Conservation Gap Within a Nature Reserve: A Case Study in the Biodiversity Hotspot of Rare and Endangered Plants at the Southern Gaoligong Mountains
高黎贡山南段稀有濒危植物保护空缺分析
Editor-in-Chief & Handling Editor: Binbin Li
ABSTRACT
enBiodiversity loss significantly impacts the stability and functioning of ecosystems, driven by factors such as climate change, human activities, and other influences. Predicting species distributions and conducting conservation gap analyses are essential for identifying key conservation areas, evaluating conservation effectiveness, and highlighting under-protected areas, thereby providing valuable insights to mitigate biodiversity loss. The Gaoligong Mountains, located within the Three Parallel Rivers of Yunnan Protected Areas (a UNESCO World Natural Heritage Site), serve as an intersection for the biota of the eastern Himalayas, the Indochina Peninsula, and the Hengduan Mountains. The region is a vitally important biological exchange corridor and a long-standing hotspot for plant geography research. Using the distribution data for 361 rare and endangered plants and eight environmental variables, this study employed the InVEST and random forest (RF) models to determine the current and future potential distribution of these species in the southern Gaoligong Mountains, considering various climate change scenarios combined with currently suitable habitats. The study also evaluated existing protected areas and used a transition matrix to quantify changes in potential habitats. The results show that the current potential suitable habitat in the southern Gaoligong Mountains spans 2987.38 km2, with only 23.14% of this area falling within protected zones, leaving a 76.86% protection gap. Additionally, areas with highly suitable habitats are predicted to decline under future climate change scenarios, emphasizing the inadequacy of existing protected areas in comprehensively safeguarding rare and endangered wild plant species. Habitat loss, primarily driven by the expansion of agricultural activities, further exacerbates this issue. To address these conservation gaps, this study recommends prioritizing the integrity of vertical zones when optimizing protected areas, thus ensuring continuous protection for rare and endangered plant species.
摘要
zh生物多样性丧失对生态系统的稳定性和功能产生了重要影响, 其原因包括气候变化、人类活动及其他因素。通过预测物种分布和开展保护空缺分析, 可以帮助确定关键区域, 评估保护措施的有效性, 并识别保护不足的区域, 从而更有效地为减少生物多样性丧失提供科学依据。位于三江并流地区的高黎贡山, 是东喜马拉雅、印度支那半岛和横断山生物区交汇的重要枢纽, 同时也是植物地理学研究的长期热点区域。我们结合361种稀有濒危植物的分布数据和8个环境变量, 利用InVEST和随机森林模型, 分析了高黎贡山南部361种稀有濒危野生植物在当前和未来气候变化情景下的潜在分布及当前适宜生境。我们评估了现有保护区, 并通过转移矩阵量化潜在生境的变化。结果表明, 高黎贡山南部当前的潜在适宜生境面积为2987.38平方公里, 其中仅有23.14%位于保护区内 (意味着有76.86%的区域存在保护空缺) 。此外, 高适宜性区域的面积在气候变化情景下将减少。这突显出现有保护区在为这些稀有濒危野生植物提供全面保护方面的不足。农业活动的扩张成为栖息地丧失的主要驱动因素。我们建议在优化保护区时, 优先考虑垂直植被带的完整性, 以确保稀有濒危植物保护的连续性和有效性。 【翻译：杨勇婧雯和谭运洪】
Plain Language Summary
enBiodiversity loss has profound effects on the stability and functioning of ecosystems, driven primarily by climate change and human activities. To improve the protection of endangered plants, we predicted species distributions and conducted a conservation gap analysis to identify critical areas and evaluate the effectiveness of current conservation efforts. Focusing on the southern Gaoligong Mountains, part of the Three Parallel Rivers of Yunnan Protected Areas, we studied the distribution of 361 rare and endangered plants. Using the InVEST and Random Forest models, we analyzed the current and future potential habitats of these plants under various climate scenarios. Our findings show that 2987.38 km² of suitable habitat currently exists in the southern Gaoligong Mountains, yet only 23.14% of this area falls within protected zones, leaving a substantial 76.86% conservation gap. Furthermore, climate change is predicted to reduce the extent of highly suitable habitats, while agricultural expansion is identified as the primary cause of habitat loss. To address these issues, we recommend prioritizing the integrity of the mountain region's vertical vegetation zones when optimizing protected areas to better conserve rare and endangered plant species.
