography, and stand height and adopting regression analysis and geo-detector model to investigate the individual and interactive effects of variables such as temperature, precipitation, elevation, slope, aspect, and forest height on forest phenology. The results indicated that because of the complexity of topography, the impacts of temperature on forest phenology were nonlinear. With fluctuation of elevation, the development of forest occurred later at the base and ridges of mountain and earlier in the valley bottom lands and mid-upper slopes. Temperature and precipitation exhibited a bilaterally strong interactive effect with slope on forest greenup. Both forest greenup and dormancy occurred earlier on shady slopes and later on sunny slopes. There may also exist an interactive effect between forest height and topographic factors on the spatial divergences of forest phenology. Future research may need to focus on whether there is a trade-off or synergy between the macroclimatic regulatory function of topography and the microclimatic regulatory function of canopy structure.

1. Introduction

The World Meteorological Organization (WMO) confirmed on 10 January 2025 that 2024 was thewarmest year on record at about 1.55 °C above pre-industrial level (https://wmo.int/news (accessed on 15 January 2025)). Vegetation phenology is considered one of the most sensitive indicators of global climate change [1,2], contributing to a more comprehensive understanding of the impact of environmental changes on terrestrial ecosystems [3,4]. Vegetation phenology refers to the natural change processes of plants, including germination, maturity, defoliation, dormancy, etc., with the periodic change of seasonal climate [5,6,7,8,9]. As a result, the relevant studies on climate change and vegetation phenology have become a hot topic in various disciplines in recent years [10,11], especially with the help of constantly evolving remote sensing technologies [12,13]. However, despite the continuous expansion of study results, we still lack a comprehensive understanding of the mechanisms underlying phenological divergences of vegetation, which may hinder the development of predictive models for terrestrial phenology [14,15,16].
Previous studies have provided ample evidence that temperature is one of the main factors promoting vegetation phenological changes [17,18,19]. Specifically, as global warming has intensified, the start of growing season (SOS) has advanced while the end of growing season (EOS) has been delayed [1]. Additional studies have also discovered that the effect of precipitation on vegetation phenology was considerable and should not be overlooked [20,21,22]. For instance, on the Tibetan Plateau, precipitation was the dominant factor among climate variables influencing the spring phenology of grasslands [21]; SOS can be delayed even under a suitable temperature condition when precipitation of pre-season was insufficient [22]. It can be seen that climatic conditions such as temperature and precipitation play a crucial role in phenological divergences of vegetation. However, the impact of climate change on vegetation phenology was not entirely consistent. For example, in different cities of the United States and China, SOS of urban vegetation showed different trends of advance and delay with the increase in pre-season temperature, respectively [23,24]. It can be inferred that climatic conditions may be coupled with other factors to affect phenological divergences of vegetation. Therefore, it is necessary and significant to explore other factors in addition to climatic conditions and their interactive effects on vegetation phenology.
At larger spatial scales, phenological divergences may be largely consistent with climate chan.es across a wide range of altitude gradients [3,25]. But at fine spatial scales, phenological patterns may also include various local processes interacting with each other, which likely contributes to significant variations in phenological patterns [26,27,28]. For examples, topography can alter microclimates by adjusting temperature and precipitation, such as processes contributing to cold air drainage into small valleys, thereby further affecting phenological variations of vegetation [29,30,31]; Continuously rising temperatures may advance SOS in low altitude areas, but with altitude gradients rising, SOS will possibly experience a nonlinear delay [32]. It can be seen that due to spatial variability, direct and indirect processes may be included at more local scales, and climatic and geographical variables may interact to affect vegetation phenology, resulting in phenological divergences of vegetation in different regions. Therefore, it is of great significance to explore how phenological divergences of vegetation are regulated by topographical variations and their interaction with climatic conditions.
A recent pioneering study demonstrated that canopy structure of temperate forests had a complex relationship with their autumn phenology [33]. One of the reasons for this phenomenon may be that canopy structure plays a crucial role in regulating microclimate conditions of natural forests, further contributing to phenological divergences [34,35]. Specifically, canopy structure can regulate temperature by dispersing heat through evapotranspiration and can also attenuate light intensity by reflecting, scattering, and absorbing solar radiation [36]. Another study suggested that variability in forest height can explain phenological divergences of plants [37]. However, due to the limited scope of obtaining canopy structure parameters through unmanned aerial vehicle field surveys or station observations, it is impossible to further investigate the spatial heterogeneity of relationship between canopy structure and forest phenology. Recently, some researchers have adopted LiDAR forest plot datasets, combined with multi-source remote sensing data and machine learning to predict and generate a map of 2020 Chinese forest stand average height [38]. This work provided a good data source for exploring the relationship between forest height and its phenology.
In addition, previous studies on forest phenology have predominantly focused on the warm temperate and mid-temperate zones [3,6,10,39,40,41], with relatively fewer studies on the driving mechanisms of forest phenology in the cold temperate zone. One study reported that temperature and soil moisture content may influence the radiation use efficiency of forests, thereby further regulating photosynthesis [42], which is closely related to forest phenology. Specifically, temperature may be more critical for the photosynthetic efficiency of forests in cold regions [43]. Soil moisture content was considered to be a significant impact factor on the radiation use efficiency of cold forests [44], but it was less significant in other cold forests [45]. Therefore, we will conduct relevant research on phenology in a forest within the cold temperate zone, aiming to provide comparative cases for studies in more diverse climatic regions.
To sum up, in order to provide a new perspective for a deeper understanding of the impact of climate changes on ecosystems and provide relevant constraints for improving vegetation phenology prediction models in the future, this study will undertake the following tasks: (1) We will collect a series of data, including regional climatic conditions (monthly temperature and precipitation), terrain factors (elevation, slope, and aspect), and forest height and, then, explore their relationships and spatial heterogeneity with forest phenology through regression analysis and spatial clustering. (2) We will further verify interactive relationships between the above factors and forest phenology through multiple regression and geo-detector model, discussing whether terrain variability can regulate climate change and, thus, contribute to phenological spatial divergence.

2. Materials and Methods

2.1. Study Area

Our study intends to select a region rich in forest resources and characterized by complex topography from the cold temperate zone of China as the study area. In light of this, Hulun Buir in northeastern China is an ideal study area. It is situated at the junction of the mid-temperate and cold temperate zones, and the majority of the Greater Khingan Mountains are located here. The Greater Khingan Mountains run through the central part of Hulun Buir in a direction from northeast to southwest, with forests as the main land use type and an altitude of 700–1700 m (Figure 1).
Hulun Buir is a prefecture-level city under the jurisdiction of Inner Mongolia Autonomous Region, China. It is located in the northeast of Inner Mongolia Autonomous Region (47°15′–53°45′ N, 116°49′–127°04′ E), adjacent to the Heilongjiang Province to the east and bordering Mongolia and Russia to the west and north (Figure 1a). The total area is approximately 253,000 km2. The western part of the Greater Khingan Mountains is the Hulun Buir grassland, which is a grassland animal husbandry economic zone with an altitude of 550–1000 m. The eastern part of the Greater Khingan Mountains is characterized by low mountains, hills, and river valley plains, and it is an agricultural economic zone dominated by planting, with an altitude of 200–500 m (Figure 1b–d). The annual average temperature is between −1.5 °C and 1.8 °C. The eastern region is a monsoon climate zone with an annual precipitation of 500–800 mm; the western region is a continental climate zone with an annual precipitation of 300–500 mm [46].

2.2. Data Collection

2.2.1. Forest Stand Mean Height and Coverage

The forest stand mean height dataset with a spatial resolution of 30 m was provided by the National Ecosystem Science Data Center, National Science & Technology Infrastructure of China (http://www.nesdc.org.cn/sdo/detail?id=67467d197e281765a3564ccc (accessed on 2 January 2025)). This dataset has adopted LiDAR forest plot datasets, combined with multi-source remote sensing data and machine learning to predict and generate a map of 2020 Chinese forest stand mean height (Figure 1a) [38]. Therefore, in order to match all the latest available datasets, we have focused our study on 2020. Then, we have defined the regions containing forest height values as forest coverage. Finally, we have adopted a 5 km × 5 km fishing net to segment forest and employed 5786 grids containing forest height values as the statistical units.

