Modeling Ecological Resilience of Alpine Forest under Climate Change in Western Sichuan

Abstract

The ecological resilience of forests is the ability to return to a stable state after being subjected to external disturbances, and it is among the critical indicators of forest status. Climate change has significant effects on forest ecological resilience and diversity. In this research, we selected Mao County as the study region, and employed the forest landscape model LANDIS-II to simulate the effect of different climate scenarios on the ecological resilience of alpine forests in western Sichuan during the next 300 years from the forest composition and structure perspective. The findings revealed that: (1) climate change will favor an increase in forest ecological resilience values in short simulations, but future climate scenarios will negatively impact the ecological resilience of forests as the simulation progresses through the middle and long term. (2) The rate of change of forest ecological resilience in the MTDF and SCF ecotones, which have a higher proportion of Fir (Abies fabri) and Spruce (Picea asperata), was greater than that in the rest of the ecotones in the short-term simulation. In contrast, it was the opposite in the medium-term simulation. The rate of change of forest ecological resilience was more significant in the long-term simulation in all four ecotones. (3) The high values of forest ecological resilience in the short- and medium-term simulations were primarily concentrated within the MTDF and SCF ecotones among the midwestern and northern parts of the study region. When the simulation proceeded to a later stage, the ecological resilience of the forests decreased significantly throughout the study region, with high values occurring only in some areas within the western parts of the study region. The research results can grasp the influence of future climate on the ecological resilience of high mountain forests within western Sichuan and provide an essential reference for the sustainable development of local forests.

1. Introduction

The forest is the most critical ecosystem on land and is essential to preserving biodiversity, improving the ecological environment, and maintaining the global carbon balance [1,2]. Forest ecosystems are inevitably subject to various internal and external disturbances [3,4], among which climate change is closely related to forest ecosystems [5], and the composition and structure of forest ecosystems are impacted by climate change on various scales [6,7]. The effects of climate change on forests are progressively worsening with global warming [8,9], and climate change directly impacts the productivity and growth rate of plants in forest ecosystems by changing environmental factors such as soil moisture content and air humidity [10,11]. At the same time, natural catastrophes will occur more frequently and more intense impacts caused by climate change will occur, such as forest fires, pests and diseases, and extreme weather, thus changing the composition and structure of forest ecosystems, increasing the vulnerability of forest ecosystems to natural disasters and threatening the ecological stability of forests [12,13].

Forest ecological resilience is an intrinsic attribute of forests, which indicates the ability of forest ecosystems to reestablish a steady state after an external disturbance [14,15], including its composition, structure, and ecosystem function, and it is a crucial goal for the sustainable development of forests [16,17]. Affected by climate change, some forests are becoming less resilient to external disturbances. Forest factors such as temperature, moisture, and soil will change with climate change [18]. The regional and temporal trends of forest ecological resilience will change as a result of these changing factors, decreasing the ability of forest ecosystems to recover from disruptions [19].

Alpine forests in western Sichuan are located on the southeastern edge of the Qinghai-Tibetan Plateau, a region with complex topography and high climate sensitivity, which is the main constituent of the Southwest Forest Region, China’s second-largest forest region [20,21]. Under the standard climate context, due to forest ecological resilience, the alpine forest ecosystems in western Sichuan have a series of abilities to maintain their original forest structure and function after disturbance [22,23]. However, as global warming is still the primary trend of the current and future climate, it will influence the forest’s ecological resilience’s dynamic changes in the region [24], and these changes may lead to the reduction of the ecological resilience of the region’s forests, which will cause damage to forest ecosystems and regional ecological security. At the same time, the impacts of climate change on forest ecological resilience in the western Sichuan Province alpine forest region may be enhanced because forest composition and structure have suffered significant damage, and the forest ecosystems are relatively fragile due to the impacts of large amounts of logging, overgrazing, forest fires, and other disturbances in the western Sichuan alpine forest region in the late twentieth century [25]. Furthermore, current research on alpine forests in west Sichuan mainly focuses on forest structure and biomass, and less research has been done on how climate change affects the ecological resilience of forests. In summary, it is essential to recognize and analyze the changes and trends of local forest ecological resilience to maintain the western Sichuan alpine forest’s ecological stability and long-term development [26].

