Next Article in Journal
Genotypic Variability in Growth and Leaf-Level Physiological Performance of Highly Improved Genotypes of Pinus radiata D. Don Across Different Sites in Central Chile
Previous Article in Journal
A System Dynamics Approach to Resilience Analysis in the Sino-Russian Timber Supply Chain
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Vegetation Growth Carryover and Lagged Climatic Effect at Different Scales: From Tree Rings to the Early Xylem Growth Season

1
School of Geography and Tourism, Xinjiang Normal University, Urumqi 830054, China
2
Institute of Desert Meteorology, China Meteorological Administration, Urumqi 830002, China
3
Key Laboratory of Tree-Ring Physical and Chemical Research, China Meteorological Administration, Urumqi 830002, China
4
College of Geography and Remote Sensing Sciences, Xinjiang University, Urumqi 830046, China
*
Authors to whom correspondence should be addressed.
Forests 2025, 16(7), 1107; https://doi.org/10.3390/f16071107
Submission received: 25 May 2025 / Revised: 21 June 2025 / Accepted: 1 July 2025 / Published: 4 July 2025
(This article belongs to the Special Issue Tree-Ring Analysis: Response and Adaptation to Climate Change)

Abstract

Vegetation growth is influenced not only by current climatic conditions but also by growth-enhancing signals and preceding climate factors. Taking the dominant species, Juniperus seravschanica Kom, in Tajikistan as the research subject, this study combines tree-ring width data with early xylem growth season data (from the start of xylem growth to the first day of the NDVI peak month), simulated using the Vaganov–Shashkin (V-S) model, a process-based tree-ring growth model. This study aims to explore the effects of vegetation growth carryover (VGC) and lagged climatic effects (LCE) on tree rings and the early xylem growth season at two different scales by integrating tree-ring width data and xylem phenology simulations. A vector autoregression (VAR) model was employed to analyze the response intensity and duration of VGC and LCE. The results show that the VGC response intensity in the early xylem growth season is higher than that of tree-ring width. The LCE duration for both the early xylem growth season and tree-ring width ranges from 0 to 11 (years or seasons), with peak LCE response intensity observed at a lag of 2–3 (years or seasons). The persistence of the climate lag effect on vegetation growth has been underestimated, supporting the use of a lag of 0–3 (years or seasons) to study the long-term impacts of climate. The influence of VGC on vegetation growth is significantly stronger than that of LCEs; ultimately indicating that J. seravschanica adapts to harsh environments by modulating its growth strategy through VGC and LCE. Investigating the VGC and LCE of multi-scale xylem growth indicators enhances our understanding of forest ecosystem dynamics.

1. Introduction

Vegetation plays a pivotal mediating role in the carbon cycle, water cycle, and energy balance of terrestrial ecosystems through its growth and dynamic processes [1,2,3]. Meanwhile, climate change has profound impacts on vegetation growth, terrestrial ecosystem balance, and the feedback mechanisms between vegetation and climate [4]. Notably, these complex feedback mechanisms are significantly regulated by the lagged effects of vegetation responses to climate change, highlighting the dynamic nature and complexity of vegetation–climate interactions [5,6]. Lagged effects provide a critical perspective for understanding the long-term impacts of climate change on ecosystems.
The impact of climate change on vegetation growth is not always synchronous: vegetation growth is not only directly influenced by current climatic conditions but may also require a certain time to respond to previous climatic changes, a phenomenon referred to as the lagged climatic effect (LCE) [7,8]. However, most studies on the effects of climate on vegetation growth have primarily focused on the immediate responses of vegetation to concurrent climatic factors, without considering the temporal lag effects of climate variables. In reality, vegetation responses to climate factors are often complex and time-sensitive [5]. Extensive records of the lagged effects of climate on tree growth at the stand level in temperate forests have been documented [9,10]. For instance, tree-ring width is significantly influenced by spring snowmelt, which may affect soil moisture supply during the early and peak vegetation growth periods [11]. Vegetation responses to diurnal asymmetry globally exhibit a delay of approximately 12 months [8]. These findings suggest that vegetation dynamics are influenced not only by current climatic conditions but also by antecedent ones, a consideration often overlooked in studies examining the impacts of climate change on vegetation dynamics.
Tree-ring width is closely related to tree growth and serves as a significant contributor to forest ecosystems, with growth variations being strongly influenced by climate [12]. Investigating tree-ring width helps understand the impacts of past climate conditions on vegetation growth [13]. However, recent studies have indicated a declining sensitivity of tree-ring width to temperature. Instead, intra-annual tree growth may be constrained by other factors, such as xylem phenology, frost events, and dust storms [14,15,16]. Additionally, carryover effects between tree-ring years may influence tree growth. The current enhanced signal of vegetation growth will affect the subsequent growth of vegetation, a phenomenon termed vegetation growth carryover (VGC) [17,18]. After considering the carryover effect, the relationship between climate and tree-ring width was improved, and it was found that the annual growth of trees is affected by carbon absorption in the second half of the previous growing season [19]. Similarly, the vegetation growth in the Northern Hemisphere is dominated by the carryover effect of enhanced growth signals from the previous season, and this effect exceeds that of climatic factors [17]. Tree-ring data not only reflect intra-annual climate conditions but also carry “memory” of prior climate conditions [20,21]. However, studying the seasonal climate–vegetation dynamics of xylem growth at an intra-annual scale is challenging. In remote and climatically harsh forest regions, long-term and consistent monitoring demands significant human and material resources, hindering efforts to quantify xylem growth responses to long-term climate changes at regional and global scales. The Vaganov–Shashkin (V-S) growth model, based on physiological and ecological processes [22,23], can simulate seasonal variations in cambial activity and xylem differentiation, enabling the analysis of long-term and multi-scale changes in xylem differentiation [22,24,25]. In this study, we combined spring phenological timing derived from the V-S model with the month of peak Normalized Difference Vegetation Index (NDVI) to determine the early xylem growth phase. This approach enabled us to investigate xylem growth dynamics at a seasonal scale. In comparison, studies of radial growth dynamics indicate that current vegetation conditions can influence subsequent growth across different growing seasons (Kang et al.). Given these variations [26], it is crucial to examine the similarities and differences in vegetation response processes to past growth rates and climatic changes across different observational scales. Although the concept of VGC has been recognized for some time, our understanding of its influence on seasonal xylem growth remains limited. Specifically, substantial knowledge gaps remain regarding how VGC interacts with lagged climatic effects (LCE) to affect xylem growth and how these interconnected factors collectively shape vegetation’s long-term responses to environmental changes.
Tajikistan, characterized by its complex terrain and distance from the oceans, exhibits a high sensitivity of tree-ring width to climate change [27]. However, studies on the carryover effects of climate on vegetation growth and dynamic lag effects in this region remain scarce. Juniperus seravschanica Kom, a key component of mountainous forests in Tajikistan, is widely distributed across the northern Khujand mountain area, the central western Alay mountain area, and the southern marginal Pamirs Plateau. Studies have confirmed that this species is highly sensitive to climate effects [15,28,29]. To better understand the dynamics of Central Asian forests and their interactions with climate, vegetation growth carryover (VGC), and lagged climatic effects (LCE), this study offers valuable perspectives for exploring vegetation growth patterns in the region. The primary objectives of this study are as follows:
(i)
Investigate how the vegetation growth carryover (VGC) under two scales of tree-ring width (TRW) and the early xylem growth season (EXGs) can enhance signal acquisition and how it affects the growth in the next stage.
(ii)
Explore the persistence status of lagged climatic effects (LCE) on wood growth under the two scales of tree-ring width (TRW) and the early xylem growth season (EXGs).
(iii)
Assess the contribution rates of the wood growth persistence effect and the climate lag effect to vegetation growth under the two scales of tree-ring width (TRW) and the early xylem growth season (EXGs).
To address these objectives, J. seravschanica was selected as the focal species. Tree-ring samples were collected from three sites in Tajikistan to analyze the relationships between VGC, LCE, and xylem growth.

2. Materials

2.1. Study Site

The study area is located in Tajikistan, which covers an area of 14.31 × 104 km2, extending approximately 700 km from east to west and 350 km from north to south. Tajikistan shares its northern borders with Uzbekistan and Kyrgyzstan, its southern border with Afghanistan, and its eastern border with China. The country is traversed by three major rivers: the Syr Darya, the Zeravshan, and the Amu Darya, primarily fed by glacial meltwater and snow accumulation [30]. Situated in Central Asia, Tajikistan is predominantly mountainous, characterized by steep peaks and deep valleys, with its southern region hosting the Pamirs Plateau—the world’s highest average-altitude plateau [31]. The climate has the characteristics of rain and heat occurring at different times, with cold and wet winters and hot and dry summers [32].
In this study, tree-ring samples of J. seravschanica were collected from three regions in Tajikistan: the northern Khujand mountain area (KZB), the central western Alay mountain area (ZTW), and the southern marginal Pamirs Plateau mountain area (DWZ) (Figure 1). The geographic details of the sampling sites are as follows: KZB at an elevation of 1575.7 m (69.612838° E, 40.646455° N), ZTW at 2147.1 m (67.510412° E, 39.3531° N), and DWZ at 2073 m (70.857156° E, 38.586636° N). These sites are characterized by complex mountainous terrain with high elevations and steep slopes. The vegetation is relatively homogeneous, dominated by J. seravschanica. These sampling locations, situated across different elevations and climatic conditions, provide valuable support for investigating the influence of regional climatic and geographic factors on tree-ring width.