2.2.2. Spring and Autumn Phenology

The phenological dataset in this study was obtained from the MODIS Land Cover Dynamics (MCD12Q2) products with a spatial resolution of 500 m, which were derived from time series of the 2-band enhanced vegetation index (EVI2) [47]. We adopted SOS of forest to represent its spring phenology and EOS to represent its autumn phenology. In order to reduce inaccuracy in subsequent analysis, we removed the extreme values in the phenological datasets based on threshold method according to previous studies [13,48]. Specifically, pixels outside the 30th to 180th days of year (DOY) in SOS were removed, and pixels outside the 240th to 350th DOY in EOS were also removed.

2.2.3. Climatic Factors

The monthly air temperature dataset with a spatial resolution of 1000 m was provided by the Institute of Geographic Sciences and Natural Resources Research, CAS, Resource and Environment Science and Data Center (https://www.resdc.cn/data.aspx?DATAID=349 (accessed on 2 January 2025)). The monthly precipitation dataset was provided by the National Tibetan Plateau, Third Pole Environment Data Center, and National Science & Technology Infrastructure of China (http://data.tpdc.ac.cn/zh-hans/data/faae7605-a0f2-4d18-b28f-5cee413766a2 (accessed on 2 January 2025)). According to previous studies [49,50,51], in order to fully cover the time series of forest phenology as much as possible, we defined 1 December of previous year to 31 May of following year as pre-season and 1 September to 30 November as autumn, and then calculated the mean values for each season.

2.2.4. Topographical Factors

The digital elevation model (DEM) adopted in our study was obtained from the Institute of Geographic Sciences and Natural Resources Research, CAS, Resource and Environment Science and Data Center (https://www.resdc.cn/data.aspx?DATAID=123 (accessed on 2 January 2025)) with a spatial resolution of 500m. The elevation values were obtained from the DEM. Finally, all raster data were resampled to the same spatial resolution of 500 m by a nearest neighbor approach.

2.3. Methodology

2.3.1. Linear and Curve Regression Analysis

In this study, univariate linear regression and curve regression models were adopted to analyze the single relationships between forest phenology and climate factors, topographical factors, and forest height, while multiple linear regression model was adopted to analyze the combined effects of these factors on forest phenology. In the binary regression analysis, we initially determined the linear and nonlinear relationships between variables by plotting scatter diagrams. Subsequently, we decided which curve model to adopt based on the fitting results of different curve regression models. Under the premise of satisfying significance (p < 0.05), we selected a curve regression model with a larger R2 and smaller and randomly distributed residuals while also considering the differences in AIC values.

2.3.2. Local Indicator of Spatial Association

Spatial cluster identification was adopted to assess spatial distribution of the relationships between forest phenology and climate factors, topographical factors, and forest height. First, we adopted air temperature, precipitation, elevation, slope, and forest height as the first variable, respectively, and forest phenology as the second variable. Then, the bivariate local Moran’s I [52] values and statistical significance were calculated for each grid unit. The local indicator of spatial association (LISA) was computed using the following equations:
P k P i = Z k i ∑ j = 1 n w i j Z P j
Z k i = X k i − X ¯ k σ k
Z P i = X P i − X ¯ P σ P
where wij is the spatial weight matrix; Xik is air temperature, precipitation, elevation, slope, and forest height of each grid unit, respectively; XiP is forest phenology; X ¯ k and X ¯ P are average values of Xik and XiP, respectively; and σk and σP are variances of Xik and XiP, respectively [53,54].
The LISA maps can present four types of spatial association: high/high, low/low, high/low, and low/high. In this study, the preceding “high or low” in each spatial cluster referred to forest phenology, while the subsequent “high or low” referred to climatic, topographic, and forest height factors, respectively. For example, the spatial clustering “high/low” of pre-season temperature and forest SOS represented a late SOS with low temperature. We have provided detailed legends in all subsequent LISA maps.

2.3.3. Geo-Detector Analysis

The geo-detector is a statistical tool designed to identify and analyze spatial stratified heterogeneity, as well as to determine the underlying factors driving such heterogeneity. The method is based on the core premise that a significant influence of an independent variable on a dependent variable will result in a strong resemblance in their spatial distribution patterns [55,56]. The geo-detector includes four detectors: factor detector, interaction detector, risk detector, and ecological detector. The specific software package and detailed information can be found on the following website: http://www.geodetector.cn (accessed on 2 January 2025).
In this study, our primary purpose of utilizing geo-detector was to assess the interactive effects of air temperature, precipitation, elevation, slope, and forest height on forest phenology. Therefore, we only adopted the interaction detector, which can be used to identify interaction between two different independent variables such as X1 and X2 on the dependent variable such as Y. By comparing q (X1) and q (X2) with q (X1 ∩ X2), one can evaluate whether the interaction between X1 and X2 will amplify or diminish their explanatory power regarding Y. The calculation method for q is as follows [57]:
q = 1 − ∑ h = 1 L N h σ h 2 N σ 2 = 1 − SSW SST       q ∈ [ 0 , 1 ]
SSW = ∑ h = 1 L N h σ h 2 ,       SST = N σ 2
where N is the number of samples constituting dependent variable Y, L is the number of strata for independent variable(s) X, stratum h ∈ [1, 2, …, L] is composed of Nh samples, σ2h h is variance in stratum h, σ2 is variance of the population, SSW is the within sum of squares, and SST is the total sum of squares. The larger the q value, the stronger the spatial stratified heterogeneity of Y. If the strata are defined by X, the q value presents the degree to which X explains Y.
The five possible types of interactions are shown in Table 1. In this study, we consider forest phenology as the dependent variable, with air temperature, precipitation, elevation, slope, and tree height as the independent variables. According to previous studies [57,58] and verification, we have categorized each independent variable into 7 types using the K-means algorithm.

3. Results

3.1. Effects of Climatic Factors on Forest Phenology

The effect of pre-season temperature on forest SOS was nonlinear. Specifically, with an increase in pre-season temperature, forest SOS exhibited a trend of first delaying and then advancing (Figure 2a). It was worth noting that the mean pre-season temperatures were all below 0 °C, and when temperatures ranged from −9 °C to −8 °C, the lower the temperature, the earlier the forest SOS occurred. In regions where SOS occurred later and temperature was lower, the terrain tended to be more elevated. Conversely, in regions where SOS occurred earlier and temperature was higher, the elevation was relatively lower (Figure 1b and Figure 2b). Moreover, in certain low-elevation areas (e.g., northeast regions), there was a positive correlation between high pre-season temperature and delayed forest SOS. At the same time, in certain high-elevation areas (e.g., north and northwest regions), there was a positive correlation between low pre-season temperature and advanced forest SOS. These results indicated that elevation may modulate the effect of pre-season temperature on forest SOS.
Pre-season precipitation had a positive linear effect on forest SOS (Figure 2c). The number of grids classified as high/high and low/low was nearly twice that of low/high and high/low (Figure 2d). In regions where forest SOS was earlier and precipitation was more plentiful, the temperatures were also relatively higher.
The effect of autumn temperatures on EOS was precisely the opposite of the effect of pre-season temperatures on SOS, that is, as autumn temperatures rose, forest EOS initially advanced and then was delayed, with a higher R2 of 0.316 (Figure 3a). The regions where autumn temperatures were positively correlated with EOS (high/high and low/low areas) overwhelmingly dominated (Figure 3b).
Although the correlation between precipitation and EOS was significant (p < 0.001), R2 and slope values were exceedingly low, merely 0.003 and −0.006, respectively (Figure 3c). The number of grids with earlier EOS and abundant precipitation was the highest, significantly surpassing the other three categories, and most of these areas were characterized by higher elevations (Figure 1b and Figure 3d).