Forest ecological resilience is a cross-scale long-term dynamic change process [27], and its analysis requires the selection of corresponding resilience indicators, as it is difficult to monitor the changes in forest ecological resilience in long-term broad geographical and temporal scales through traditional field observation [28]. Therefore, the use of modeling simulation studies has become the most common method of research into the dynamic change of forest landscapes at home and abroad [29,30,31]. For example, Mina et al. predicted the changes in forest composition and productivity using the LANDIS-II model in southeastern Canada under different climatic scenarios from 2000 to 2200; Duan et al. [32] used the LANDIS PRO model to simulate the changes in species composition ratios of hardwood forests in the central U.S. due to climate change in the period of 2000 to 2300; Dai et al. [33] combined the LANDIS-II and PnET-II models to simulate and predict the response of plantation forests in southern China under three future climatic scenarios to climate change. Therefore, in this paper, we selected the LANDIS-II forest landscape model to simulate the dynamic changes of forest ecological resilience in the alpine forests of western Sichuan under different climate scenarios for 300 years and combined the changes of two elasticity indicators, aboveground biomass of the forests in the study region, and the ratio of the maximum age of the trees to the longevity of the tree species to carry out the computation of forest ecological resilience and quantify the long term. The forest ecological resilience was calculated to quantify the long-term spatial and temporal dynamic changes of forest ecological resilience indexes under different climatic conditions in the Alpine forests in western Sichuan to give the sustainable development of the region’s forest resources a scientific basis.

2. Materials and Methods

2.1. Overview of the Study Region

The geographical range of this study region is 102°56′ to 104°10′ E, 31°25′ to 32°16′ N, located in Mao County, Aba Tibetan and Qiang Autonomous Prefecture, Sichuan Province. It belongs to the alpine forest region in western Sichuan, having a total area of about 3903.28 square units (Figure 1). The research area is situated on the southeastern edge of the Qinghai-Tibet Plateau, in the northern section of the Hengduan Mountains. The local climate is complex, and high mountains and gorges dominate the terrain. The elevation varies greatly, ranging from 890 to 5320 m (Figure 1b). The Mao County area is rich in biodiversity. Dominant tree species in the forest communities of this area include Fir (Abies fabri), Spruce (Picea asperata), Chinese hemlock (Tsuga chinensis), Chinese pine (Pinus tabuliformis), Liaodong oak (Quercus wutaishanica Mayr), Ring-cupped oak (Cyclobalanopsis glauca), Birch (Betula spp.), and Maple (Acer spp.). Among them, Fir and Spruce dominate the study region. According to elevation, forest type, and soil data, this study divides the research area into five ecotones: evergreen broad-leaved forest (EBF), mixed evergreen, deciduous broad-leaved forest (MEDBF), mixed coniferous and deciduous broad-leaved forest (MTDF), subalpine coniferous forest (SCF), and non-forest area (NF) (Figure 1c).

2.2. Research Methodology

2.2.1. Climate Data

To investigate the effects of climate change on the ecological resilience of forests, we considered current climate scenarios as well as three future climate change scenarios (RCP2.6, RCP4.5, RCP8.5) in our study. The current climate scenario was set as the baseline to extrapolate current climate conditions (Baseline) into the future and set to standard climate (SC). Among them, the temperature and precipitation information is used as the observation data of the monthly average of the meteorological observation stations in the Mao County area during the multi-year period of 1986–2016. Future climate change data were rederived from the Coupled Model Intercomparison Project Phase 5 (CMIP5) framework proposed by the World Climate Research Program; they are compiled by averaging projections from multiple climate models, which are then reduced and interpolated to a 30-inch grid by the MarkSimGCM weather generator (http://gisweb.ciat.cgiar.org/MarkSimGCM/, accessed on 9 January 2023). Since MarkSimGCM can only generate data up to 2095, we assumed that the climate parameter data in the study area remained constant after 2095. Of the three future climate change scenarios we selected, the RCP2.6 scenario assumes that temperatures continue to increase by 2 °C or less until 2050 and that the radiative forcing decreases to 2.6 W/${\mathrm{m}}^2$ by the end of the century; the RCP4.5 scenario assumes that temperatures continue to increase until 2070, after which the rate of increase slows down, and the radiative forcing stabilizes at 4.5 W/${\mathrm{m}}^2$ by the end of the century, and the RCP8.5 scenario assumes a more extreme climate, which indicates that by the end of the century, the warming will reach 5 °C.