2.2. Datasets

2.2.1. Climate Data

The ERA5-Land dataset, released by the European Centre for Medium-Range Weather Forecasts (ECMWF) (https://cds.climate.copernicus.eu, accessed on 10 March 2025), is a high-resolution dataset developed as the next generation of reanalysis data, based on the land component of the ERA5 reanalysis product. This dataset covers the period from 1950 to the present, with a spatial resolution of 0.1° × 0.1° and a temporal resolution of daily intervals. In this study, we extracted daily maximum temperature (MaxTEM), daily minimum temperature (MinTEM), daily mean temperature (TEM), and daily precipitation (PRE) from 1982 to 2022 for the sampling site coordinates (Figure 2). These meteorological data were used to analyze the response of J. seravschanica to climate change at the three sampling sites over the period 1982–2022.
The annual mean precipitation at the sampling sites (KZB, ZTW, and DWZ) is approximately 1.63 mm, 2.01 mm, and 3.58 mm, respectively, with peak precipitation occurring in March at 3.15 mm, 3.93 mm, and 5.65 mm, respectively. The annual mean temperatures for these sites are approximately 12.62 °C, 7.50 °C, and 1.31 °C, with high temperatures concentrated from June to August. The average temperature peaks in July, reaching 25.28 °C, 19.10 °C, and 13.86 °C at KZB, ZTW, and DWZ, respectively. Meteorological data from the three sampling sites indicate that summer precipitation is extremely scarce, with the majority of precipitation concentrated in spring and winter. These seasons also experience lower temperatures, while summer is characterized by hot and dry conditions (Figure S1), consistent with previous research findings [32,33,34].

2.2.2. Tree-Ring Material

In August 2023, cores from Juniperus seravschanica Kom were collected in Tajikistan. To minimize interference from non-climatic factors on tree growth, sampling focused on juniper individuals unaffected by disturbances such as fire, earthquakes, animal activity, or human intervention. For each tree, the first core was collected at breast height along the slope aspect KZB (S), ZTW (N), and DWZ (WS) using a 10 mm-inner-diameter increment borer, and then the second core was collected from a direction perpendicular to the first one. At least 20 trees (over 40 cores) were sampled at each site. The cores were sealed in paper tubes labeled with site information, sampling date, and core ID, then transported to the laboratory for processing.
In the laboratory, the cores from the three sampling sites were air-dried, mounted, and polished. Ring widths were visually cross-dated under a microscope and measured using a LINTAB6 tree-ring measuring system.
The COFECHA program was used to perform cross-dating and quality checks on tree-ring width. The ARSTAN program, developed by the International Tree-Ring Data Bank, was then applied to generate a chronology from the cross-dated tree-ring width data [35]. Negative exponential functions were fitted to the tree-ring data to remove growth- and age-related non-climatic trends (Table 1). Detrended series were then averaged using a bi-weight robust mean method to produce standardized chronologies that preserve low-frequency signals.
A standard deviation (SD) and mean sensitivity (MS) greater than 0.193 indicate a strong climatic influence on vegetation growth. The sparse distribution and low canopy density of J. seravschanica result in low values for the mean correlation coefficient among all series (R1) and the mean correlation coefficient between trees (R2) (Pearson correlation analysis). Sample depth represents the number of samples available for each year. According to the subsample signal (SSS), when the value in the early chronology exceeds the threshold of 0.85, this period can be used to represent the climatic signal of the sampling site. The effective time periods of the chronologies were determined as KZB (1896–2023), ZTW (1846–2023), and DWZ (1929–2023) (Table 1 and Figure 3). The signal-to-noise ratio (SNR) of ZTW was higher than that of KZB and DWZ, indicating that ZTW contains more climatic information. The regional representativeness of the tree-ring chronologies was evaluated using the expressed population signal (EPS), where an EPS value > 0.85 indicates reliable regional representativeness. The EPS values of all three sampling sites exceeded this threshold (Table 1), confirming the reliability of our chronology data. In this study, GIMMS remote sensing phenology data were used to validate the results simulated by the V-S model. To ensure the reliability of the V-S model data, the tree-ring data were truncated to the period of 1982–2022, so as to align the time with the remote sensing data.

2.3. Remote Sensing Data

2.3.1. GIMMS NDVI

The NDVI data for KZB and ZTW were derived from the PKU GIMMS NDVI dataset (version 1.2), which provides globally consistent biweekly NDVI data at a 1/12° spatial resolution for the period 1982–2022. This dataset is designed to mitigate uncertainties caused by NOAA satellite orbital drift and AVHRR sensor degradation in long-term NDVI observations. Based on the GIMMS NDVI3g product, the dataset was generated using a biome-specific back-propagation neural network (BPNN) model trained on 3.6 million high-quality global samples. To extend the dataset to 2022, a pixel-level random forest fusion method was applied to integrate MODIS NDVI (MOD13C1) data, significantly enhancing temporal consistency and eliminating satellite system biases. The dataset demonstrates strong consistency with MODIS NDVI in both pixel values and global vegetation trends, providing critical foundational support for global change research (https://gee-community-catalog.org/projects/gimms_ndvi/, accessed on 25 March 2025).
For DWZ, NDVI data were sourced from the NASA/GIMMS/3GV0 dataset (1982–2013), as the PKU dataset lacked NDVI values for December–April at the DWZ sampling site. All data were obtained using the Google Earth Engine platform.

2.3.2. Remote Sensing Phenological Data

Remote sensing satellite observations play a crucial role in detecting vegetation canopy phenology [36,37,38]. This study utilized the Northern Hemisphere Vegetation Key Phenology Dataset (1982–2022), which was derived from the GIMMS NDVI 3G+ data. To enhance data quality and temporal smoothing, multiple optimization algorithms, including Savitzky–Golay (SG) filtering, were applied to process the vegetation index (NDVI). Key phenological metrics of the growing season, such as the start of the season (SOS), peak of the season (POS), and end of the season (EOS), were extracted using dynamic-threshold methods and logistic regression. By integrating high-spatiotemporal-resolution remote sensing data with ground observations, this approach enables precise detection of vegetation’s seasonal dynamics, providing a comprehensive understanding of the spatiotemporal patterns of vegetation phenology. The data is sourced from the National Earth System Science Data Center, National Science and Technology Infrastructure of China (http://www.geodata.cn, accessed on 25 March 2025).

3. Method and Analysis

3.1. Simulation of Xylem Phenology and Determination of the Early Xylem Growth Season

3.1.1. Xylem Phenological Simulation

The Vaganov–Shashkin (V-S) model is based on the principle of ecological limiting factors, assuming that tree radial growth is constrained by sunlight, temperature, and soil moisture. The model calculates the relative growth rate of trees using these limiting factors, which drives cambial activity to determine the number (and size) of cells, ultimately converting this into a standardized tree-ring width series [23]. The model comprises two main modules. (1) Environmental module: This module uses daily mean temperature, daily precipitation, and latitude as inputs to calculate daily relative growth rates determined by light, temperature, and soil moisture (derived through a water balance equation). The total relative growth rate is then calculated based on the principle of limiting factors. (2) Cambial activity module: In this module, the total relative growth rate is used to estimate the growth rate of cells in the tree cambium [39].
We iteratively adjusted the simulation parameters for the three sampling sites until the simulation results showed a significant correlation with the observed tree-ring width chronologies. Specifically, based on previous simulation experience, we selected parameters sensitive to the environmental module and soil moisture and defined their respective value ranges. Among these, four parameters of the temperature–growth function were determined: the minimum temperature for growth (T1), the lower (T2) and upper (T3) limits of the optimal temperature, and the maximum temperature for growth (T4). These parameters must satisfy the relationship T1 < T2 < T3 < T4 [25].The correlation coefficients between the simulated results and observed tree-ring width chronologies were calculated, and the parameter set corresponding to the highest correlation coefficient was selected as the optimal configuration. To minimize the impact of simulation initialization, the first simulation year was excluded during the comparison of the simulated and observed data. The detailed parameter settings are provided in Table S1.