3.2. Effects of Topographical Factors on Forest Phenology

With the continuous increase in elevation, forest SOS showed a trend of first advancing and then being delayed (Figure 4a). Across the entire study area, regions where elevation was positively correlated with forest SOS constituted the predominant type. However, in the northern and northeastern parts of the study area, there were a few regions where elevation was negatively correlated with forest SOS (Figure 4b). This result may be the reason why elevation had a nonlinear impact on forest SOS.
Slope exhibited a significant linear negative correlation with forest SOS. For every 1° increase in slope, forest SOS advanced by approximately 1.3 days (Figure 4c). Furthermore, in LISA map, the number of high/low and low/high grids significantly surpasses those of high/high and low/low grids (Figure 4d).
In contrast to the nonlinear relationship between elevation and forest SOS, elevation exhibited a significant linear negative correlation with forest EOS, with a higher R2 value of 0.203 (Figure 5a). In the LISA map, the effect of elevation on forest EOS was also similar, with forest EOS occurring later in low-elevation areas and earlier in high-elevation areas. Only 686 grids exhibited a positive correlation between two variables (Figure 5b).
Although the correlation between slope and EOS was significant, the level of significance was relatively low (p < 0.05), and the R2 value was exceedingly minimal, merely 0.001 (Figure 5c). By overlaying Figure 5b,d, it was evident that in northeastern and southern regions of the study area, there was a negative correlation between topographic factors and forest EOS, indicating that forest EOS may occur later in terrains with low elevation and gentle slope, and vice versa.
Since the aspect data utilized in this study were categorical rather than continuous like elevation and slope data, it was not possible to generate scatter plots or LISA maps. However, we have calculated the average forest phenology for each aspect category within every grid and the entire study area and have illustrated these findings using tables and box plots (Figure 6 and Table 2).
From the perspective of the entire study area, forest SOS occurred relatively earlier in the first quadrant (north, northeast, and east aspects), reaching its earliest point in the due east aspect. Conversely, it was relatively later in the third quadrant (south, southwest, and west aspects), with the latest occurrence in the due west aspect. Forest EOS occurred relatively earlier in the northern aspects (northwest, north, and northeast), reaching its earliest point in the due north aspect. In contrast, it was relatively later in the southern aspects (southeast, south, and southwest), with the latest occurrence in the due south aspect (Table 2).
According to the statistical results from each grid, the occurrence of numerous outliers in average values of SOS and EOS indicated a high level of spatial heterogeneity in forest phenology. Forest SOS was similarly earlier in the first quadrant (north, northeast, and east aspects) and later in the third quadrant (south, southwest, and west aspects), with the earliest occurrence in northeast aspect and the latest in southwest aspect (Figure 6a). Forest EOS was similarly earlier in the northern aspects (northwest, north, and northeast) and later in the southern aspects (southeast, south, and southwest), with the earliest occurrence in north aspect and the latest in southeast aspect (Figure 6b).
In comparison, whether considering the entire study area or each grid, the trends in forest phenology were quite similar, suggesting that the phenological characteristics of forest remained largely consistent across these two scales.

3.3. Effects of Forest Height on Forest Phenology

Overall, there was a significant linear negative correlation between tree height and its SOS, with SOS advancing by about 0.68 days for every 1m increase in tree height (Figure 7a). From the LISA map, it was evident that high/high spatial clustering predominantly occurred in high-elevation areas, whereas low/low spatial clustering was more concentrated in low-elevation areas. This suggested that the combination of tree height and topography may regulate the process of forest SOS (Figure 1b and Figure 7b).
Similarly, there was a significant linear negative correlation between tree height and its EOS, with EOS advancing by about 0.53 days for every 1m increase in tree height (Figure 8a). Unlike the spatial clustering of SOS and tree height, in high-elevation areas, taller trees exhibited an earlier EOS, whereas in low-elevation areas, shorter trees experienced a later EOS (Figure 1b and Figure 8b).

3.4. Effects of Topographic Factors on Climate

In order to further verify whether topography can alter forest phenology by modulating macroclimate, we conducted a regression analysis between climatic factors and topographic factors (Figure 9 and Figure 10). The results exhibited significant negative correlations between elevation and air temperatures in both seasons, with R2 values above 0.3. However, it was worth noting that the scatter points at lower parts of scatter plots, or those with lower temperatures, exhibited some degree of separation (Figure 9a and Figure 10a). Moreover, as elevation rose, pre-season precipitation showed a trend of first decreasing and then increasing (Figure 9b).
We have statistically analyzed the mean temperatures and precipitations across two seasons for eight aspects of the entire study area (Table 3). The pre-season temperatures in the northeast and east aspects were relatively higher, followed by the south and southeast aspects, while the autumn temperatures were higher in the northeast, west, and north aspects but lower in the south and southeast aspects. The rankings of precipitation across eight aspects were quite similar for both seasons, with the most abundant precipitation occurring in the southeast aspect, followed by the east and south slopes and, finally, the west and northwest slopes.

3.5. Interactive Effects of Various Factors on Forest Phenology

In order to investigate the combined effects of different factors on forest phenology, we incorporated climatic, topographical, and forest height variables into the multiple linear regression model, and the resulting parameter estimates are detailed in Table 4. The VIF values for all independent variables range from a minimum of 1.088 to a maximum of 2.024, indicating a low level of multicollinearity among them. Except for non-significant effect of forest height on its SOS (p = 0.062 > 0.05), the other independent variables had significant effects on both forest SOS and EOS (p < 0.001). By comparing the β values of two sets of multiple linear regression models, it was evident that temperature and slope had a stronger impact on forest phenology compared to precipitation and elevation. Although the effect of forest height on its EOS was relatively weaker, it should not be overlooked.
In this study, we employed the interaction detector to investigate whether the components of different climatic, topographical, and forest height factors exhibited interactions that either weaken or enhance forest phenology (Table 5).
The interactions between pre-season temperature and the other four variables on forest SOS were relatively strong, with q values all exceeding 0.3. Notably, the q values for the interaction effects of pre-season temperature with elevation and slope on forest SOS even surpassed 0.4, indicating that pre-season temperature and topography significantly promoted or inhibited the greenup of forest. The interactive effect type between pre-season temperature and precipitation on forest SOS was “Enhance, Bilaterally”, indicating that suitable temperature combined with adequate precipitation may promote forest greenup. In addition, although the q values of precipitation ∩ slope and elevation ∩ slope were not as high, these effects cannot be ignored.
Similarly, the interactions between autumn temperature and the other four variables on forest EOS were also relatively strong, and their q values were relatively close. The interactive effect type between autumn temperature and forest height on EOS was “Enhance, Bilaterally”, with a q value close to 0.35, indicating that forest height may affect the temperature distribution within woodland, thereby modulating its autumn phenological process.