2.2.2. Introduction to the Model

In this paper, we combined the forest landscape model LANDIS-II and the ecosystem process model PnET-II to simulate the changes in forest ecological resilience of alpine forests within the west Sichuan region under standard climate and climate change scenarios in the next 300 years. The forest landscape model LANDIS-II is a spatially intuitive forest landscape model, which is primarily employed for the simulation of natural succession, seed dispersal, forest disturbances (climate change, pests, diseases, forest fires, etc.), anthropogenic disturbances (harvesting, planting, etc.), and their interactions with forest ecosystems [34]. LANDIS-II is a way to view the landscape as a grid of interacting image elements. Initial species information is obtained from remotely sensed imagery or tree species distribution maps [35]. Each pixel is divided into ecotones with similar environments and the same stand conditions, and the life history parameters and seed germination ability of tree species in different environmental zones will determine their growth [36]. There are various extension modules in the LANDIS-II model to achieve other research objectives; this paper uses the model’s Biomass Succession and Age-only Succession modules to realize the succession of forest biomass and tree age [37]. The Biomass Succession module simulates the change of biomass of the selected tree species over time in each pixel to output the raster map of biomass in different simulated years, taking into account the life history parameters of each species, while the Age-only Succession module acts on the aging and mortality of tree species in each pixel in the study region and outputs raster maps with different age parameters. The PnET-II model is used to simulate the dynamic processes of cycling of carbon, nitrogen, and water within forest ecosystems [38]. By inputting climate data, soil parameters, and physiological parameters of selected tree species in the study region, we can simulate and estimate each species’ ANPP and SEP parameters required to operate the LANDIS-II model under different climate change scenarios [23].

2.2.3. Indicator Selection

Forest ecological resilience is a long-term dynamic change process, affected by a variety of factors, and it is more difficult to measure forest ecological resilience directly, so the primary method is to grasp the dynamic change of forest ecological resilience values by selecting indicator factors that can reflect the key characteristics of local forest ecological resilience [39]. The forest ecological resilience indicators chosen in this paper were based on the research of Luo Xu et al. [40] and Mehdi et al. [41], combined with the situation of the study region, from the aspects of forest structure and forest composition. The forest ecological resilience index was selected in line with the research characteristics of the region, namely, forest aboveground biomass (AGB) and the ratio of the maximum age of each species to the longevity of each species. Among them, forest aboveground biomass is an important index reflecting the function of the forest ecosystem [42]. The ratio of the maximum age of each species to the longevity of each species is selected because the maximum tree age of each tree species is more sensitive to external interference. It can show more obvious changes in the ecological resilience of forests in the face of external disturbances [43]. Finally, the coefficient of variation method was used to determine the weighting coefficients of different forest ecological resilience indicators. The climate scenarios were selected concerning the typical concentration pathways of IPCC, in which the three future climate change scenarios of RCP2.6, RCP4.5, and RCP8.5 were selected, and the current climate scenario was set as SC (Standard climate).

2.2.4. Model Parameterization

In this paper, the primary 16 tree species in the region were selected to simulate forest dynamic succession of forests from 2016 to 2316 in 300 years with a time step of 5 years. The climatic data for the PnET-II model were selected from the monthly average of the meteorological observation stations in Mao County from 1986 to 2016, the soil data were obtained from the China Soil Database, and the physiological parameters of the selected tree species were obtained from the forestry survey data in Mao County. The operation of the LANDIS-II model requires GIS graphic parameter files and attribute parameter files. The required GIS graphic parameter files, namely, the initial tree species distribution raster data, are from the forest map and field survey data in Mao County 2016. The resolution is 100 m × 100 m, and the number of rows and columns of the simulated map is 1166 × 942. The attribute parameter files, the life history attributes of the selected tree species, the tree species establishment coefficient, etc., are mainly obtained from field surveys, inquiries with local forestry experts, and references to relevant literature [44,45] (

Table 1. Life history parameters of main tree species in Mao County.