3.1.2. Determination of the Early Xylem Growth Season

In this study, the V-S model, based on physiological and ecological processes, was used to simulate tree radial growth at a daily resolution. The V-S model can generate outputs including standardized tree-ring width chronologies, daily growth rates, and key dates related to xylem phenology and tree-ring formation, such as the timing of xylem cell division and maturation. Compared to its performance in simulating the end of the growing season (autumn phenology), the V-S model demonstrates better accuracy in modeling the start of the growing season (spring phenology) [40,41].
The consistency between xylem phenology and remote sensing phenology has been validated using various approaches, including satellite observations [42], field measurements [43], and model simulations [44]. In this study, we conducted a correlation analysis between simulated xylem phenology and remote sensing phenology, which also demonstrated strong consistency. The results are as follows: The simulated start of the vegetation growing season (SOS) shows a significant positive correlation with the GIMMS-derived SOS, with correlation coefficients of 0.63, 0.645, and 0.683 for KZB, ZTW, and DWZ, respectively. This indicates that the V-S model demonstrates strong simulation accuracy for spring xylem phenology at the KZB, ZTW, and DWZ sites (Figure 4).
The determination of the early xylem growth season (EXGs) [45]: The first day of the month with the maximum NDVI value is designated as the onset of the peak growth period (Figure 5a). The EXGs is defined as the period from the start of the growing season to the first day of the month with the maximum NDVI value (Figure 5b). This method effectively integrates physiological indicators with remote sensing data, enabling the quantification of the response of tree growth to environmental changes. The effects of VGC and LCEs on vegetation growth are analyzed using data from two scales: the EXGs and TRW.

3.2. Vector Autoregression Model

The vector autoregression (VAR) model is commonly used to describe the inter-relationships and dynamic processes among multivariate time series. In this study, a time-series model, VAR(q), was applied. This model is suitable for capturing the dynamic patterns among time series and was therefore utilized to analyze the response relationship between climatic factors and vegetation growth [46,47].
The analysis was conducted using STATA 18. The specific workflow was as follows. (1) Stationarity test of the original series: This study employed the augmented Dickey–Fuller (ADF) test to assess stationarity, where a unit root r < 1 indicates a stable model. If the data series was found to be non-stationary, differencing was applied [48]. In this study, all data series passed the ADF stationarity test (Figure S2). (2) Determination of the optimal lag order: The optimal lag length was determined by comparing multiple selection criteria [49]. In this study, the minimum values of four criteria were all less than 1, leading to the selection of a first-order lag for constructing the VAR model (Figure S3). (3) VAR model construction: The model was applied to impulse response analysis and forecast error variance decomposition. First, impulse response analysis was used to investigate the effects of vegetation growth carryover (VGC) and the lagged climatic effect (LCE). Second, variance decomposition was conducted to evaluate the contribution of each structural shock to endogenous variable fluctuations, thereby further assessing the importance of different structural shocks.
In this study, prior to applying the VAR model, it was necessary to perform key indicator analysis, stationarity testing, and Granger causality testing to ensure that the model’s internal parameters met the research requirements [46]. The primary focus of the VAR model evaluation is on the overall stability of the system rather than the significance of individual regression coefficients. Only when the VAR modeling system is stable can impulse response functions and variance decomposition be used to quantify the dynamic responses of vegetation indices to intrinsic and extrinsic random disturbances [50,51].
The structure of the VAR model depends on two key parameters: the number of variables n and the optimal lag order q. The general mathematical expression of the VAR(q) model is as follows:
VAR q = A 0 + q = 1 k A q B C i , t q + q = 1 k E q D C j , t q + α t i = 1 , 2 , , 4 ; j = 1 , 2 , , 5
In this context, B C i and B C t q represent the vectors of endogenous variables at time t and lagged time tq, respectively, encompassing TRW and growth and the early xylem growing season (EXGs).   D C t q denotes the lagged exogenous variables at time tq, including climate factors such as MaxTEM, MinTEM, TEM, and PRE. Endogenous variables are determined by relationships specified within the model and represent outcomes or responses explained by the system. These variables are crucial for understanding the interactions and feedback mechanisms within the studied ecosystem. In contrast, exogenous variables are treated as constants, serving as external inputs that influence the endogenous variables but are not determined by the system’s equations. The coefficients A 0 ,   A q , and E q represent parameters estimated within the model, while α t denotes the error term.

3.2.1. Impulse Response Analysis

The impulse response function represents the impact of a one-time shock to a disturbance term on the current and future values of endogenous variables [52].
IRF i j y = t = 1 k β i j t
The term IRF i j y represents the impulse response of variable i to a shock in variable j after y periods. The element β i j t in the matrix, located at row i and column j, indicates the response of a variable after y periods to a shock in variable j at time t. This can be interpreted as how a shock to variable j (TRW, EXGs, and climate factors) affects the response of variable i (TRW and EXGs) over time.
IRF i j y describes the dynamic response direction of vegetation indicators to an impulse, where positive values indicate a positive response, and negative values indicate a negative response.

3.2.2. Forecast Error Variance Decomposition

Variance decomposition partitions the variations in endogenous variables into components attributable to shocks in the VAR(q) model, providing quantitative information on the relative importance of each random disturbance affecting the variables in the system [53]. It can be interpreted as a method to decompose and quantify the impacts of VGC and LCEs on vegetation growth, identifying the dominant factors and their relative significance in the process.
The formula is as follows [54]:
RVC i j S = q = 0 S 1 β q , i j σ i i j = 1 K q = 0 S 1 ( β q , i j ) 2 σ i i
The variance of β j t can be decomposed into K uncorrelated components. Therefore, to quantify the contribution of each disturbance to the variance of β j t , the relative variance contribution (RVC) is introduced. Under a limited number of S terms, the relative contribution of the variance of variable i, based on shock-induced variance, to the variance of β j t is used as a measure of the influence of variable i on variable j. σ i i represents the variance of variable i. A higher RVC i j S indicates a stronger influence of variable i on variable j, whereas a lower RVC i j S suggests a weaker influence.

4. Results

4.1. Xylem Vegetation Growth Carryover Effect

The vegetation growth carryover (VGC) was obtained through the IRF function. The overall trend in the effects of VGC on TWR and EXGs was similar, with a strong positive response in the early stages that gradually diminished to zero over time (Figure 6).
The influence of VGC on TRW growth and EXGs declined progressively over time. For TRW, the VGC effects disappeared entirely after six years. At the KZB site, the impacts decreased by 68.14%, 52.64%, 45.25%, 43.20%, and 42.73% year-on-year over six years. Similarly, at ZTW, the reductions were 57.01%, 46.79%, 43.58%, 42.81%, and 42.64%. In contrast, at DWZ, the impacts decreased significantly during the first two years, by 74.00% and 83.22%, respectively. This was followed by minor positive fluctuations in years three and four, with increases of 32.13% and 5.73%, before declining again by 26.99% in year six. The initial sharp decline at DWZ, followed by slight recovery and subsequent decrease, underscores the site-specific mechanisms for adapting to environmental factors. This general pattern highlights the transient nature of VGC’s influence on radial growth under typical climatic conditions (Figure 6a). The average response intensity over the six-year lag period was highest at KZB (0.40), followed by ZTW (0.035) and DWZ (0.028).
Compared to TRW, the VGC response intensity for the EXGs was more pronounced. The influence of VGC on the early xylem growth season at KZB, ZTW, and DWZ also declined over time, with year-on-year reductions of 91.20%, 98.13%, and 83.61%, respectively, over four years. During the early xylem growing season, the mean response intensities of VGC were 2.44, 1.59, and 1.70 for KZB, ZTW, and DWZ, respectively. These effects diminished to zero by the sixth, fourth, and tenth seasonal cycles, respectively (Figure 6b,c).

4.2. Lagged Climatic Effect Impact on Xylem

The lagged climatic effect (LCE) was obtained through the IRF function. Compared to explanations based on TRW that explore the entire growing season, data from the EXGs provides a stronger interpretation of climate responses. The TRW and EXGs of J. seravschanica in the three regions exhibited varying lag effects on climate across annual and seasonal scales (Figure 7).
For TRW, examined on an annual scale, the lagged climatic effect (LCE) persisted for 4–10 years, with peaks occurring within 2–3 years. The lagged response direction of TRW to MaxTEM varied among the three regions: KZB exhibited a very low lagged response intensity, while ZTW and DWZ showed inconsistent directions in lagged response intensity. The lagged response intensity and direction for MinTEM and TEM were consistent across regions, with MinTEM exhibiting the longest lag duration, up to 11 years.
Regarding the effect of PRE on TRW across the three regions, KZB and ZTW exhibited opposite response directions, while DWZ showed a minor shift from a negative to a positive response. The lag duration for ZTW was longer than that for DWZ, whereas KZB exhibited a lag of 4 years in its response to PRE.
At the scale of the xylem growth season, using seasonal intervals for investigation, the LCE on early vegetation growth persisted for 4 to 11 seasons, with peaks occurring within 2 to 3 seasons (Figure 8). The lagged effect of MaxTEM on early vegetation growth showed consistent response directions for ZTW and DWZ, with both exhibiting a bimodal transition from negative to positive responses, with lag times of six and eight seasons, respectively. In contrast, KZB displayed a unimodal positive response, with a lag time of six seasons.
For MinTEM, the lagged effect showed consistent positive response directions for KZB and ZTW, both with a lag time of 6 seasons, while DWZ exhibited a negative response with the longest lag time of 11 seasons. The lagged effect of TEM showed inconsistent response directions: KZB and DWZ had opposite response directions, whereas ZTW exhibited a bimodal response with both positive and negative peaks. The lag times across the three regions ranged from 4 to 11 seasons. For PRE, all responses showed positive peaks, with lag effects ceasing after between 7 and 11 seasons.
These results indicate that the response directions and intensities vary across different research scales and climate variables. Temperature and precipitation exert significant but variable impacts on all studied variables, underscoring the critical role of climate in vegetation dynamics.