4. Discussion

In this study, we established regression and spatial models between two climatic factors (temperature and precipitation) and the phenology of cold-temperate mountain forests, utilizing satellite phenological products and climatic datasets. Our research has found that temperature exhibited a nonlinear relationship with the spring and autumn phenology of forests across spatial dimensions (Figure 2a and Figure 3a). This deviated somewhat from linear results of previous studies obtained in urban areas with relatively homogeneous topography [29,51,59], indicating that topography may play a certain role in the process of temperature regulation on forest phenology. Just as some other related studies have reiterated, topography and vegetation characteristics hold dual significance for forest phenology [3,60,61,62]. In addition, the interactive effects of pre-season temperature and topography on forest SOS were relatively stronger than those of autumn temperature and topography on forest EOS (Table 5). Therefore, our study contributed to enhancing the understanding of climatic impact mechanisms underlying the spatial divergences of forest phenology, that is, when exploring the relationship between temperature and forest phenology, particularly the relationship between pre-season temperature and forest SOS, it is quite essential to take into account the topography of the study area.
Our results indicated that forest SOS occurred earlier when temperatures were lower (from −9 °C to −8 °C) (Figure 2a), primarily in the northern and northwestern parts of the study area (Figure 2b). This finding was inconsistent with some previous research outcomes. It may be related to the arrival of a warm spring following low winter temperatures. Rapid warming can contribute to an earlier SOS, but this situation was not directly caused by low temperatures; rather, it is the result of temperature changes [63]. Moreover, temperature may be more critical for the photosynthetic efficiency of forests in cold regions [43].
As indicated by the R2 of regression model, the effect of precipitation on forest phenology was not as significant as that of temperature (Figure 2 and Figure 3). This could be because in more humid regions, the lower the risk of drought, the higher the temperature sensitivity of forest phenology, potentially maximizing thermal benefits, and the weaker the correlation between forest phenology and pre-season precipitation [20]. Our study found that the effect of precipitation on forest SOS was positive, and high precipitation did not necessarily delay forest EOS in high-elevation regions (Figure 2c and Figure 3d). The reason for these phenomena may be that, under unsuitable temperature conditions, although soil moisture content is sufficient, forest SOS may be also more likely to occur later, and EOS may occur earlier [22].
Our study has found that the relationship between elevation and forest SOS was non-linear, whereas it was linear with forest EOS (Figure 4a and Figure 5a), which was highly consistent with the results of previous studies [3,8]. The terrain undulation is considered to support phenological related microclimates, and in regions where the undulation of altitude is minimal, there may exist enduring microclimates [27,64]. The elevation fluctuations often have a close relationship with temperature variability [8,65], and our study has confirmed this point (Figure 9a and Figure 10a). With elevation fluctuations, the development of forest occurred later at the base and ridges of the mountain, and earlier in the valley bottom lands and mid-upper slopes (Figure 4b). Moreover, the dormancy of forests occurred earlier at the ridges and later on the slopes and at the mountain base (Figure 5b).
Previous studies have consistently regarded elevation as the primary topographic factor regulating vegetation phenology, whereas the effects of slope and aspect on vegetation phenology were also significant and should not be overlooked [61,66]. Our research has found that an increase in slope may lead to an advancement in forest SOS and a delay in forest EOS (Figure 4b and Figure 5b). The reason for these phenomena may be that, on steep slopes, the direct angle of sunlight is greater, contributing to faster soil warming [65]. Additionally, snow accumulation at the end of winter and beginning of spring tends to melt more readily [67]. Under these conditions of suitable temperature and ample soil moisture, forests are more likely to turn green. Another finding of ours provided evidence for this as well, that is, temperature and precipitation exhibited a strong interactive effect, combined with slope, on forest SOS, and the type of interaction was “Enhance, Bilaterally” in both cases (Table 5). However, on steep slopes at high elevations, forest SOS can still be delayed (Figure 4d) and EOS can still be advanced (Figure 5d), which may be related to the more intense solar radiation on steep slopes at high altitudes [68]. Accumulated solar radiation may lead to surface warming, increase soil moisture evaporation, and, consequently, delay the forest’s SOS [22]. Excessive solar radiation and warming environment may promote higher vegetation productivity, contributing to earlier carbon limitation and EOS [10,69].
Aspect, as another crucial topographic factor, can directly influence the amount of solar radiation received by forest ecosystems, thereby affecting ecological characteristics such as air temperature, soil moisture, evapotranspiration rates, etc. [70,71]. Our study has discovered that, at both the scale of entire study area and the grid scale, forest SOS and EOS occurred earlier on shady slopes and later on sunny slopes (Table 2 and Figure 6). Among these findings, the discovery related to forest EOS aligned with a previous study, that is, ample sunlight on sunny slopes allows trees to grow for a longer period and enter dormancy later [72]. However, it is important to emphasize that soil temperature and evaporation rates can increase with the slope on southern slopes [69]. Intense evaporation may lead to soil moisture stress, resulting in an earlier arrival of forest EOS [73]. The discovery related to forest SOS was inconsistent with some previous studies. To this end, we calculated the average temperature and precipitation for each slope aspect within the study area. The results showed that temperatures in the northeast and east aspects were higher (Table 3), which may be the reason why forest SOS occurred earlier on shady slopes. The distribution of temperature across slope aspects may be related to factors such as microclimate, vegetation type, human activities, etc. [30,67,72,74].
Unlike external environmental factors such as climate and topography, tree height is an intrinsic physiological characteristic of the forest. Relevant studies have also found that phenological divergences of forests can be explained by differences in their height [37,75] as tree height may be associated with local microclimates [34]. Our study found that forest height displayed negatively correlated with both its SOS and EOS, which appeared to be consistent with previous studies [3,33]. As height increases, the regulatory capacity of canopy structure on microclimate may diminish [76]. Therefore, the extent to which microclimate influences phenology may vary along the gradient of forest height. We have also found that forest height may interactively influence its phenology in conjunction with topographic factors (Table 5). We hypothesize that this is because the regulatory effect of topography on macroclimate may have trade-offs or synergies with the microclimate regulation within the canopy. Future works on vegetation phenology should take into account microclimatic factors within the canopy and delve deeper into the driving mechanisms of canopy structure on phenology. In addition, some studies have already begun to investigate the driving mechanisms of phenology at the local scale of forests, such as the impact of landscape fragmentation and abundance, natural disturbances, and human management interventions on the forest and its phenology [30,77]. Exploring the variation characteristics of phenology at a more refined scale may be a valuable direction in the future.

5. Conclusions

Remote sensing technology has made it possible to monitor the spatial variations in forest phenology. The factors contributing to the spatial differentiation of forest phenology may be diverse, including but not limited to climate, topography, and canopy structure, among others. In light of this, the present study focused on montane forests of the cold temperate zone as its study area, investigating the individual and interactive effects of factors such as climate, topography, and tree height on forest phenology. The results of this study indicated that temperature variation was a pivotal factor affecting the spatial divergences of forest phenology. However, due to the modulating role of topographic factors on temperature, the effects of temperature on forest phenology may be nonlinear. This is particularly true when investigating the mechanisms by which temperature affects forest SOS where the impact of topographic elements must be taken into account. Owing to the study area’s location within a relatively humid zone, forest phenology may not be as sensitive to precipitation as it is to temperature. Due to the complexity of elevation, slope, and aspect, there is a spatial heterogeneity in ecological parameters such as temperature, soil moisture, solar radiation, etc., which further contribute to the spatial divergences of forest phenology. Structural parameters such as forest height may affect forest phenology through the regulation of microclimates, and there is also an interactive effect between forest height and topographic factors on the spatial divergences of forest phenology. Our study can provide relevant constraints for improving vegetation phenology prediction models in the future and serve as a reference for formulating environmental restoration strategies.

Author Contributions

Conceptualization, Y.T. and B.L.; methodology, Y.T., B.L. and Y.Y.; software, Y.T. and L.W.; investigation, Y.T. and B.L.; writing—original draft preparation, Y.T.; writing—review and editing, L.W., B.L. and Y.Y.; supervision, L.W. and D.X.; project administration, L.W. and Y.Y.; funding acquisition, L.W. and Y.Y. All authors have read and agreed to the published version of the manuscript.

Funding

This study was supported by the National Natural Science Foundation of China (No. 42171246).

Data Availability Statement

The data presented in this study are available on request from the corresponding author.