Scientific Name	Lng	Mat	Shd	EffD	MaxD
Fir (Abies fabri)	300	60	4	50	150
Spruce (Picea asperata)	300	60	4	50	150
Chinese hemlock (Tsuga chinensis)	400	80	4	100	150
Huashan pine (Pinus armandii)	200	35	2	30	100
Chinese pine (Pinus tabuliformis)	150	35	2	30	100
Liaodong oak (Quercus wutaishanica Mayr)	300	40	2	50	200
Brown oak (Quercus semicarpifolia)	250	40	2	50	200
Cork oak (Quercus variabilis)	200	20	3	50	300
Ring-cupped oak (Cyclobalanopsis glauca)	250	20	2	50	200
Minjiang cypress (Cupressus chengiana)	300	30	3	200	500
Locust (Sophora japonica)	150	15	1	300	1500
Chinese toon (Toona sinensis)	120	15	1	300	1000
Maple (Acer spp.)	200	20	3	50	200
Birch (Betula spp.)	150	15	1	200	1500
Aspen (Populus spp.)	150	15	1	300	1500
Alder (Alnus cremastogyne)	150	15	2	200	1000

Lng: Longevity; Mat: Maturity; Shd: Shade tolerance; EffD: Effective seeding distance; MaxD: Maxi mum seeding distance.

). The ANPP and SEP data of the tree species are derived from the calculation results of the PnET-II model.

2.2.5. Data Analysis

In this study, three future climate change scenarios (RCP2.6, RCP4.5, RCP8.5) and the standard climate (SC) were selected to simulate the 300-year dynamic successional changes of the alpine forests in western Sichuan without considering other disturbances, and the aboveground biomass of the forests in the study region (AGB) and the ratio of the maximum age of each species to the longevity of each species reflected the spatial and temporal dynamic changes of the ecological resilience of the forests. The LANDIS-II is a stochastic process-based model that simulates forest development on large spatiotemporal scales. To mitigate the model’s uncertainty, we adjusted the random seed parameters of the model to simulate the dynamic succession of forests in the study region during 2016–2316 under each scenario three times repeatedly. Multiple comparisons were used to analyze the significant differences between the standard climate and future climate change scenarios and between individual future climate change scenarios. Application ANOVA was used to verify significant differences in the effects of different climate scenarios on forest ecological resilience within each ecotone at three stages: short- (0–50 years), medium- (50–150 years), and long-term (150–300 years) periods.

Because the selected elasticity indicators have different magnitudes, it is necessary first to normalize the data and determine the weight coefficients of different elasticity indicators by using the coefficient of variation method, and finally apply the weighted summation method of ArcGIS to spatially superimpose the processed data, so that the forest ecological resilience indicators of each raster can be obtained in the end; the calculation formulas are as follows:

$$R_a = {\sum }_{i=1}^n{X}_i{Y}_i(i=1, 2, 3\dots \dots , n)$$

where $R_a$ represents the forest ecological resilience indicator in the grid, and its size represents the recovery ability of the forest in the region after external disturbance, with a range of [0, 1]. $X_i$ is the weight value of the ith ecological resilience indicator for the ecological resilience of the forest, and $Y_i$ is the standardized value of the specific indicator factor.

3. Results

3.1. Effects of Climate Change on Overall Forest Ecological Resilience

Under different climate scenarios, the ecological resilience of forests at the landscape level basically showed a consistent fluctuating trend over the simulation period (Figure 2), all of which showed a sharp upward trend in the first 50 years and a slow fluctuating upward trend in the 50–125 years. The ecological resilience of forests under the influence of the standard climate scenario reaches a peak of 0.97 at 130 years, while the remaining three climate change scenarios all reach a wave peak at 110 years of the simulation. The changes in forest ecological resilience under all four climate scenarios are relatively flat during the simulated 125–225 years, but the simulation shows a sharp downward trend from 225 years onwards. At the beginning of the simulation, the ecological resilience of forests under the RCP4.5 and RCP2.6 climate scenarios is slightly higher than that under the standard climate and RCP8.5 climate scenarios. At the mid to end of the short-term simulation (25–50 years), the changes and values of forest ecological resilience under three climate change scenarios are broadly comparable to those under the standard climate scenario. The impacts of four climate scenarios on forest ecological resilience gradually diverged during the mid-term simulations. From the 75th year of the simulation, the standard climate scenario’s level of forest ecological resilience gradually outpaces that of the other three climate change scenarios. During a later phase in the simulation (150–300 years), the differences in the ecological resilience of forests under the various climate scenarios become more significant, with SC > RCP2.6 > RCP4.5 > RCP8.5, and forests under the influence of SC having significantly more ecological resilience than those under the other three climate change impacts.