4.3. Effects of Lagged Climatic Effect and Vegetation Growth Carryover on Xylem

By conducting decomposition and quantification of the influence of VGC and LCEs on vegetation growth through forecast error variance decomposition, the relative importance of the VGC and LCE factors on vegetation growth was determined. The growth indices of different regions (three sampling sites) and vegetation growth levels (TRW and EXGs) are more influenced by the preceding vegetation growth state (VGC, gray area) than by the lagged effects of climate variability (LCE, non-gray area) (Figure 9). From the first to the 10th lagged period, the relative influence of VGC on vegetation growth gradually weakens, while the relative influence of climate variability progressively increases.
From the perspective of TRW, the average VGC contribution rates across the 10 lagged periods (10 years) for the three regions, KZB, ZTW, and DWZ, decrease from 100% to 81.15%, 64.37%, and 79.61%, respectively (Figure 9a–c). Similarly, the average VGC contribution rates during the EGXs for the three regions decrease from 100% to 93.91%, 96.88%, and 63.84%, respectively (Figure 9d–f).
The remaining variations in TRW and the EXGs can be attributed to the lagged effects of climate factors (Figure 9).

5. Discussion

5.1. Impact of Vegetation Growth Carryover and Lagged Climatic Effect on Growth Scale of Tree-Ring Width

This study uses tree-ring width (TRW) as a quantitative indicator of radial growth to explore the dynamic characteristics of vegetation growth carryover (VGC) and the lagged climatic effect (LCE) on radial growth.
By employing the impulse response function (IRF) of the vector autoregression model VAR(q), this study further quantifies the long-term impact of climate on vegetation growth. The IRF not only captures the dynamic responses of vegetation throughout the entire growth process but also accounts for latent or unobservable changes. In contrast, traditional methods such as multiple linear regression (MLR) or partial correlation analysis, while capable of testing the significance of individual variables, often lose statistical significance for coefficients when VGC or LCE weakens or disappears [55,56]. These methods are insufficient for continuously analyzing the dynamic relationships between climate and vegetation growth. The VAR(q) model, under the optimal lag order, reveals long-term responses of climate variables to vegetation growth with greater precision, and the IRF provides continuous time-series response results. Studies indicate that, compared to traditional MLR and partial correlation analysis, the VAR(q) model is better at quantifying VGC and LCE effects on vegetation growth [46,47].
The VGC effect gradually diminishes over time, enabling the identification of longer-term vegetation growth states and climate responses, consistent with previous studies [46,57]. The results of this study demonstrate that under the conditions of VGC and the LCE, the radial growth of trees is mainly influenced by VGC, with a small effect from the LCE (Figure 6a). For the three sampling sites of KZB, ZTW, and DWZ, the VGC duration was 6 years for all of them, indicating good consistency in the influence of enhanced “signals” from the TRW growth of J. seravschanica in Tajikistan on subsequent vegetation. The VGC of KZB was stronger than that of ZTW and DWZ, which may be attributed to the earlier onset of xylem phenology in KZB (Figure 5). This led to a longer growing season than in ZTW and DWZ, allowing more photosynthetic products to accumulate and thus promoting tree radial growth [58].
Previous studies on LCE have typically analyzed the correlation between the previous year’s climate variables and the current year’s tree-ring width. Significant correlations indicate a lagged response of trees to climate [59]. However, such analyses are often limited to single-year lagged effects, with less focus on multi-year lagged effects [28,29]. The VAR model can explore the continuous lagging effects of climate and has certain advantages in this aspect of research.
Our findings reveal differences in the direction and intensity of TRW's responses to the LCE (Figure 7). TRW exhibits consistent directional responses but the responses to minimum temperature, mean temperature, and precipitation are of varying intensities. However, the response to maximum temperature shows a weak intensity at KZB, while ZTW and DWZ exhibit inconsistent response directions. This could be attributed to the harsh environmental conditions where J. seravschanica grows, characterized by scarce precipitation, high summer temperatures, and low winter temperatures. Differences in elevation further influence radial growth [60,61,62].

5.2. Impact of Vegetation Growth Carryover and Lagged Climatic Effect on the Early Xylem Growth Season

This study found that the response of tree-ring width to VGC and LCE emphasizes vegetation growth on a large temporal scale, whereas the response of the EXGs to VGC and LCEs highlights vegetation growth on a finer temporal scale.
In the EXGs of KZB, ZTW, and DWZ, the VGC in KZB was stronger than that in ZTW and DWZ (Figure 6a), likely influenced by the earlier onset of spring phenology in KZB. Both remote sensing and ground-based observations have confirmed that climate warming has significantly advanced the onset of spring greening in vegetation across mid-to-high latitudes in the Northern Hemisphere [63,64]. Related studies on the carbon cycle indicate that earlier spring phenology enhances vegetation productivity and ecosystem carbon sequestration by extending the photosynthetic period and increasing the photosynthetic rate [65,66].
The VGC duration in DWZ was longer than that in KZB and ZTW (Figure 6d), potentially due to DWZ's location on the Pamirs Plateau. High-altitude plants have evolved adaptive traits and morphologies to cope with unique plateau climates. However, vegetation in these regions experiences delayed growth onset, shorter effective growth periods, and slower growth rates, leading to a prolonged impact of earlier growth on subsequent development [67,68]. This further underscores the broad adaptability of J. seravschanica to arid continental climates.
The LCE on early vegetation growth in KZB, ZTW, and DWZ persisted for 4–11 seasons, peaking at 2–3 seasons (Figure 8). Traditional multiple linear regression models and partial correlation analyses cannot capture such dynamic signals, demonstrating the advantages of our approach.
For MaxTEM lagged effects, ZTW and DWZ exhibited consistent response directions, transitioning from negative to positive responses with a double-peaked pattern. This may be attributed to the initial adverse effects of high temperatures, which negatively impact vegetation growth. However, as these regions are at higher altitudes, sustained high temperatures accelerate snowmelt and increase soil moisture, resulting in a shift to positive responses [69,70]. For MinTEM lagged effects, DWZ showed a negative response, with the longest lag duration, 11 seasons; this might be because this sampling point is located on the Pamirs Plateau. The minimum temperature at KZB (6.40 °C) is higher than that of ZTW (1.25 °C) and DWZ (−4.77 °C). The annual average minimum temperature of DWZ is lower than that of KZB and ZTW, likely leading to severe frost damage, which significantly impacts vegetation growth and prolongs the lagged response [71]. The lagged effects of TEM displayed inconsistent response directions, with KZB and DWZ exhibiting opposite responses, while ZTW demonstrated a double-peaked response, with both positive and negative values. The lag duration across the three regions ranged from 4 to 11 seasons. Unique microclimates in each sampling site likely contributed to the varied responses of J. seravschanica to TEM.
Responses to PRE consistently exhibited positive peaks, with lag effects ceasing after 7–11 seasons. Water availability plays a critical role in the radial growth of J. seravschanica, with sufficient moisture exerting long-lasting effects on growth. Among the climatic factors influencing radial growth, precipitation was the most significant, aligning with findings from Opała-Owczarek et al. [15].

5.3. The Impact of Xylem Vegetation Carryover and Lagged Climatic Effect on the Growth Strategy of Juniperus seravschanica

In this study, the climatic conditions for the growth of J. seravschanica showed that moisture was more abundant during the growth initiation period (spring) than in summer and autumn, while summer and autumn were characterized by high temperatures and drought (Figure 3). Under such harsh climatic conditions, both tree-ring width and early xylem growth season indicate growth enhancement signals produced by vegetation under the influence of the prevailing climatic factors, and these signals continue to affect vegetation growth in subsequent periods. It was also found that vegetation growth exhibits a long-term dynamic lag response to climate effect. Additionally, previous studies [46] have shown that the VGC of different tree species responds independently to climatic factors, and its action process is not directly affected by the LCE.
Our results demonstrate that VGC contributes over 63% to both radial growth and the EXGs of J. seravschanica (Figure 9). It is demonstrated that under the current climatic conditions, the accumulated growth enhancement signals of VGC play a significant role in the radial growth (TRW) of J. seravschanica in the next stage. At the same time, it is also indicated that in the EXGs, the accumulated growth enhancement signals of VGC will have a relatively high contribution to the current xylem growth peak season. Lian et al. [17] found that while spring warming is the dominant factor increasing vegetation growth, in summer and autumn, vegetation growth is primarily governed by VGC signals from the preceding season, with a greater effect than that of climatic factors. In arid regions, researchers often underestimate the impact of the previous season’s vegetation growth enhancement signals on the current season’s radial growth of vegetation, resulting in discrepancies in calculating vegetation–climate responses [27,72].
In the TRW data, the climatic LCE contribution in ZTW was stronger than that in KZB and DWZ (Figure 9a–c), highlighting a more robust response of vegetation growth to climatic factors in ZTW. It is indicated that both the climatic conditions (LCE) and current-year climate conditions had significant influences in ZTW. The tree-ring index of ZTW had an extremely significant negative correlation with the climatic conditions of the current season (r = −0.45, p = 0.003), TEM had an extremely significant negative correlation (r = −0.42, p = 0.007), and PRE had a significant positive correlation (r = 0.31, p = 0.049) (Figure 10a). The interaction among the climatic factors led to a higher contribution of the LCE to the ZTW sample plots.
For the EXGs, the contribution of the LCE was very low in KZB and ZTW. Conversely, in DWZ, the average contribution rate of TEM (LCE) to vegetation growth was 19.38% (Figure 9f). DWZ is at the edge of the Pamirs Plateau in Central Asia, where the climate environment is harsh. The EXGs cannot immediately respond to the current-season temperature at DWZ (Figure 10b). The J. seravschanica in this region needs a certain amount of time to convert the temperature into a factor promoting growth, indicating that TEM (LCE) is an important factor affecting the EXGs of DWZ.
In similar harsh environments, previous studies on high-altitude tree-ring responses in the Tianshan and Karakoram regions have often been limited by the lack of reliable high-altitude observational data, precluding climate calibration. These studies indicate mixed climate responses for low-altitude junipers, while high-altitude junipers are primarily influenced by summer temperatures. Under the harsh conditions of high altitude, vegetation requires more temperature accumulation to promote growth. The conclusion of this study is consistent with that of Esper et al. [73,74,75].
In summary, J. seravschanica in arid regions can adjust its adaptive strategies to the harsh environment through VGC and the LCE, thereby exerting a significant influence on the subsequent growth of vegetation.