Acknowledgments

We would like to extend our gratitude to all the anonymous reviewers for their peer review of this article.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Piao, S.; Liu, Q.; Chen, A.; Janssens, I.A.; Fu, Y.; Dai, J.; Liu, L.; Lian, X.; Shen, M.; Zhu, X. Plant phenology and global climate change: Current progresses and challenges. Glob. Change Biol. 2019, 25, 1922–1940. [Google Scholar] [CrossRef] [PubMed]
  2. Chuine, I.; Morin, X.; Bugmann, H. Warming, photoperiods, and tree phenology. Science 2010, 329, 277–278. [Google Scholar] [CrossRef]
  3. Adams, B.; Matthews, S.; Iverson, L.; Prasad, A.; Peters, M.; Zhao, K. Spring phenological variability promoted by topography and vegetation assembly processes in a temperate forest landscape. Agric. For. Meteorol. 2021, 308, 108578. [Google Scholar] [CrossRef]
  4. Zhang, X.Y.; Friedl, M.A.; Schaaf, C.B.; Strahler, A.H.; Hodges, J.C.F.; Gao, F.; Reed, B.C.; Huete, A. Monitoring vegetation phenology using MODIS. Remote Sens. Environ. 2003, 84, 471–475. [Google Scholar] [CrossRef]
  5. Broich, M.; Huete, A.; Tulbure, M.G.; Ma, X.; Xin, Q.; Paget, M.; Restrepo-Coupe, N.; Davies, K.; Devadas, R.; Held, A. Land surface phenological response to decadal climate variability across Australia using satellite remote sensing. Biogeosciences 2014, 11, 5181–5198. [Google Scholar] [CrossRef]
  6. Yang, Y.; Guan, H.; Shen, M.; Liang, W.; Jiang, L. Changes in autumn vegetation dormancy onset date and the climate controls across temperate ecosystems in China from 1982 to 2010. Glob. Change Biol. 2015, 21, 652–665. [Google Scholar] [CrossRef]
  7. Wang, X.; Xiao, J.; Li, X.; Cheng, G.; Ma, M.; Zhu, G.; Arain, M.A.; Black, T.A.; Jassal, R.S. No trends in spring and autumn phenology during the global warming hiatus. Nat. Commun. 2019, 10, 2389. [Google Scholar] [CrossRef]
  8. Li, X.; Du, H.; Zhou, G.; Mao, F.; Zhang, M.; Han, N.; Fan, W.; Liu, H.; Huang, Z.; He, S.; et al. Phenology estimation of subtropical bamboo forests based on assimilated MODIS LAI time series data. ISPRS-J. Photogramm. Remote Sens. 2021, 173, 262–277. [Google Scholar] [CrossRef]
  9. Garonna, I.; de Jong, R.; Stockli, R.; Schmid, B.; Schenkel, D.; Schimel, D.; Schaepman, M.E. Shifting relative importance of climatic constraints on land surface phenology. Environ. Res. Lett. 2018, 13, 024025. [Google Scholar] [CrossRef]
  10. Zani, D.; Crowther, T.W.; Mo, L.; Renner, S.S.; Zohner, C.M. Increased growing-season productivity drives earlier autumn leaf senescence in temperate trees. Science 2020, 370, 1066–1071. [Google Scholar] [CrossRef]
  11. Beil, I.; Kreyling, J.; Meyer, C.; Lemcke, N.; Malyshev, A. Late to bed, late to rise-Warmer autumn temperatures delay spring phenology by delaying dormancy. Glob. Change Biol. 2021, 27, 5806–5817. [Google Scholar] [CrossRef] [PubMed]
  12. Wu, L.; Ma, X.; Dou, X.; Zhu, J.; Zhao, C. Impacts of climate change on vegetation phenology and net primary productivity in arid Central Asia. Sci. Total Environ. 2021, 796, 149055. [Google Scholar] [CrossRef]
  13. Zhou, D.; Zhao, S.; Zhang, L.; Liu, S. Remotely sensed assessment of urbanization effects on vegetation phenology in China’s 32 major cities. Remote Sens. Environ. 2016, 176, 272–281. [Google Scholar] [CrossRef]
  14. Liu, Z.; Zhou, Y.; Feng, Z. Response of vegetation phenology to urbanization in urban agglomeration areas: A dynamic urban-rural gradient perspective. Sci. Total Environ. 2023, 864, 161109. [Google Scholar] [CrossRef] [PubMed]
  15. Basler, D. Evaluating phenological models for the prediction of leaf-out dates in six temperate tree species across central Europe. Agric. For. Meteorol. 2016, 217, 10–21. [Google Scholar] [CrossRef]
  16. Richardson, A.D.; Keenan, T.F.; Migliavacca, M.; Ryu, Y.; Sonnentag, O.; Toomey, M. Climate change, phenology, and phenological control of vegetation feedbacks to the climate system. Agric. For. Meteorol. 2013, 169, 156–173. [Google Scholar] [CrossRef]
  17. Ren, P.; Liu, Z.; Zhou, X.; Peng, C.; Xiao, J.; Wang, S.; Li, X.; Li, P. Strong controls of daily minimum temperature on the autumn photosynthetic phenology of subtropical vegetation in China. For. Ecosyst. 2021, 8, 31. [Google Scholar] [CrossRef] [PubMed]
  18. Meng, L.; Mao, J.; Zhou, Y.; Richardson, A.D.; Lee, X.; Thornton, P.E.; Ricciuto, D.M.; Li, X.; Dai, Y.; Shi, X.; et al. Urban warming advances spring phenology but reduces the response of phenology to temperature in the conterminous United States. Proc. Natl. Acad. Sci. USA 2020, 117, 4228–4233. [Google Scholar] [CrossRef]
  19. Kabano, P.; Lindley, S.; Harris, A. Evidence of urban heat island impacts on the vegetation growing season length in a tropical city. Landsc. Urban Plan. 2021, 206, 103989. [Google Scholar] [CrossRef]
  20. Shen, M.; Piao, S.; Cong, N.; Zhang, G.; Janssens, I.A. Precipitation impacts on vegetation spring phenology on the Tibetan Plateau. Glob. Change Biol. 2015, 21, 3647–3656. [Google Scholar] [CrossRef]
  21. Cheng, M.; Wang, Y.; Zhu, J.; Pan, Y. Precipitation dominates the relative contributions of climate factors to grasslands spring phenology on the Tibetan Plateau. Remote Sens. 2022, 14, 517. [Google Scholar] [CrossRef]
  22. Du, J.; He, Z.; Piatek, K.B.; Chen, L.; Lin, P.; Zhu, X. Interacting effects of temperature and precipitation on climatic sensitivity of spring vegetation green-up in arid mountains of China. Agric. For. Meteorol. 2019, 269, 71–77. [Google Scholar] [CrossRef]
  23. Wang, X.; Du, P.; Chen, D.; Lin, C.; Zheng, H.; Guo, S. Characterizing urbanization-induced land surface phenology change from time-series remotely sensed images at fine spatio-temporal scale: A case study in Nanjing, China (2001–2018). J. Clean Prod. 2020, 274, 122487. [Google Scholar] [CrossRef]
  24. Li, X.; Zhou, Y.; Asrar, G.R.; Meng, L. Characterizing spatiotemporal dynamics in phenology of urban ecosystems based on Landsat data. Sci. Total Environ. 2017, 605, 721–734. [Google Scholar] [CrossRef]
  25. Fitzjarrald, D.R.; Acevedo, O.C.; Moore, K.E. Climatic consequences of leaf presence in the eastern United States. J. Clim. 2001, 14, 598–614. [Google Scholar] [CrossRef]
  26. Chmura, H.E.; Kharouba, H.M.; Ashander, J.; Ehlman, S.M.; Rivest, E.B.; Yang, L.H. The mechanisms of phenology: The patterns and processes of phenological shifts. Ecol. Monogr. 2019, 89, e01337. [Google Scholar] [CrossRef]
  27. Melaas, E.K.; Friedl, M.A.; Zhu, Z. Detecting interannual variation in deciduous broadleaf forest phenology using Landsat TM/ETM+ data. Remote Sens. Environ. 2013, 132, 176–185. [Google Scholar] [CrossRef]
  28. Liang, L.; Schwartz, M.D.; Fei, S. Validating satellite phenology through intensive ground observation and landscape scaling in a mixed seasonal forest. Remote Sens. Environ. 2011, 115, 143–157. [Google Scholar] [CrossRef]