3.2. Climate Change’s Effects on Forests’ Ecological Resilience across Different Ecotones

The four forest ecotones in the research area responded differently to different climate scenarios (

Table 2. Dynamic changes of forest ecological resilience under different climate scenarios in different ecotones.

Simulated Scenarios		EBF	MEDBF	MTDF	SCF
S	SC	0.156	0.148	0.536	0.695
	RCP2.6	${0.123}^{ *}$	${0.103}^{ *}$	${0.507}^{ *}$	${0.627}^{ *}$
	RCP4.5	${0.127}^{ *}$	${0.101}^{ *}$	${0.509}^{ *}$	${0.548}^{ *}$
	RCP8.5	${0.142}^{ *}$	${0.106}^{ *}$	${0.527}^{ *}$	${0.598}^{ *}$
M	SC	0.748	0.817	0.338	0.059
	RCP2.6	${0.797}^{ *}$	${0.829}^{ *}$	${0.371}^{ *}$	${0.106}^{ *}$
	RCP4.5	${0.793}^{ *}$	${0.848}^{ *}$	${0.362}^{ *}$	${0.153}^{ *}$
	RCP8.5	${0.782}^{ *}$	${0.832}^{ *}$	${0.385}^{ *}$	${0.162}^{ *}$
L	SC	0.596	0.583	0.359	0.636
	RCP2.6	${0.644}^{ *}$	${0.614}^{ *}$	${0.635}^{ *}$	${0.615}^{ *}$
	RCP4.5	${0.646}^{ *}$	${0.621}^{ *}$	${0.671}^{ *}$	${0.680}^{ *}$
	RCP8.5	${0.630}^{ *}$	${0.624}^{ *}$	${0.671}^{ *}$	0.624

$*$ Significantly different between simulated climate change scenarios and SC climate scenarios (p < 0.05) S: 0–50a; M: 50–150a; L: 150–300a.

). Rates of change in forest ecological resilience dynamics under climate change scenarios were significantly different from SC in all forest ecotones in the short- and medium-term studies. In the long-term study, the dynamic rate of change of forest ecological resilience in the RCP8.5 scenario within SCF ecotone was not significantly different from SC, and the rest of the categories were all significantly different from SC. In the short-term simulation, the forest ecological resilience of the climate scenarios within the EBF and MEDBF ecotones changed less, with an average rate of change of 0.137 and 0.114, respectively. In contrast, the rate of change in forest ecological resilience within the MTDF and SCF ecotones was higher. The respective average rates of change were 0.520 and 0.617. In the medium-term simulation, the rate of change of forest ecological resilience was relatively low for each climate scenario in the MTDF and SCF ecotones, while the rate of change of forest ecological resilience was large for EBF and MEDBF ecotones. Finally, in the long-term simulations, the dynamic rate of change of forest ecological resilience was larger in all ecotones under different climate scenarios.

Different climate change scenarios significantly affected forest ecological resilience within different ecotones (Figure 3). In the short-term simulation (0–50 years), all three climate change scenarios significantly differed from SC’s forest ecological resilience. Still, a non-significant difference existed between RCP2.6 and RCP4.5 within the EBF ecotone. The forest ecological resilience of SC was lower than the resilience values of RCP2.6 and RCP4.5 in the SCF ecotone. In the mid-term simulation (50–150 years), except for the SCF ecotones where the forest ecological resilience values for the four climate scenarios were arranged in a stepwise order of SC > RCP2.6 > RCP4.5 > RCP8.5, the differences among forest ecological resilience values in the other three ecotones for the four climate scenarios were small. Except for the EBF ecotone, where non-significant differences were found between the forest ecological resilience numbers of RCP4.5 and those for SC and RCP2.6 within the EBF ecotone, the non-significant difference between RCP2.6 and RCP8.5, and a non-significant difference between RCP2.6 and RCP4.5 within the MEDBF ecotone, there was a significant difference between all climate scenarios. During long-term simulation (150–300 years), forest ecological resilience values of SC within the four ecotones were higher than those of other climate change scenarios, and the differences were larger within MEDBF, MTDF, and SCF, among which the most obvious performance was found in the MTDF and SCF ecotones, whose forest ecological resilience was higher than the lowest one by 0.273 and 0.191, respectively. The forest ecological resilience behaved as SC > RCP2.6 > RCP4.5 > RCP8.5, and significant differences existed between climate scenarios, except that there was a non-significant difference between RCP2.6, RCP4.5 and SC, and RCP8.5 within the EBF ecotone and non-significant difference between RCP2.6 and RCP8.5 within the MEDBF ecotone.