6. Conclusions and Research Uncertainties

6.1. Conclusions

Using tree-ring width (TRW) data and the early xylem growth season (EXGs), this study investigated the vegetation growth carryover (VGC) and the time-lag effect of climate on vegetation growth across two temporal scales. The key findings are as follows:
The response intensity of VGC differed between the two scales, with the VGC response during the EXGs being significantly stronger than that of the TRW. This can be attributed to the rapid cell growth during the EXGs, which facilitates the swift conversion and accumulation of carbon dioxide and solar radiation, thereby exerting a profound and lasting influence on subsequent vegetation growth.
Similarly, the intensity of the lagged climatic effect (LCE) varied between the two scales, with the LCE response for the EXGs being much stronger than that for the TRW. The duration of the LCE ranged from 0 to 11 (years or seasons) across both scales, with peak response intensity occurring during 2 to 3 (years or seasons). Among the climatic factors driving vegetation growth, under the current climate conditions, the VGC enhancement signal obtained from the previous vegetation growth will have a much stronger impact on the next stage of vegetation growth than the LCE.
These findings provide a more comprehensive and precise methodological foundation for understanding the mechanisms of long-term climatic impacts on vegetation growth processes.

6.2. Research Uncertainties

(1)
The V-S model considers only the effects of climatic factors on xylem phenology, excluding non-climatic factors that may influence cell division and tracheid differentiation [76,77].
(2)
The growth onset time simulated by the V-S model and the first day of the month with the maximum NDVI value are used to define the start of the peak growth period, thereby determining the time range of the early xylem growth season. However, this method involves cross-scale definitions and is subject to uncertainties.

Supplementary Materials

The following supporting information can be downloaded at: https://www.mdpi.com/article/10.3390/f16071107/s1, Figure S1: KZB, ZTW, and DWZ Walter-Lieth; Figure S2: Augmented Dickey–Fuller, ADF; Figure S3: The order of the lag term is determined; Table S1: Growth parameters simulated by VS model.

Author Contributions

Conceptualization, J.C. and K.L.; methodology, Z.H. software, T.Z.; validation, K.G. and T.H.; formal analysis, K.G.; investigation, Y.W.; resources, T.Z.; data curation, J.C.; writing—original draft preparation, J.C.; writing—review and editing, J.C.; visualization, J.S. and B.L.; supervision, Y.W.; project administration, Y.W. and T.Z.; funding acquisition, Y.W. and T.Z. All authors have read and agreed to the published version of the manuscript.

Funding

This research was funded by the Natural Key R&D Program of China (2023YFE0102700). It was also supported by the Tianshan Talent Training Program-Young Scientific and Technological Innovation Talent (2023TSYCCX0076), Natural Science Foundation of Xinjiang Uigur Autonomous Region-Science Fund for Distinguished Young Scholars (2022D01E105) and Regional Collaborative Innovation Project of Xinjiang (2023E01005).

Data Availability Statement

If necessary, you can consult the author for more information.

Acknowledgments

We would like to express our deepest gratitude to the editor and the anonymous reviewers for their careful review of this article. We are truly honored by their efforts.

Conflicts of Interest

The authors declare no conflicts of interest.