简明语言摘要
zh生物多样性减少对生态系统的稳定性和功能有重要影响, 主要原因包括气候变化和人类活动等。为了更好地保护濒危植物, 我们通过预测物种分布和分析保护空缺, 找到了关键区域, 并评估了现有保护措施的有效性。在三江并流地区的高黎贡山南部, 我们研究了361种稀有濒危植物的分布, 结合气候变化预测, 利用InVEST和随机森林模型分析了它们的潜在栖息地情况。研究发现, 目前高黎贡山南部约2987.38平方公里的适宜生境中, 仅有23.14%位于保护区内, 这意味着仍有76.86%的区域缺乏保护。此外, 气候变化将导致高适宜性栖息地面积减少, 而农业活动则是导致栖息地丧失的主要原因。我们建议在优化保护区时, 优先考虑高黎贡山垂直植被带的完整性, 以更好地保护这些珍稀濒危植物的生存环境。
Summary
en
-
Currently, 76.86% of suitable habitats in the southern Gaoligong Mountains lie outside protected areas, highlighting the urgent need to expand conservation coverage.
-
Highly suitable habitats are projected to decline under future climate scenarios, with agricultural expansion emerging as the leading driver of habitat loss, emphasizing the need for stricter controls on agricultural activities in key ecological regions.
-
Conservation planning should prioritize the integrity of the mountains' vertical vegetation zones to ensure habitat connectivity and stability for rare and endangered plant species.
实践者要点
zh
-
高黎贡山南部当前的适宜栖息地中, 76.86%的区域未被保护区覆盖, 需要增加保护区面积。
-
未来气候变化将导致高适宜性栖息地减少, 农业扩张是栖息地丧失的主要驱动因素, 应限制农业开发对关键生态区域的侵占。
-
建议在规划和调整保护区时, 注重保护高黎贡山垂直植被带的完整性, 以确保稀有濒危植物栖息地的连贯性和稳定性。
1 Introduction
Biodiversity loss is recognized as one of the three major environmental concerns (UNEP 2021), and over the past 50 years, global biodiversity has experienced significant declines (IPBES 2019). Tropical forests, which sustain at least two-thirds of the world's biodiversity (Raven 1988), have been particularly affected by deforestation and forest degradation driven by human activities such as logging, hunting, agricultural expansion, and urbanization (Giam 2017). These pressures are illustrated by sharp declines in wildlife populations worldwide, such as global populations of amphibians, birds, fish, mammals, and reptiles, which decreased by an average of 68% between 1970 and 2016 (WWF 2016).
Human-driven large-scale tropical forest development has resulted in habitat loss of significant population declines for approximately 40% of the world's plant species (Newbold et al. 2015). In the Brazilian Amazon, numerous rare species face extinction due to habitat destruction (Feeley and Silman 2009). In response to these challenges, the Convention on Biological Diversity (CBD) established the Aichi Targets to safeguard and restore global biodiversity through the development of cooperative policies and strategies (CBD 2020). However, despite the strengthening of conservation measures at the global level, many countries struggle to meet these targets, and biodiversity loss continues to escalate at regional and national levels (IPBES 2019). In December 2022, the United Nations Biodiversity Conference (COP15) adopted the Kunming-Montreal Global Biodiversity Framework, a pivotal initiative aimed at reversing biodiversity loss and restoring ecosystems.
Among the many factors influencing biodiversity, including land use changes and human activities, climate change is considered one of the most significant threats (Li et al. 2018). It drives habitat shifts and increases the frequency of extreme climatic events, posing substantial challenges to species' adaptability (Loarie et al. 2009). Studies have shown that many species have migrated to polar regions or high-altitude areas in response to climate change—a trend expected to persist in the coming decades (Parmesan and Yohe 2003). However, traditional biodiversity conservation strategies often fail to account for the rapid impacts of climate change on species distribution, exacerbating extinction risks for wild plants at regional levels (Araújo et al. 2004; Araújo and Rahbek 2006). Research indicates that over a million terrestrial species could face extinction within the next 50 years (Thomas et al. 2004).