  29. Jia, W.; Zhao, S.; Zhang, X.; Liu, S.; Henebry, G.M.; Liu, L. Urbanization imprint on land surface phenology: The urban-rural gradient analysis for Chinese cities. Glob. Change Biol. 2021, 27, 2895–2904. [Google Scholar] [CrossRef]
  30. Li, X.; Du, H.; Zhou, G.; Mao, F.; Zhu, D.; Zhang, M.; Xu, Y.; Zhou, L.; Huang, Z. Spatiotemporal patterns of remotely sensed phenology and their response to climate change and topography in subtropical bamboo forests during 2001–2017: A case study in Zhejiang Province, China. GISci. Remote Sens. 2023, 60, 2163575. [Google Scholar] [CrossRef]
  31. Yuan, M.; Wang, L.; Lin, A.; Liu, Z.; Li, Q.; Qu, S. Vegetation green up under the influence of daily minimum temperature and urbanization in the Yellow River Basin, China. Ecol. Indic. 2020, 108, 105760. [Google Scholar] [CrossRef]
  32. Vitasse, Y.; Signarbieux, C.; Fu, Y.H. Global warming leads to more uniform spring phenology across elevations. Proc. Natl. Acad. Sci. USA 2018, 115, 1004–1008. [Google Scholar] [CrossRef] [PubMed]
  33. Wu, X.; Niu, C.; Liu, X.; Hu, T.; Feng, Y.; Zhao, Y.; Liu, S.; Liu, Z.; Dai, G.; Zhang, Y.; et al. Canopy structure regulates autumn phenology by mediating the microclimate in temperate forests. Nat. Clim. Change 2024, 14, 1299–1305. [Google Scholar] [CrossRef]
  34. Jucker, T.; Hardwick, S.; Both, S.; Elias, D.; Ewers, R.; Milodowski, D.; Swinfield, T.; Coomes, D. Canopy structure and topography jointly constrain the microclimate of human-modified tropical landscapes. Glob. Change Biol. 2018, 24, 5243–5258. [Google Scholar] [CrossRef]
  35. Smith-Tripp, S.M.; Eskelson, B.N.I.; Coops, N.C.; Schwartz, N.B. Canopy height impacts on the growing season and monthly microclimate in a burned forest of British Columbia, Canada. Agric. For. Meteorol. 2022, 323, 109067. [Google Scholar] [CrossRef]
  36. De Frenne, P.; Lenoir, J.; Luoto, M.; Scheffers, B.R.; Zellweger, F.; Aalto, J.; Ashcroft, M.B.; Christiansen, D.M.; Decocq, G.; De Pauw, K.; et al. Forest microclimates and climate change: Importance, drivers and future research agenda. Glob. Change Biol. 2021, 27, 2279–2297. [Google Scholar] [CrossRef]
  37. Osada, N.; Hiura, T. Intraspecific differences in spring leaf phenology in relation to tree size in temperate deciduous trees. Tree Physiol. 2019, 39, 782–791. [Google Scholar] [CrossRef]
  38. Chen, Y.; Yang, H.; Yang, Z.; Yang, Q.; Liu, W.; Huang, G.; Ren, Y.; Cheng, K.; Xiang, T.; Chen, M.; et al. Enhancing high-resolution forest stand mean height mapping in China through an individual tree-based approach with close-range lidar data. Earth Syst. Sci. Data 2024, 16, 5267–5285. [Google Scholar] [CrossRef]
  39. Lorer, E.; Verheyen, K.; Blondeel, H.; De Pauw, K.; Sanczuk, P.; De Frenne, P.; Landuyt, D. Forest understorey flowering phenology responses to experimental warming and illumination. New Phytol. 2024, 241, 1476–1491. [Google Scholar] [CrossRef]
  40. Keenan, T.; Gray, J.; Friedl, M.; Toomey, M.; Bohrer, G.; Hollinger, D.; Munger, J.; O’Keefe, J.; Schmid, H.; Suewing, I.; et al. Net carbon uptake has increased through warming-induced changes in temperate forest phenology. Nat. Clim. Change 2014, 4, 598–604. [Google Scholar] [CrossRef]
  41. Osada, N. Differential springtime branch warming controls intra-crown nitrogen allocation and leaf photosynthetic traits in understory saplings of a temperate deciduous species. Oecologia 2021, 196, 331–340. [Google Scholar] [CrossRef] [PubMed]
  42. Balzarolo, M.; Valdameri, N.; Fu, Y.; Schepers, L.; Janssens, I.; Campioli, M. Different determinants of radiation use efficiency in cold and temperate forests. Glob. Ecol. Biogeogr. 2019, 28, 1649–1667. [Google Scholar] [CrossRef]
  43. Garbulsky, M.F.; Peñuelas, J.; Papale, D.; Ardö, J.; Goulden, M.L.; Kiely, G.; Richardson, A.D.; Rotenberg, E.; Veenendaal, E.M.; Filella, I. Patterns and controls of the variability of radiation use efficiency and primary productivity across terrestrial eco-systems. Glob. Ecol. Biogeogr. 2010, 19, 253–267. [Google Scholar] [CrossRef]
  44. Chasmer, L.; McCaughey, H.; Barr, A.; Black, A.; Shashkov, A.; Treitz, P.; Zha, T. Investigating light-use efficiency across a jack pine chronosequence during dry and wet years. Tree Physiol. 2008, 28, 1395–1406. [Google Scholar] [CrossRef]
  45. Mäkelä, A.; Pulkkinen, M.; Kolari, P.; Lagergren, F.; Berbigier, P.; Lindroth, A.; Loustau, D.; Nikinmaa, E.; Vesala, T.; Hari, P. Developing an empirical model of stand GPP with the LUE approach: Analysis of eddy covariance data at five contrasting conifer sites in Europe. Glob. Change Biol. 2008, 14, 92–108. [Google Scholar] [CrossRef]
  46. Inner Mongolia Statistical Bureau. Inner Mongolia Statistical Yearbook 2020; Inner Mongolia Statistical Bureau: Hohhot, China, 2020. [Google Scholar]
  47. Ganguly, S.; Friedl, M.A.; Tan, B.; Zhang, X.; Verma, M. Land surface phenology from MODIS: Characterization of the Collection 5 global land cover dynamics product. Remote Sens. Environ. 2010, 114, 1805–1816. [Google Scholar] [CrossRef]
  48. Ren, Q.; He, C.; Huang, Q.; Zhou, Y. Urbanization impacts on vegetation phenology in China. Remote Sens. 2018, 10, 1905. [Google Scholar] [CrossRef]
  49. Yin, P.; Li, X.; Mao, J.; Johnson, B.A.; Wang, B.; Huang, J. A comprehensive analysis of the crop effect on the urban-rural differences in land surface phenology. Sci. Total Environ. 2023, 861, 160604. [Google Scholar] [CrossRef]
  50. Meng, L.; Zhou, Y.; Roman, M.O.; Stokes, E.C.; Wang, Z.; Asrar, G.R.; Mao, J.; Richardson, A.D.; Gu, L.; Wang, Y. Artificial light at night: An underappreciated effect on phenology of deciduous woody plants. PNAS Nexus 2022, 1, pgac046. [Google Scholar] [CrossRef]
  51. Wang, L.; De Boeck, H.J.; Chen, L.; Song, C.; Chen, Z.; McNulty, S.; Zhang, Z. Urban warming increases the temperature sensitivity of spring vegetation phenology at 292 cities across China. Sci. Total Environ. 2022, 834, 155154. [Google Scholar] [CrossRef]
  52. Anselin, L. Local indicators of spatial association-LISA. Geogr. Anal. 1995, 27, 93–115. [Google Scholar] [CrossRef]
  53. Liu, B.; Tian, Y.; Guo, M.; Tran, D.; Alwah, A.; Xu, D. Evaluating the disparity between supply and demand of park green space using a multi-dimensional spatial equity evaluation framework. Cities 2022, 121, 103484. [Google Scholar] [CrossRef]
  54. Li, H.; Wang, Q.; Shi, W.; Deng, Z.; Wang, H. Residential clustering and spatial access to public services in Shanghai. Habitat Int. 2015, 46, 119–129. [Google Scholar] [CrossRef]
  55. Wang, J.F.; Hu, Y. Environmental health risk detection with GeogDetector. Environ. Modell. Softw. 2012, 33, 114–115. [Google Scholar] [CrossRef]
  56. Wang, J.F.; Xu, C.D. Geodetector: Principle and prospective. Acta Geogr. Sin. 2017, 72, 116–134. [Google Scholar]