References

  1. Meyer, W.B.; Turner, B.L. Human population growth and global land-use/cover change. Annu. Rev. Ecol. Evol. Syst. 1992, 23, 39–61. [Google Scholar] [CrossRef]
  2. Nemani, R.R.; Keeling, C.D.; Hashimoto, H.; Jolly, W.M.; Piper, S.C.; Tucker, C.J.; Myneni, R.B.; Running, S.W. Climate-driven increases in global terrestrial net primary production from 1982 to 1999. Science 2003, 300, 1560–1563. [Google Scholar] [CrossRef]
  3. Silvestro, R.; Mencuccini, M.; García-Valdés, R.; Antonucci, S.; Arzac, A.; Biondi, F.; Buttò, V.; Camarero, J.J.; Campelo, F.; Cochard, H.J. Partial asynchrony of coniferous forest carbon sources and sinks at the intra-annual time scale. Nat. Commun. 2024, 15, 6169. [Google Scholar] [CrossRef]
  4. Heimann, M.; Reichstein, M.J.N. Terrestrial ecosystem carbon dynamics and climate feedbacks. Nature 2008, 451, 289–292. [Google Scholar] [CrossRef]
  5. Anderegg, W.R.; Schwalm, C.; Biondi, F.; Camarero, J.J.; Koch, G.; Litvak, M.; Ogle, K.; Shaw, J.D.; Shevliakova, E.; Williams, A.J.S. Pervasive drought legacies in forest ecosystems and their implications for carbon cycle models. Science 2015, 349, 528–532. [Google Scholar] [CrossRef]
  6. Anderegg, W.R.; Trugman, A.T.; Badgley, G.; Konings, A.G.; Shaw, J.J. Divergent forest sensitivity to repeated extreme droughts. Nat. Clim. Change 2020, 10, 1091–1095. [Google Scholar] [CrossRef]
  7. Liu, M.; Wang, H.; Zhai, H.; Zhang, X.; Shakir, M.; Ma, J.; Sun, W.; Meteorology, F. Identifying thresholds of time-lag and accumulative effects of extreme precipitation on major vegetation types at global scale. Agric. For. Meteorol. 2024, 358, 110239. [Google Scholar] [CrossRef]
  8. Wen, Y.; Liu, X.; Pei, F.; Li, X.; Du, G. Non-uniform time-lag effects of terrestrial vegetation responses to asymmetric warming. Agric. For. Meteorol. 2018, 252, 130–143. [Google Scholar] [CrossRef]
  9. Liu, Y.; Song, H.; Sun, C.; Song, Y.; Cai, Q.; Liu, R.; Lei, Y.; Li, Q. The 600-mm precipitation isoline distinguishes tree-ring-width responses to climate in China. Natl. Sci. Rev. 2019, 6, 359–368. [Google Scholar] [CrossRef]
  10. Wu, X.; Li, X.; Liu, H.; Ciais, P.; Li, Y.; Xu, C.; Babst, F.; Guo, W.; Hao, B.; Wang, P. Uneven winter snow influence on tree growth across temperate China. Glob. Change Biol. 2019, 25, 144–154. [Google Scholar] [CrossRef]
  11. Liu, H.; Park Williams, A.; Allen, C.D.; Guo, D.; Wu, X.; Anenkhonov, O.A.; Liang, E.; Sandanov, D.V.; Yin, Y.; Qi, Z. Rapid warming accelerates tree growth decline in semi-arid forests of Inner Asia. Glob. Change Biol. 2013, 19, 2500–2510. [Google Scholar] [CrossRef]
  12. Körner, C.J.S. A matter of tree longevity. Science 2017, 355, 130–131. [Google Scholar] [CrossRef]
  13. Babst, F.; Bodesheim, P.; Charney, N.; Friend, A.D.; Girardin, M.P.; Klesse, S.; Moore, D.J.; Seftigen, K.; Björklund, J.; Bouriaud, O. When tree rings go global: Challenges and opportunities for retro-and prospective insight. Quat. Sci. Rev. 2018, 197, 1–20. [Google Scholar] [CrossRef]
  14. Xue, H.; Shi, F.; Gennaretti, F.; Fu, Y.H.; He, B.; Wu, X.; Guo, Z. Evidence of advancing spring xylem phenology in Chinese forests under global warming. Sci. China Earth Sci. 2023, 66, 2187–2199. [Google Scholar] [CrossRef]
  15. Opała-Owczarek, M.; Owczarek, P.; Rahmonov, O.; Małarzewski, Ł.; Chen, F.; Niedźwiedź, T. Divergence in responses of juniper tree rings to climate conditions along a high-mountain transect in the semi-arid Fann Mountains, Pamirs-Alay, western Tajikistan. Ecol. Indic. 2023, 150, 110280. [Google Scholar] [CrossRef]
  16. Liu, Q.; Piao, S.; Janssens, I.A.; Fu, Y.; Peng, S.; Lian, X.; Ciais, P.; Myneni, R.B.; Peñuelas, J.; Wang, T.J.N.c. Extension of the growing season increases vegetation exposure to frost. Nat. Commun. 2018, 9, 426. [Google Scholar] [CrossRef]
  17. Lian, X.; Piao, S.; Chen, A.; Wang, K.; Li, X.; Buermann, W.; Huntingford, C.; Peñuelas, J.; Xu, H.; Myneni, R.B. Seasonal biological carryover dominates northern vegetation growth. Nat. Commun. 2021, 12, 983. [Google Scholar] [CrossRef]
  18. Lian, X.; Peñuelas, J.; Ryu, Y.; Piao, S.; Keenan, T.F.; Fang, J.; Yu, K.; Chen, A.; Zhang, Y.; Gentine, P. Diminishing carryover benefits of earlier spring vegetation growth. Nat. Ecol. Evol. 2024, 8, 218–228. [Google Scholar] [CrossRef]
  19. Wong, C.Y.; Young, D.J.; Latimer, A.M.; Buckley, T.N.; Magney, T.S. Importance of the legacy effect for assessing spatiotemporal correspondence between interannual tree-ring width and remote sensing products in the Sierra Nevada. Remote. Sens. Environ. 2021, 265, 112635. [Google Scholar] [CrossRef]
  20. Alfaro-Sánchez, R.; Nguyen, H.; Klesse, S.; Hudson, A.; Belmecheri, S.; Köse, N.; Diaz, H.; Monson, R.; Villalba, R.; Trouet, V. Climatic and volcanic forcing of tropical belt northern boundary over the past 800 years. Nat. Geosci. 2018, 11, 933–938. [Google Scholar] [CrossRef]
  21. Zhang, Z. Tree-rings, a key ecological indicator of environment and climate change. Ecol. Indic. 2015, 51, 107–116. [Google Scholar] [CrossRef]
  22. Tychkov, I.I.; Sviderskaya, I.V.; Babushkina, E.A.; Popkova, M.I.; Vaganov, E.A.; Shishov, V.V. How can the parameterization of a process-based model help us understand real tree-ring growth? Trees 2019, 33, 345–357. [Google Scholar] [CrossRef]
  23. Shishov, V.V.; Tychkov, I.I.; Popkova, M.I.; Ilyin, V.A.; Bryukhanova, M.V.; Kirdyanov, A.V. VS-oscilloscope: A new tool to parameterize tree radial growth based on climate conditions. Dendrochronologia 2016, 39, 42–50. [Google Scholar] [CrossRef]
  24. He, M.; Yang, B.; Wang, Z.; Bräuning, A.; Pourtahmasi, K.; Oladi, R. Climatic forcing of xylem formation in Qilian juniper on the northeastern Tibetan Plateau. Trees 2016, 30, 923–933. [Google Scholar] [CrossRef]
  25. Anchukaitis, K.J.; Evans, M.N.; Hughes, M.K.; Vaganov, E.A. An interpreted language implementation of the Vaganov–Shashkin tree-ring proxy system model. Dendrochronologia 2020, 60, 125677. [Google Scholar] [CrossRef]
  26. Kang, J.; Yang, Z.; Yu, B.; Ma, Q.; Jiang, S.; Shishov, V.V.; Zhou, P.; Huang, J.-G.; Ding, X.J.; Meteorology, F. An earlier start of growing season can affect tree radial growth through regulating cumulative growth rate. Agric. For. Meteorol. 2023, 342, 109738. [Google Scholar] [CrossRef]
  27. Jiao, L.; Jiang, Y.; Zhang, W.-T.; Wang, M.-C.; Zhang, L.-N.; Zhao, S.-D. Divergent responses to climate factors in the radial growth of Larix sibirica in the eastern Tianshan Mountains, northwest China. Trees 2015, 29, 1673–1686. [Google Scholar] [CrossRef]
  28. Fan, Y.; Shang, H.; Yu, S.; Wu, Y.; Li, Q. Understanding the representativeness of Tree rings and their Carbon isotopes in characterizing the Climate Signal of Tajikistan. Forests 2021, 12, 1215. [Google Scholar] [CrossRef]
  29. Gou, X.-X.; Zhang, T.-W.; Yu, S.-L.; Zhang, R.-B.; Jiang, S.-X.; Guo, Y.-L. Response of radial growth of Juniperus seravschanica to climate changes in different environmental conditions. J. Ecol. 2021, 40, 1574. [Google Scholar] [CrossRef]
  30. Kure, S.; Jang, S.; Ohara, N.; Kavvas, M.; Chen, Z. Hydrologic impact of regional climate change for the snow-fed and glacier-fed river basins in the Republic of Tajikistan: Statistical downscaling of global climate model projections. Hydrol. Process. 2013, 27, 4071–4090. [Google Scholar] [CrossRef]
  31. Nowak, A.; Nowak, S.; Nobis, M. The Pamirs-Alai mountains (middle Asia: Tajikistan). In Plant Biogeography and Vegetation of High Mountains of Central and South-West Asia; Springer: Berlin/Heidelberg, Germany, 2020; pp. 1–42. [Google Scholar] [CrossRef]
  32. Aalto, J.; Kämäräinen, M.; Shodmonov, M.; Rajabov, N.; Venäläinen, A. Features of Tajikistan's past and future climate. Int. J. Clim. 2017, 37, 4949–4961. [Google Scholar] [CrossRef]
  33. Kure, S.; Jang, S.; Ohara, N.; Kavvas, M.; Chen, Z. Hydrologic impact of regional climate change for the snowfed and glacierfed river basins in the Republic of Tajikistan: Hydrological response of flow to climate change. Hydrol. Process. 2013, 27, 4057–4070. [Google Scholar] [CrossRef]
  34. Yang, T.; Li, Q.; Chen, X.; De Maeyer, P.; Yan, X.; Liu, Y.; Zhao, T.; Li, L. Spatiotemporal variability of the precipitation concentration and diversity in Central Asia. Atmos. Res. 2020, 241, 104954. [Google Scholar] [CrossRef]
  35. Holmes, R.L. Computer-Assisted Quality Control in Tree-Ring Dating and Measurement; Tree-Ring Society: Tucson, AZ, USA, 1983. [Google Scholar]
  36. Huang, Y.; Jiang, N.; Shen, M.; Guo, L. Effect of preseason diurnal temperature range on the start of vegetation growing season in the Northern Hemisphere. Ecol. Indic. 2020, 112, 106161. [Google Scholar] [CrossRef]
  37. Chen, H.; Zhao, J.; Zhang, H.; Zhang, Z.; Guo, X.; Wang, M. Detection and attribution of the start of the growing season changes in the Northern Hemisphere. Sci. Total. Environ. 2023, 903, 166607. [Google Scholar] [CrossRef]
  38. Xu, X.; Riley, W.J.; Koven, C.D.; Jia, G.; Zhang, X. Earlier leaf-out warms air in the north. Nat. Clim. Change 2020, 10, 370–375. [Google Scholar] [CrossRef]
  39. Huang, J.; van den Dool, H.M.; Georgarakos, K.P. Analysis of model-calculated soil moisture over the United States (1931–1993) and applications to long-range temperature forecasts. J. Clim. 1996, 9, 1350–1362. [Google Scholar] [CrossRef]
  40. Furze, M.E.; Huggett, B.A.; Aubrecht, D.M.; Stolz, C.D.; Carbone, M.S.; Richardson, A.D. Whole-tree nonstructural carbohydrate storage and seasonal dynamics in five temperate species. New Phytol. 2019, 221, 1466–1477. [Google Scholar] [CrossRef]
  41. Hartmann, F.P.; Rathgeber, C.B.; Fournier, M.; Moulia, B.J. Modelling wood formation and structure: Power and limits of a morphogenetic gradient in controlling xylem cell proliferation and growth. Ann. For. Sci. 2017, 74, 14. [Google Scholar] [CrossRef]
  42. Antonucci, S.; Rossi, S.; Deslauriers, A.; Morin, H.; Lombardi, F.; Marchetti, M.; Tognetti, R.J.A.; Meteorology, F. Large-scale estimation of xylem phenology in black spruce through remote sensing. Agric. For. Meteorol. 2017, 233, 92–100. [Google Scholar] [CrossRef]
  43. Huang, J.G.; Deslauriers, A.; Rossi, S.J.N.P. Xylem formation can be modeled statistically as a function of primary growth and cambium activity. New Phytol. 2014, 203, 831–841. [Google Scholar] [CrossRef]
  44. Yang, B.; He, M.; Shishov, V.; Tychkov, I.; Vaganov, E.; Rossi, S.; Ljungqvist, F.C.; Bräuning, A.; Grießinger, J.J.P.o.t.N.A.o.S. New perspective on spring vegetation phenology and global climate change based on Tibetan Plateau tree-ring data. Proc. Natl. Acad. Sci. USA 2017, 114, 6966–6971. [Google Scholar] [CrossRef]
  45. Piermattei, A.J. A new perspective on tree growing season determination. J. Biogeogr. 2024, 51, 2334–2337. [Google Scholar] [CrossRef]
  46. Tang, W.; Liu, S.; Jing, M.; Healey, J.R.; Smith, M.N.; Farooq, T.H.; Zhu, L.; Zhao, S.; Wu, Y. Vegetation growth responses to climate change: A cross-scale analysis of biological memory and time lags using tree ring and satellite data. Glob. Change Biol. 2024, 30, e17441. [Google Scholar] [CrossRef]
  47. Tang, W.; Liu, S.; Kang, P.; Peng, X.; Li, Y.; Guo, R.; Jia, J.; Liu, M.; Zhu, L. Quantifying the lagged effects of climate factors on vegetation growth in 32 major cities of China. Ecol. Indic. 2021, 132, 108290. [Google Scholar] [CrossRef]
  48. Lee, C.-C.; Chang, C.-P. New evidence on the convergence of per capita carbon dioxide emissions from panel seemingly unrelated regressions augmented Dickey–Fuller tests. Energy 2008, 33, 1468–1475. [Google Scholar] [CrossRef]
  49. Maziarz, M. A review of the Granger-causality fallacy. J. Philos. Econ. Reflect. Econ. Soc. Issues 2015, 8, 86–105. [Google Scholar] [CrossRef]
  50. Hansen, P.R. Structural changes in the cointegrated vector autoregressive model. J. Econ. 2003, 114, 261–295. [Google Scholar] [CrossRef]
  51. Lütkepohl, H. Vector autoregressive models. In Handbook of Research Methods and Applications in Empirical Macroeconomics; Edward Elgar Publishing: Cheltenham, UK, 2013; pp. 139–164. [Google Scholar]
  52. Haslbeck, J.M.; Bringmann, L.F.; Waldorp, L.J. A tutorial on estimating time-varying vector autoregressive models. Multivar. Behav. Res. 2021, 56, 120–149. [Google Scholar] [CrossRef]
  53. Volpicella, A.J.; Statistics, E. SVARs identification through bounds on the forecast error variance. J. Bus. Econ. Stat. 2022, 40, 1291–1301. [Google Scholar] [CrossRef]
  54. Gujarati, D.N. Essentials of Econometrics; Sage Publications: Thousand Oaks, CA, USA, 2021. [Google Scholar]
  55. Yuan, Y.; Bao, A.; Jiapaer, G.; Jiang, L.; De Maeyer, P.J. Phenology-based seasonal terrestrial vegetation growth response to climate variability with consideration of cumulative effect and biological carryover. Sci. Total. Environ. 2022, 817, 152805. [Google Scholar] [CrossRef]