In addition to climate change, human activities, such as agricultural expansion, deforestation, and urbanization, have further degraded natural habitats (Corlett and Westcott 2013; Mohammed and Behailu 2019; Mahmoud and Gan 2018; Scanes 2018; Zhao et al. 2006), accelerating biodiversity loss (Finlayson et al. 2005; Pereira, Navarro, and Martins 2012). According to the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) assessment, approximately 25% of animal and plant species are currently threatened, with nearly 1 million species on the verge of extinction (IPBES 2019). In India, for instance, the Western Ghats, a renowned biodiversity hotspot, has suffered extensive habitat destruction due to the expansion of tea and coffee plantations and illegal logging. Consequently, many endemic plant species in the region are now critically endangered (Das et al. 2006).
Conservation areas are the key to safeguarding biodiversity (Dudley 2008). They provide safe habitats for rare and endangered species (McNeely 2020) and mitigate the impacts of human activities on these species and their ecosystems, effectively preventing species extinction and maintaining ecological balance (Du, Fathollahi-Fard, and Wong 2023). However, climate change-induced shifts in species distributions present significant challenges to such protected areas. In the future, more species may relocate to unprotected areas due to unsuitable climatic conditions within currently protected zones, heightening extinction risks (Jenkins, Pimm, and Joppa 2013). The species that lose viable habitats within protected areas may face greater extinction threats, underscoring the limitations of static conservation strategies (Wessely et al. 2017). Therefore, transformative measures, such as expanding the global network of protected areas, have become increasingly urgent (IPBES 2019). Historically, biodiversity conservation priorities have mainly focused on species-based approaches at large spatial scales, often overlooking localized, targeted analyses that identify specific species or habitats for conservation (Brooks et al. 2006; Pimm, Jenkins, and Li 2018; Yang et al. 2019). Therefore, carrying out regional biodiversity conservation planning rooted in geographical perspectives must be made a priority.
The Gaoligong Mountains, situated within one of the world's key biodiversity hotspots in southwestern China (Myers et al. 2000), form the core area of the UNESCO World Natural Heritage Site, the Three Parallel Rivers of Yunnan Protected Areas (Li and Qi 2017). It serves as a corridor for biota exchange between the Qinghai-Tibet Plateau and the Indo-China Peninsula, with a distinct north–south alignment (Xiao et al. 2023). Its unique geographical location and diverse climatic conditions have established the Gaoligong Mountains as an essential genetic repository for subtropical, temperate, and cold temperate plants, hosting a rich array of biodiversity (Li and Guo 2000; Yi et al. 2021).
Previous studies have predominantly focused on plant diversity in the northern Gaoligong Mountains (Li et al. 2007; Li, Dao, and Li 2011). However, despite its designation as a nature reserve in 1983, the region, particularly the southern Gaoliogong Mountains, continues to face escalating threats. Increasing land use intensity has profoundly impacted wild plant habitats, leading to severe habitat degradation and posing direct risks to endangered plant species (Bellard, Cassey, and Blackburn 2016; Oliver and Morecroft 2014; Shearman, Bryan, and Laurance 2012; Venter et al. 2016). While some studies have investigated the seed plant flora and plant communities in the southern region of the Gaoligong Mountains (Schweinfurth and Weilie 1984), there is a notable lack of research addressing the specific impacts of climate change and human activities on endangered plants. This research gap hinders the development and implementation of effective biodiversity conservation strategies for this vulnerable area.
In this context, this study aims to evaluate the current and future impacts of climate change and human activities on the potential habitats of endangered plants in the southern Gaoligong Mountains (2041–2060) using species distribution and InVEST models. A conservation gap analysis was performed to compare simulated biodiversity hotspots with existing protected areas, assessing their effectiveness in safeguarding established conservation zones. The study's specific objectives are (1) to investigate the potential shifts in the distribution of rare and endangered wild plant species in the southern Gaoligong Mountains under current and projected climate scenarios (2041–2060) and (2) to evaluate the adequacy of existing protected areas in safeguarding the potential habitats of rare and endangered wild plant species. The study also aims to provide critical insights and actionable guidelines to preserve these unique, fragile, and globally significant biodiversity hotspots amidst ongoing global change.