  57. Yin, J.; Wu, X.; Shen, M.; Zhang, X.; Zhu, C.; Xiang, H.; Shi, C.; Guo, Z.; Li, C. Impact of urban greenspace spatial pattern on land surface temperature: A case study in Beijing metropolitan area, China. Landsc. Ecol. 2019, 34, 2949–2961. [Google Scholar] [CrossRef]
  58. Xue, Z.; Meng, X.; Liu, B. Spatiotemporal evolution and driving factors of ecosystem services in the upper Fenhe watershed, China. Ecol. Indic. 2024, 160, 111803. [Google Scholar] [CrossRef]
  59. Yang, J.; Luo, X.; Jin, C.; Xiao, X.; Xia, J. Spatiotemporal patterns of vegetation phenology along the urban–rural gradient in Coastal Dalian, China. Urban For. Urban Green. 2020, 54, 126784. [Google Scholar] [CrossRef]
  60. Klosterman, S.; Melaas, E.; Wang, J.; Martinez, A.; Frederick, S.; O’Keefe, J.; Orwig, D.A.; Wang, Z.; Sun, Q.; Schaaf, C.; et al. Fine-scale perspectives on landscape phenology from unmanned aerial vehicle (UAV) photography. Agric. For. Meteorol. 2018, 248, 397–407. [Google Scholar] [CrossRef]
  61. Misra, G.; Buras, A.; Heurich, M.; Asam, S.; Menzel, A. LiDAR derived topography and forest stand characteristics largely explain the spatial variability observed in MODIS land surface phenology. Remote Sens. Environ. 2018, 218, 231–244. [Google Scholar] [CrossRef]
  62. Xie, Y.; Ahmed, K.F.; Allen, J.M.; Wilson, A.M.; Silander, J.A. Green-up of deciduous forest communities of northeastern North America in response to climate variation and climate change. Landsc. Ecol. 2015, 30, 109–123. [Google Scholar] [CrossRef]
  63. Zhang, F.; Liu, B.; Henderson, M.; Shen, X.; Su, Y.; Zhou, W. Changing spring phenology of northeast China forests during rapid warming and short-term slowdown periods. Forests 2022, 13, 2173. [Google Scholar] [CrossRef]
  64. Elmore, A.J.; Guinn, S.M.; Minsley, B.J.; Richardson, A.D. Landscape controls on the timing of spring, autumn, and growing season length in mid-Atlantic forests. Glob. Change Biol. 2012, 18, 656–674. [Google Scholar] [CrossRef]
  65. Tian, R.; Liu, L.; Zheng, J.; Li, J.; Han, W.; Liu, Y. Combined Effects of Meteorological Factors, Terrain, and Greenhouse Gases on Vegetation Phenology in Arid Areas of Central Asia from 1982 to 2021. Land 2024, 13, 180. [Google Scholar] [CrossRef]
  66. Hua, X.B.; Ohlemüller, R.; Sirguey, P. Differential effects of topography on the timing of the growing season in mountainous grassland ecosystems. Environ. Adv. 2022, 8, 100234. [Google Scholar] [CrossRef]
  67. An, S.; Zhang, X.; Chen, X.; Yan, D.; Henebry, G. An Exploration of Terrain Effects on Land Surface Phenology across the Qinghai–Tibet Plateau Using Landsat ETM+ and OLI Data. Remote Sens. 2018, 10, 1069. [Google Scholar] [CrossRef]
  68. Li, R.; Zhao, L.; Wu, T.; Ding, Y.; Xin, Y.; Zou, D.; Xiao, Y.; Jiao, Y.; Qin, Y.; Sun, L. Temporal and spatial variations of global solar radiation over the Qinghai-Tibetan Plateau during the past 40 years. Theor. Appl. Climatol. 2013, 113, 573–583. [Google Scholar] [CrossRef]
  69. Li, D.; Stucky, B.J.; Baiser, B.; Guralnick, R. Urbanization delays plant leaf senescence and extends growing season length in cold but not in warm areas of the Northern Hemisphere. Glob. Ecol. Biogeogr. 2022, 31, 308–320. [Google Scholar] [CrossRef]
  70. Desta, F.; Colbert, J.J.; Rentch, J.S.; Gottschalk, K.W. Aspect induced differences in vegetation, soil, and microclimatic characteristics of an Appalachian watershed. Castanea 2004, 69, 92–108. [Google Scholar] [CrossRef]
  71. Radcliffe, D.C.; Matthews, S.N.; Hix, D.M. Beyond oak regeneration: Modeling mesophytic sapling density drivers along topographic, edaphic, and stand-structural gradients in mature oak-dominated forests of Appalachian Ohio. Can. J. For. Res. 2020, 50, 1215–1227. [Google Scholar] [CrossRef]
  72. Ding, C.; Li, Y.; Xie, Q.; Li, H.; Zhang, B. Impacts of terrain on land surface phenology derived from Harmonized Landsat 8 and Sentinel-2 in the Tianshan Mountains, China. GISci. Remote Sens. 2023, 60, 2242621. [Google Scholar] [CrossRef]
  73. Vitasse, Y.; Baumgarten, F.; Zohner, C.M.; Kaewthongrach, R.; Fu, Y.H.; Walde, M.G.; Moser, B. Impact of microclimatic conditions and resource availability on spring and autumn phenology of temperate tree seedlings. New Phytol. 2021, 232, 537–550. [Google Scholar] [CrossRef] [PubMed]
  74. Fan, X.; Li, W.; Wu, X.; Yao, M.; Niu, F.; Lin, Z. Heterogeneity of Surface heat exchange of slopes and potential drivers of the initiation of thaw slump, Qinghai-Tibet Plateau. Int. J. Disaster Risk Sci. 2023, 14, 549–565. [Google Scholar] [CrossRef]
  75. de Souza, A.C.P.; da Costa, R.C. Differences in leaf phenology between juvenile and adult individuals of two tree species in a seasonally dry tropical woodland. Aust. Ecol. 2020, 45, 240–248. [Google Scholar] [CrossRef]
  76. Vinod, N.; Slot, M.; McGregor, I.R.; Ordway, E.M.; Smith, M.N.; Taylor, T.C.; Sack, L.; Buckley, T.N.; Anderson-Teixeira, K.J. Thermal sensitivity across forest vertical profiles: Patterns, mechanisms, and ecological implications. New Phytol. 2023, 237, 22–47. [Google Scholar] [CrossRef]
  77. Timmers, R.; van Kuijk, M.; Verweij, P.A.; Ghazoul, J.; Hautier, Y.; Laurance, W.F.; Arriaga-Weiss, S.L.; Askins, R.A.; Battisti, C.; Berg, Å.; et al. Conservation of birds in fragmented landscapes requires protected areas. Front. Ecol. Environ. 2022, 20, 361–369. [Google Scholar] [CrossRef]
Forests 16 00490 g001
Figure 1. The study area, (a) location and forest height of Hulun Buir in China; (b) the elevation of Hulun Buir; (c) the slope of Hulun Buir; and (d) the aspect of Hulun Buir.
Figure 1. The study area, (a) location and forest height of Hulun Buir in China; (b) the elevation of Hulun Buir; (c) the slope of Hulun Buir; and (d) the aspect of Hulun Buir.
Forests 16 00490 g001
Forests 16 00490 g002
Figure 2. Scatter plots of SOS with pre-season temperature (a) and precipitation (c). LISA maps of SOS with pre-season temperature (b) and precipitation (d).
Figure 2. Scatter plots of SOS with pre-season temperature (a) and precipitation (c). LISA maps of SOS with pre-season temperature (b) and precipitation (d).
Forests 16 00490 g002
Forests 16 00490 g003
Figure 3. Scatter plots of EOS with autumn temperature (a) and precipitation (c). LISA maps of EOS with autumn temperature (b) and precipitation (d).
Figure 3. Scatter plots of EOS with autumn temperature (a) and precipitation (c). LISA maps of EOS with autumn temperature (b) and precipitation (d).
Forests 16 00490 g003
Forests 16 00490 g004
Figure 4. Scatter plots of SOS with elevation (a) and slope (c). LISA maps of SOS with elevation (b) and slope (d).
Figure 4. Scatter plots of SOS with elevation (a) and slope (c). LISA maps of SOS with elevation (b) and slope (d).
Forests 16 00490 g004
Forests 16 00490 g005