  56. Zhang, X.; Cao, Q.; Chen, H.; Quan, Q.; Li, C.; Dong, J.; Chang, M.; Yan, S.; Liu, J. Effect of vegetation carryover and climate variability on the seasonal growth of vegetation in the upper and middle reaches of the Yellow River Basin. Remote. Sens. 2022, 14, 5011. [Google Scholar] [CrossRef]
  57. Huang, M.; Wang, X.; Keenan, T.F.; Piao, S.J. Drought timing influences the legacy of tree growth recovery. Glob. Change Biol. 2018, 24, 3546–3559. [Google Scholar] [CrossRef]
  58. Gao, S.; Liang, E.; Liu, R.; Babst, F.; Camarero, J.J.; Fu, Y.H.; Piao, S.; Rossi, S.; Shen, M.; Wang, T.J.; et al. An earlier start of the thermal growing season enhances tree growth in cold humid areas but not in dry areas. Nat. Ecol. Evol. 2022, 6, 397–404. [Google Scholar] [CrossRef]
  59. Wu, G.; Xu, G.; Chen, T.; Liu, X.; Zhang, Y.; An, W.; Wang, W.; Fang, Z.-A.; Yu, S.J. Age-dependent tree-ring growth responses of Schrenk spruce (Picea schrenkiana) to climate—A case study in the Tianshan Mountain, China. Dendrochronologia 2013, 31, 318–326. [Google Scholar] [CrossRef]
  60. Lavergne, A.; Daux, V.; Villalba, R.; Barichivich, J. Temporal changes in climatic limitation of tree-growth at upper treeline forests: Contrasted responses along the west-to-east humidity gradient in Northern Patagonia. Dendrochronologia 2015, 36, 49–59. [Google Scholar] [CrossRef]
  61. Gou, X.; Zhang, T.; Yu, S.; Liu, K.; Zhang, R.; Shang, H.; Qin, L.; Fan, Y.; Jiang, S.; Zhang, H. Climate response of Picea schrenkiana based on tree-ring width and maximum density. Dendrochronologia 2023, 78, 126067. [Google Scholar] [CrossRef]
  62. Zhang, H.; Cai, Q.; Liu, Y. Altitudinal difference of growth-climate response models in the north subtropical forests of China. Dendrochronologia 2022, 72, 125935. [Google Scholar] [CrossRef]
  63. Liu, Q.; Fu, Y.H.; Zhu, Z.; Liu, Y.; Liu, Z.; Huang, M.; Janssens, I.A.; Piao, S.J. Delayed autumn phenology in the Northern Hemisphere is related to change in both climate and spring phenology. Glob. Change Biol. 2016, 22, 3702–3711. [Google Scholar] [CrossRef]
  64. Schwartz, M.D.; Ahas, R.; Aasa, A.J. Onset of spring starting earlier across the Northern Hemisphere. Glob. Change Biol. 2006, 12, 343–351. [Google Scholar] [CrossRef]
  65. Keeling, C.D.; Chin, J.; Whorf, T.J.N. Increased activity of northern vegetation inferred from atmospheric CO2 measurements. Nature 1996, 382, 146–149. [Google Scholar] [CrossRef]
  66. Piao, S.; Friedlingstein, P.; Ciais, P.; Viovy, N.; Demarty, J. Growing season extension and its impact on terrestrial carbon cycle in the Northern Hemisphere over the past 2 decades. Glob. Biogeochem. Cycles 2007, 21, 22151–22156. [Google Scholar] [CrossRef]
  67. Yu, H.; Luedeling, E.; Xu, J. Winter and spring warming result in delayed spring phenology on the Tibetan Plateau. Proc. Natl. Acad. Sci. USA 2010, 107, 22151–22156. [Google Scholar] [CrossRef]
  68. Liu, N.; Peng, S.-Z.; Chen, Y.-M. Temporal effects of climate factors on vegetation growth on the Qingzang Plateau, China. Chin. J. Plan Ecol. 2022, 46, 18. [Google Scholar] [CrossRef]
  69. Li, Y.; Chen, Y. The continuing shrinkage of snow cover in High Mountain Asia over the last four decades. In Proceedings of the EGU General Assembly Conference Abstracts, Vienna, Austria, 23–28 August 2023; p. EGU–16661. [Google Scholar]
  70. Wang, X.; Wang, T.; Guo, H.; Liu, D.; Zhao, Y.; Zhang, T.; Liu, Q.; Piao, S. Disentangling the mechanisms behind winter snow impact on vegetation activity in northern ecosystems. Glob. Change Biol. 2018, 24, 1651–1662. [Google Scholar] [CrossRef]
  71. Jiang, B.; Shi, Y.; Peng, Y.; Jia, Y.; Yan, Y.; Dong, X.; Li, H.; Dong, J.; Li, J.; Gong, Z.J. Cold-induced CBF–PIF3 interaction enhances freezing tolerance by stabilizing the phyB thermosensor in Arabidopsis. Mol. Plant 2020, 13, 894–906. [Google Scholar] [CrossRef]
  72. Che, C.; Xiao, S.; Peng, X.; Ding, A.; Su, J. Radial growth of Korshinsk peashrub and its response to drought in different sub-arid climate regions of northwest China. J. Environ. Manag. 2023, 326, 116708. [Google Scholar] [CrossRef]
  73. Esper, J. Long-term tree-ring variations in Juniperus at the upper timber-line in the Karakorum (Pakistan). Holocene 2000, 10, 253–260. [Google Scholar] [CrossRef]
  74. Esper, J.; Cook, E.R.; Krusic, P.J.; Peters, K.; Schweingruber, F.H. Tests of the RCS method for preserving low-frequency variability in long tree-ring chronologies. Tree-Ring Res. 2003, 59, 81–98. [Google Scholar]
  75. Esper, J.; Frank, D.C.; Wilson, R.J.; Büntgen, U.; Treydte, K.J.T. Uniform growth trends among central Asian low-and high-elevation juniper tree sites. Trees 2007, 21, 141–150. [Google Scholar] [CrossRef]
  76. Vaganov, E.A.; Hughes, M.K.; Shashkin, A.V. Growth Dynamics of Conifer Tree Rings: Images of Past and Future Environments; Springer Science & Business Media: Berlin/Heidelberg, Germany, 2006; Volume 183. [Google Scholar]
  77. Tumajer, J.; Kašpar, J.; Kuželová, H.; Shishov, V.V.; Tychkov, I.I.; Popkova, M.I.; Vaganov, E.A.; Treml, V.J. Forward modeling reveals multidecadal trends in cambial kinetics and phenology at treeline. Front. Plant Sci. 2021, 12, 613643. [Google Scholar] [CrossRef]
Figure 1. The sampling sites are located in the northern Khujand mountain area, the central western Alay mountain area, and the southern marginal Pamirs Plateau. The white graph element source from ArcMap 10.8: https://services.arcgisonline.com/arcgis, accessed on 3 June 2025.
Figure 1. The sampling sites are located in the northern Khujand mountain area, the central western Alay mountain area, and the southern marginal Pamirs Plateau. The white graph element source from ArcMap 10.8: https://services.arcgisonline.com/arcgis, accessed on 3 June 2025.
Forests 16 01107 g001
Figure 2. (ac) represent KZB, ZTW, and DWZ, respectively. The average precipitation (PRE) and monthly average temperatures (MaxTEM: average maximum temperature, TEM: average temperature, MinTEM: average minimum temperature) from January to December during 1982–2022 are shown.
Figure 2. (ac) represent KZB, ZTW, and DWZ, respectively. The average precipitation (PRE) and monthly average temperatures (MaxTEM: average maximum temperature, TEM: average temperature, MinTEM: average minimum temperature) from January to December during 1982–2022 are shown.
Forests 16 01107 g002
Figure 3. Chronologies (solid lines) with sample depth (dashed lines). (a) KZB; (b) ZTW; (c) DWZ.
Figure 3. Chronologies (solid lines) with sample depth (dashed lines). (a) KZB; (b) ZTW; (c) DWZ.
Forests 16 01107 g003
Figure 4. Verifies the reliability of simulated xylem phenology and remote sensing phenology.
Figure 4. Verifies the reliability of simulated xylem phenology and remote sensing phenology.
Forests 16 01107 g004
Figure 5. (a) Multi-year monthly average NDVI values from 1982 to 2022 for KZB, ZTW, and DWZ. The month with the maximum NDVI value for KZB and ZTW is May, while for DWZ it is July. (b) The trend in the early xylem growth season (EXGs) from 1982 to 2022.
Figure 5. (a) Multi-year monthly average NDVI values from 1982 to 2022 for KZB, ZTW, and DWZ. The month with the maximum NDVI value for KZB and ZTW is May, while for DWZ it is July. (b) The trend in the early xylem growth season (EXGs) from 1982 to 2022.
Forests 16 01107 g005
Figure 6. Response curves of vegetation growth carryover (VGC). (a) Response curve for tree-ring width (TRW). (bd) Response curves for the early xylem growth season (EXGs). The light gray shaded areas indicate the duration of the carryover effects.
Figure 6. Response curves of vegetation growth carryover (VGC). (a) Response curve for tree-ring width (TRW). (bd) Response curves for the early xylem growth season (EXGs). The light gray shaded areas indicate the duration of the carryover effects.
Forests 16 01107 g006
Figure 7. The lagged climatic effect (LCE) curves of tree-ring width (TRW). The Y-axis represents the response intensity of each vegetation indicator, while the X-axis denotes the lag period of vegetation growth in response to climate variability. The calculations are based on the impulse response function (IRF). The shaded light gray area indicates the duration of the lag period.
Figure 7. The lagged climatic effect (LCE) curves of tree-ring width (TRW). The Y-axis represents the response intensity of each vegetation indicator, while the X-axis denotes the lag period of vegetation growth in response to climate variability. The calculations are based on the impulse response function (IRF). The shaded light gray area indicates the duration of the lag period.
Forests 16 01107 g007
Figure 8. The lagged climatic effect (LCE) response curves during the early xylem growth season (EXGs). The Y-axis represents the response intensity of vegetation indices, while the X-axis denotes the lag period for vegetation growth in response to climate variability. Calculations were performed using the impulse response function (IRF). The light gray area indicates the duration of the lag period.
Figure 8. The lagged climatic effect (LCE) response curves during the early xylem growth season (EXGs). The Y-axis represents the response intensity of vegetation indices, while the X-axis denotes the lag period for vegetation growth in response to climate variability. Calculations were performed using the impulse response function (IRF). The light gray area indicates the duration of the lag period.
Forests 16 01107 g008
Figure 9. The influence of vegetation growth carryover (VGC) and lagged climatic effect (LCE) on climate-related indicators. (ac) Growth indices for single-site tree-ring width (TRW) at KZB, ZTW, and DWZ, lagged annually. (df) The early xylem growth season (EXGs) for KZB, ZTW, and DWZ, lagged seasonally. The Y-axis represents the contributions of mean MaxTEM, MinTEM, TEM, PRE, and VGC. For (ac), the Y-axis starts at 60%, as 0%–60% of the contribution is attributed to VGC. For (df), the Y-axis starts at 70%, as 0%–70% of the contribution is attributed to VGC. The contributions of VGC and LCE factors to vegetation growth variables were calculated using forecast error variance decomposition analysis.
Figure 9. The influence of vegetation growth carryover (VGC) and lagged climatic effect (LCE) on climate-related indicators. (ac) Growth indices for single-site tree-ring width (TRW) at KZB, ZTW, and DWZ, lagged annually. (df) The early xylem growth season (EXGs) for KZB, ZTW, and DWZ, lagged seasonally. The Y-axis represents the contributions of mean MaxTEM, MinTEM, TEM, PRE, and VGC. For (ac), the Y-axis starts at 60%, as 0%–60% of the contribution is attributed to VGC. For (df), the Y-axis starts at 70%, as 0%–70% of the contribution is attributed to VGC. The contributions of VGC and LCE factors to vegetation growth variables were calculated using forecast error variance decomposition analysis.
Forests 16 01107 g009
Figure 10. Correlation analysis of TRW and EXGs with the climate of the current year or season. (a) Correlation analysis of TRW with MaxTEM, MinTEM, TEM, and PRE of the current year in three sample plots. (b) Correlation analysis of EXGs with MaxTEM, MinTEM, TEM, and PRE of the current season in three sample plots. ** indicates extremely significant correlation (p < 0.01), * indicates significant correlation (p < 0.05).
Figure 10. Correlation analysis of TRW and EXGs with the climate of the current year or season. (a) Correlation analysis of TRW with MaxTEM, MinTEM, TEM, and PRE of the current year in three sample plots. (b) Correlation analysis of EXGs with MaxTEM, MinTEM, TEM, and PRE of the current season in three sample plots. ** indicates extremely significant correlation (p < 0.01), * indicates significant correlation (p < 0.05).
Forests 16 01107 g010
Table 1. Main characteristic parameters of tree-ring width standardized chronology.
Table 1. Main characteristic parameters of tree-ring width standardized chronology.
TypeKZBZTWDWZ
Latitude40.646455° N39.3531° N38.586636° N
Longitude69.612838° E67.510412° E70.857156° E
Altitude/m1575.72147.12073
Slope directionSNWS
Sample core/tree52/2640/2039/21
Sample depth1896–20231846–20231929–2023
The first year of subsample signal strength
(SSS > 0.85)
189618461929
Standard deviation (SD)0.2470.3610.193
Mean sensitivity (MS)0.2000.2790.196
Mean correlation coefficient among all series (R1)0.2840.2250.195
Mean correlation coefficient between trees (R2)0.2700.2030.160
Mean correlation coefficient within trees (R3)0.8210.7870.659
Express population signal (EPS)0.9360.8830.882
Signal-to-noise ratio (SNR)14.5017.5517.51
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