2 Materials and Methods
2.1 Research Area
The southern Gaoligong Mountains (98°34′–98°50′ E, 24°56′–26°09′ N) are located in western Yunnan Province, encompassing the northern part of Longling County, the entirety of Tengchong City, and the western part of Longyang District (Li and Li 2020; Liang et al. 2023). Nestled in the southwest of the Yunnan-Guizhou Plateau, the region forms the southern edge of the Hengduan Mountains and is bordered by Myanmar to the northwest and south (Figure 1). Elevations in the area range from 540 m in the southeast to 3722 m in the northwest, boasting a relative elevation variation of 3182 m (Ren 2011). This area is characterized by a lowland mountainous subtropical monsoon climate (Li et al. 2023). The western slopes, including Tengchong County and Longling County, are influenced by warm and humid airflows from the southwest, resulting in higher precipitation, which makes this area the wettest in the region. In contrast, the eastern slopes, sheltered by the Gaoligong Mountains, receive comparatively less rainfall (Sun et al. 2011). Due to this difference, the western slopes harbor a greater abundance of endemic seed plants than the eastern slopes (Yang et al. 2016). The region's distinctive geographical, geological, and topographic features, combined with its unique landforms, contribute to a diverse climate and ecological environment. This makes it a renowned area for studying vegetation-climate interrelations.
2.2 Species Distribution Model Prediction
2.2.1 Species Occurrence Data
The selection of endangered plant species was based on the following criteria: (1) the List of National Protected Key Wild Plants (2021) issued by China's State Forestry Administration, Grassland Bureau, and Ministry of Agriculture and Rural Affairs, which includes national first- and second-level protected wild plants; (2) the List of Plant Species with Extremely Small Populations in Yunnan Province (2022), jointly published by the Yunnan Provincial Forestry and Grassland Bureau, the Provincial Department of Agriculture and Rural Affairs, and the Provincial Department of Science and Technology; (3) the Implementation Plan of Rescuing and Conserving China's Wild Plants with Extremely Small Populations (2011–2015), released by the State Forestry Administration, delineating 120 species of wild plants with extremely small populations; (4) endemic species documented in Plant resources and geography of Gaoligong Mountain (2020) by researcher Heng Li and her team from the Kunming Institute of Botany, Chinese Academy of Sciences; and (5) rare and endangered endemic plants featured in Comprehensive Scientific Research on Gaoligong Mountain in Yunnan (2023) by Jie Zhou's research team from the Kunming Branch of the Chinese Academy of Sciences.
In accordance with the biodiversity characteristics of the study area and relevant species protection lists, we conducted research using data from the Chinese Virtual Herbarium (CVH, https://www.cvh.ac.cn/), the Global Biodiversity Information Facility (GBIF, https://www.gbif.org/), and the China National Specimen Information Infrastructure (NSII, http://www.nsii.org.cn) to obtain verified species distribution data. In addition, field surveys were conducted in September 2021 and May 2022. This comprehensive effort yielded 1485 records of 361 plant species (Table S1).
To minimize spatial sampling bias, we implemented sampling strategies based on the actual collection environment of the study area, ensuring that the coordinates of sampling locations for each species were at least 1 km apart. After this process, 908 records were retained for subsequent analyses (Table S2).
2.2.2 Environmental Variables
The study used 19 bioclimatic variables and three topographic variables (slope, aspect, and altitude). The bioclimatic variables under current and future climate conditions were sourced from WorldClim version 2.1 (http://www.worldclim.org) at a resolution of 30 arc-seconds. These variables represent climate averages for the period 1970–2000.
The topographic variables—elevation, slope, and aspect—were derived from a digital elevation model (DEM) obtained from the Geospatial Data Cloud (http://www.gscloud.cn/), also at a spatial resolution of 30 arc-seconds. ArcGIS was employed to extract these terrain variables.
Spatial autocorrelation between environmental variables can complicate the analysis of the relationship between species distribution and environmental factors (Wan et al. 2020). To minimize the influence of multicollinearity, Pearson rank correlation analysis was conducted using the “findCorrelation” function in the “caret” package in R (Kuhn 2008). Variables exhibiting a Pearson correlation coefficient greater than 0.7 were excluded (Dormann et al. 2013) (Table S3). Following this analysis, nine environmental variables were retained for modeling: altitude, aspect, slope, bio3 (isothermality), bio4 (temperature seasonality), bio6 (minimum temperature of the coldest month), bio7 (annual temperature range), bio12 (annual precipitation), and bio14 (precipitation of the driest month).