Figure 5. Scatter plots of EOS with elevation (a) and slope (c). LISA maps of EOS with elevation (b) and slope (d).
Figure 5. Scatter plots of EOS with elevation (a) and slope (c). LISA maps of EOS with elevation (b) and slope (d).
Forests 16 00490 g005
Forests 16 00490 g006
Figure 6. Box plots of SOS (a) and EOS (b) in different aspects. The green dashed lines and numbers in each box represent mean values, the orange lines represent median values, and the top and bottom edges of each box represent 75th and 25th percentiles, respectively.
Figure 6. Box plots of SOS (a) and EOS (b) in different aspects. The green dashed lines and numbers in each box represent mean values, the orange lines represent median values, and the top and bottom edges of each box represent 75th and 25th percentiles, respectively.
Forests 16 00490 g006
Forests 16 00490 g007
Figure 7. Scatter plot of SOS with forest height (a). LISA map of SOS with forest height (b).
Figure 7. Scatter plot of SOS with forest height (a). LISA map of SOS with forest height (b).
Forests 16 00490 g007
Forests 16 00490 g008
Figure 8. Scatter plot of EOS with forest height (a). LISA map of EOS with forest height (b).
Figure 8. Scatter plot of EOS with forest height (a). LISA map of EOS with forest height (b).
Forests 16 00490 g008
Forests 16 00490 g009
Figure 9. Scatter plots of elevation and slop with pre-season temperature (a,b) and pre-season precipitation (c,d). Scatter plots without fitted curves indicated a non-significant relationship.
Figure 9. Scatter plots of elevation and slop with pre-season temperature (a,b) and pre-season precipitation (c,d). Scatter plots without fitted curves indicated a non-significant relationship.
Forests 16 00490 g009
Forests 16 00490 g010
Figure 10. Scatter plots of elevation and slope with autumn temperature (a,b) and autumn precipitation (c,d). Scatter plots without fitted curves indicated a non-significant relationship.
Figure 10. Scatter plots of elevation and slope with autumn temperature (a,b) and autumn precipitation (c,d). Scatter plots without fitted curves indicated a non-significant relationship.
Forests 16 00490 g010
Table 1. Five possible types of interactions in the interaction detector.
Table 1. Five possible types of interactions in the interaction detector.
NO.Mean Forest SOS (DOY)Mean Forest EOS (DOY)
1q (X1 ∩ X2) < Min (q (X1), q (X2))The factors nonlinearly weaken each other.
2Min (q (X1), q (X2)) < q (X1 ∩ X2) < Max (q (X1), q (X2))One factor nonlinearly weakens the other.
3q (X1 ∩ X2) > Max (q (X1), q (X2))The factors bilaterally enhance each other.
4q (X1 ∩ X2) = q (X1) + q (X2)The factors are independent of each other.
5q (X1 ∩ X2) > q (X1) + q (X2)The factors nonlinearly enhance each other.
Table 2. The average forest phenology for each aspect type across the entire study area.
Table 2. The average forest phenology for each aspect type across the entire study area.
Aspect TypeMean Forest SOS (DOY)Mean Forest EOS (DOY)
North117.99276.85
Northeast117.76277.40
East117.65278.42
Southeast117.86279.00
South118.24279.25
Southwest118.43278.98
West118.46278.17
Northwest118.34277.15
Table 3. Mean temperature and precipitation values across two seasons for various slope aspects.
Table 3. Mean temperature and precipitation values across two seasons for various slope aspects.
AspectMean Pre-Season Temperature (°C)Mean Pre-Season Precipitation (mm)Mean Autumn Temperature (°C)Mean Autumn Precipitation (mm)
North−6.218534.790.3933336.51
Northeast−6.140562.500.4103403.40
East−6.148569.490.3873411.87
Southeast−6.200571.830.3383427.94
South−6.199565.760.3543424.64
Southwest−6.202551.190.3753389.15
West−6.224525.080.4033319.87
Northwest−6.267520.500.3843301.01
Table 4. Parameters of multiple linear regression between forest phenology and different independent variables.
Table 4. Parameters of multiple linear regression between forest phenology and different independent variables.
Independent VariableβpVIF
Forest SOS: R2 = 0.329
Intercept118.371<0.001–
Temperature0.389<0.0011.796
Precipitation0.060<0.0011.305
Elevation0.009<0.0012.024
Slope−1.484<0.0011.317
Forest Height−0.113=0.0621.548
Forest EOS: R2 = 0.315
Intercept282.275<0.001–
Temperature0.621<0.0011.532
Precipitation−0.010<0.0011.088
Elevation−0.007<0.0011.815
Slope0.559<0.0011.254
Forest Height0.196<0.0011.368
Table 5. Results of the interaction between climatic, topographical, and forest height factors on forest phenology.
Table 5. Results of the interaction between climatic, topographical, and forest height factors on forest phenology.
Interactive VariablesqInteractive Type
Forest SOS
Temperature ∩ Precipitation0.350EnhanceB
Temperature ∩ Elevation0.435N
Temperature ∩ Slope0.408B
Temperature ∩ Forest Height0.307N
Precipitation ∩ Elevation0.258N
Precipitation ∩ Slope0.302B
Precipitation ∩ Forest Height0.174N
Elevation ∩ Slope0.323N
Elevation ∩ Forest Height0.175N
Slope ∩ Forest Height0.211B
Forest EOS
Temperature ∩ Precipitation0.366EnhanceN
Temperature ∩ Elevation0.366B
Temperature ∩ Slope0.372N
Temperature ∩ Forest Height0.349B
Precipitation ∩ Elevation0.304N
Precipitation ∩ Slope0.078N
Precipitation ∩ Forest Height0.143N
Elevation ∩ Slope0.288N
Elevation ∩ Forest Height0.251B
Slope ∩ Forest Height0.118N
“B” denoted two interactive variables bilaterally enhance each other, and “N” denoted nonlinear enhancement of two interactive variables.
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Tian, Y.; Wang, L.; Liu, B.; Yao, Y.; Xu, D. Phenological Spatial Divergences Promoted by Climate, Terrain, and Forest Height in a Cold Temperate Forest Landscape: A Case Study of the Greater Khingan Mountain in Hulun Buir, China. Forests 2025, 16, 490. https://doi.org/10.3390/f16030490

AMA Style

Tian Y, Wang L, Liu B, Yao Y, Xu D. Phenological Spatial Divergences Promoted by Climate, Terrain, and Forest Height in a Cold Temperate Forest Landscape: A Case Study of the Greater Khingan Mountain in Hulun Buir, China. Forests. 2025; 16(3):490. https://doi.org/10.3390/f16030490

Chicago/Turabian Style

Tian, Yu, Lei Wang, Bingxi Liu, Yunlong Yao, and Dawei Xu. 2025. "Phenological Spatial Divergences Promoted by Climate, Terrain, and Forest Height in a Cold Temperate Forest Landscape: A Case Study of the Greater Khingan Mountain in Hulun Buir, China" Forests 16, no. 3: 490. https://doi.org/10.3390/f16030490

APA Style

Tian, Y., Wang, L., Liu, B., Yao, Y., & Xu, D. (2025). Phenological Spatial Divergences Promoted by Climate, Terrain, and Forest Height in a Cold Temperate Forest Landscape: A Case Study of the Greater Khingan Mountain in Hulun Buir, China. Forests, 16(3), 490. https://doi.org/10.3390/f16030490

Note that from the first issue of 2016, this journal uses article numbers instead of page numbers. See further details here.

Article Metrics

Back to TopTop