Chen, J.; Wang, Y.; Zhang, T.; Liu, K.; Guo, K.; Hou, T.; Song, J.; He, Z.; Liang, B. Vegetation Growth Carryover and Lagged Climatic Effect at Different Scales: From Tree Rings to the Early Xylem Growth Season. Forests 2025, 16, 1107. https://doi.org/10.3390/f16071107

AMA Style

Chen J, Wang Y, Zhang T, Liu K, Guo K, Hou T, Song J, He Z, Liang B. Vegetation Growth Carryover and Lagged Climatic Effect at Different Scales: From Tree Rings to the Early Xylem Growth Season. Forests. 2025; 16(7):1107. https://doi.org/10.3390/f16071107

Chicago/Turabian Style

Chen, Jiuqi, Yonghui Wang, Tongwen Zhang, Kexiang Liu, Kailong Guo, Tianhao Hou, Jinghui Song, Zhihao He, and Beihua Liang. 2025. "Vegetation Growth Carryover and Lagged Climatic Effect at Different Scales: From Tree Rings to the Early Xylem Growth Season" Forests 16, no. 7: 1107. https://doi.org/10.3390/f16071107

APA Style

Chen, J., Wang, Y., Zhang, T., Liu, K., Guo, K., Hou, T., Song, J., He, Z., & Liang, B. (2025). Vegetation Growth Carryover and Lagged Climatic Effect at Different Scales: From Tree Rings to the Early Xylem Growth Season. Forests, 16(7), 1107. https://doi.org/10.3390/f16071107

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