3.3. Forest Ecological Resilience under Four Climate Scenarios: Spatial and Temporal Dynamics

The spatial and temporal distribution of forest ecological resilience under different climate change scenarios varied significantly, showing an overall trend of first growing and then dropping, with a pattern of distribution where the west has a high and the east has a low (Figure 4). At the beginning of the simulation, the spatiotemporal distribution of forest ecological resilience at the landscape scale was similar under all four climate scenarios, with most of the forest’s ecological resilience values concentrated below 0.2. Still, some high forest ecological resilience values appeared in the study region’s central and western regions. The high values were mainly concentrated above 0.6. After 50 years of simulation, the forest ecological resilience under the four climate scenarios was increased as a whole, and the range of areas with higher values of forest ecological resilience was expanded, except for the eastern part of the study region, which had a certain distribution in the rest of the regions. Compared with the forest ecological resilience under the SC influence, the forest ecological resilience value under the RCP4.5 scenario increased in most areas. During the mid to late stages of the simulation, there is a significant increase in regions with forest ecological resilience values of 0.6–0.8 under different climate scenarios. The spatial distribution of forest ecological resilience values was mainly expanded in the study region’s central, western, and northern regions. The number of regions with low forest ecological resilience values under the influence of SC was the least compared with the three future climate change scenarios. When succession reaches a late stage, the ecological resilience of forests under the four climate scenarios declines significantly. The forest ecological resilience values were mainly concentrated below 0.2. The proportion of areas with values greater than 0.2 under the SC scenario accounted for 26.65%, which was the largest among the four climate scenarios. In contrast, the proportion of areas with values of ecological resilience of forests greater than 0.2 under the rest of the climate change scenarios accounted for less than 20%.

4. Discussion

Forest ecological resilience can well reflect the changes in forest ecosystems when subjected to external influences and is an important indicator for observing the self-regulation ability of forests [46]. In this paper, we effectively modeled the natural succession of forests under the standard climate scenario and three future climate scenarios using alpine forests in western Sichuan as the research object between 2016 and 2316 using a model at the landscape scale, and calculated spatiotemporal changes in forest ecological resilience, reflecting how climate change has affected the ecological resilience of forests. The findings suggest that future climate scenarios will significantly influence the ecological resilience of forests in the alpine forest region of western Sichuan. From the changes in the overall forest ecological resilience of the region (Figure 1), in the short-term simulation, the order of forest ecological resilience was RCP4.5 > RCP2.6 > SC > RCP8.5, which indicates that a certain amount of climate change would be beneficial in increasing the ecological resilience of the region’s forests, but when the simulation was carried out in the middle- and long-term, the future climate change scenarios might negatively affect changes to the ecological resilience of forests. Therefore, we believe that the appropriate warming in the short-term simulation will be beneficial to the accumulation of the research area’s forest’s entire aboveground biomass and the increase in the ratio of the maximum age of each species to the longevity of each species in the area. However, in the medium- and long-term simulations, the excessive temperature increase will exceed the optimal light and temperature levels of the tree species [47], which will adversely affect the forest’s aboveground biomass and the living conditions of the maximum tree age possible for each species of tree and ultimately lead to a decrease in the forest’s ecological resilience value. Overall, the ecological resilience of the forests in the study region is higher in the simulation period of 50–130 because the forests in the study region suffered from massive logging, overgrazing, and forest fires at the end of the 20th century, which resulted in serious damage to the forest ecosystems. As the forest succession continues, the forest ecosystem is gradually rebuilt, the forest aboveground biomass (AGB) increases, and the ratio of the maximum age of each species to the longevity of each species increases, which will directly lead to an increase in the ecological resilience of the forest. In the process of succession, the pioneer species of Birch, Locust, Aspen, etc., which are less shade-tolerant in the study region, will be gradually replaced by other species that are more shade-tolerant in the study region.

The changes in forest ecological resilience within each ecotone were not quite the same at different simulation stages. In the short-term simulation, the rate of change of forest ecological resilience within MTDF and SCF is higher than that of the other two ecotones, and this is because the main tree species types within these two ecotones are coniferous forests, and the major dominant species in the research area are the Fir and Spruce, which accounts for more than 65% of the study region’s forest land, resulting in a strong response to forest ecological resilience in short-term simulations and a significant increase in forest ecological resilience. For this reason, the ecological resilience rate of change of forests in the MTDF and SCF ecotones was lower than that of the EBF and MEDBF in the middle stage of the simulation. As forest succession continues, in the later stages of the simulation, biomass growth is slow and the ratio of maximum age to the longevity of each tree species decreases. The ecological resilience of the forests in the study region shows an overall decreasing trend, so the rate of change in the ecological resilience of the forests in the four ecotones is larger.

Regarding the spatial distribution of forest ecological resilience, the high-value areas of forest ecological resilience in the short-term and medium-term simulations are mainly concentrated in the MTDF and SCF areas in the study region’s central, western, and northern parts. This is due to the fact that the concentration of the main tree species of the research area, its aboveground biomass, and the ratio of the maximum age of each species to the longevity of each species are high, so the values of forest ecological resilience were high in the simulations. At the later stage of the simulation, the ecological resilience of forests in the whole study region decreased significantly, and only in the research area’s western region were there some areas with higher values. We found that some areas with higher values of ecological resilience appeared stably in the western part of the research area at higher elevations in different periods and under the influence of different climatic scenarios. Therefore, we randomly selected an area a in the high-value area and area b in the central part of the research area for comparison and viewed the dynamics of the continuity of forest ecological resilience on a landscape scale in area ab during the 300-year period (Figure 5). The value of forest ecological resilience in the region a kept fluctuating around 0.8 in the short- and medium-term simulations, and then started to decline in the long-term simulations, but still remained above 0.6. On the other hand, the ecological resilience of forests in region b is lower than that of the region and starts to decrease in the medium-term simulation, and the ecological resilience of forests under the influence of various climate change scenarios in the long-term simulation is below 0.4. This may be due to the fact that area a is located at a higher altitude, which is suitable for the survival of the dominant species of Fir and Spruce in the region, and that the ecological resilience of the forest in this area is higher at the beginning of the simulation. This in turn may be due to the fact that the forests in this area have been subjected to relatively low harvesting in recent decades and that they are rich in internal age gradient and species composition, the biological community is relatively stable, and the anti-interference ability is relatively strong.

In this study, based on selecting relevant resilience indicators and estimating forest ecological resilience, we measured the effects of different climates scenarios on the ecological resilience of alpine forests within western Sichuan, China, at the landscape scale using the LANDIS-II (forest landscape model). Because of the variability of forest ecological resilience indicators and the complexity of the climate system, this paper only analyzes the trend in the ecological resilience of forests under scenarios of future climate change in terms of aboveground biomass and the ratio of maximum age to the longevity of each species and is not an accurate prediction of forest ecological resilience in this region. Moreover, this study was carried out without considering other factors affecting forests, such as deforestation, pests and diseases, forest fires, etc. However, in actual forest succession, anthropogenic deforestation will change the natural state of forests [48], pests and diseases will change the distribution area of species in forests [49], and forest fires will change the composition and structure of forests [50]. Therefore, we will synthesize the various influencing factors to conduct in-depth research in the following work.

5. Conclusions

In this study, we simulated the natural succession of forests at the landscape scale for standard and future climate scenarios by combining the LANDIS-II forest landscape model and the PnET-II ecosystem process model from 2016 to 2316 and spatiotemporal changes in the ecological resilience of alpine forests in western Sichuan were calculated. Our simulation and calculation results showed that: only within the short-term simulation, certain Alpine forests in western Sichuan will be more ecologically resilient due to climate change, while the adverse impacts of climate change scenarios on the ecological resilience of the area increase as the simulation time increases; meanwhile, with an increase in simulation time, climate change scenarios will have increasingly negative effects on the ecological resilience of regional forests, and the intensity of the negative impacts is shown as RCP8.5 > RCP4.5 > RCP2.6. On the other hand, since the main tree species of the study region are Fir and Spruce, the ecological resilience of the MTDF and SCF ecotones where they are distributed has a more significant variation, indicating that the protection of the dominant tree species of the region is effective in improving the ecological resilience of the study region in the event of climate change disturbances.