2.2.3 Model Selection and Validation
After processing the species distribution and environmental variable data for the study area, the “sdm” package in R was used for model selection, ecological niche modeling, and simulating potential distribution areas for predictive analysis (Naimi and Araújo 2016). This software package supports the application of 21 models, including the random forest (RF) model, maximum entropy (MaxEnt) model, boosted regression tree (BRT) model, classification and regression tree (CART) model, generalized additive model (GAM), and generalized linear model (GLM), among others. Model performance was assessed using 75% of the distribution data as the training set and the remaining 25% as the test set, with 10 repetitions and cross-validation employed as the run type.
The area under the receiver operating characteristic (AUC-ROC) curve was used to evaluate model quality. The Jackknife test was used to assess the importance of environmental variables (Phillips and Dudík 2008). A higher AUC value signifies greater model prediction accuracy, while true skill statistic (TSS) values exceeding 0.6 are considered excellent (Komac et al. 2016). Based on these metrics, the Random Forest model exhibited the highest AUC and TSS values (0.94 and 0.76, respectively) and was selected for subsequent species distribution model analyses (Table S4). Bio12 and bio7 were identified as the most influential variables contributing to the model (Table S5). ArcGIS was then used to process the model output.
Potentially suitable areas were reclassified using the maximum TSS threshold: values below the presence threshold of 0.36 were categorized as 0 (unsuitable), while values above 0.36 were categorized as 1 (suitable) (Peng et al. 2023). This method, widely recognized for its effectiveness and conservatism, is frequently applied to differentiate suitable from unsuitable areas, aiding the identification of critical conservation zones (Huang et al. 2020). The resulting predictions were further classified into four distinct categories of potential habitats: “unsuitable” (< 0.36), “low-suitability” (0.36–0.5), “medium-suitability” (0.5–0.75), and “high-suitability” (> 0.75) (Anand, Oinam, and Singh 2021).
2.2.4 Predicting Potential Habitat Under Future Climate Change
To evaluate the potential impact of future climate changes on endangered plants in the southern Gaoligong Mountains, we used the RF model to predict their potential habitat for the period 2041–2060. Future climate projections were downscaled from the CMIP6 data set of the Sixth Assessment Report (AR6) by the Intergovernmental Panel on Climate Change (IPCC) using the EC-Earth3 Earth System Model (EC-Earth3-Veg). EC-Earth3-Veg was selected for its superior performance among global climate models, as evaluated by GCMEval, a model selection tool tailored to the study area (Parding et al. 2020). In addition, two shared socioeconomic pathways (SSPs), SSP2-4.5 and SSP5-8.5, representing moderate and high emission scenarios, respectively, were used to predict suitable areas for all studied species. These projections were then compared with the current hotspot layer to assess changes in potential habitat suitability.
2.3 InVEST Prediction
Habitat quality refers to an ecosystem's ability to provide suitable conditions for the survival and development of individual organisms and populations, forming the foundation for maintaining biodiversity (Wang and Cheng 2022). Species distribution models often overlook the actual state of land use and cover, which can lead to overpredictions. The integrated valuation of ecosystem services and tradeoffs (InVEST) model, developed by Stanford University, the World Wildlife Fund, and The Nature Conservancy, provides a suite of tools for assessing ecosystem service (Nelson et al. 2009). This model addresses exaggerated outputs by incorporating land use vulnerability into habitat quality assessments, particularly when monitoring data are lacking (Gong et al. 2019; Moradi et al. 2023).
In this study, the habitat quality module of the InVEST model was used to calculate habitat quality within the study area. This model evaluates the quality of diverse habitat patches by estimating the intensity of human interference in habitats from various threat sources (Zhang and Li 2022). Specifically, the InVEST model integrates four primary factors to evaluate habitat quality: the relative impact of each threat source, the relative sensitivity of each habitat type to these sources, the distance from natural habitats to threat sources, and the degree of legal protection of grid cells (Tallis et al. 2011).
Based on the specific conditions of the study area, relevant literature, and expert opinions, the identified threat sources include cropland, towns, residential areas, other construction land, and bare rock land (Huang et al. 2020; Zhao and Si 2022). The sensitivity of each land use type to stress factors was determined and is summarized in Table S6. Land use/cover data were sourced from the Resource and Environmental Science and Data Center of the Chinese Academy of Sciences (https://www.resdc.cn/). The habitat quality calculation formulas used in the InVEST model are provided in Appendix Formula (1) and Formula (2). The model output values range from 0 to 1, with higher values indicating better habitat quality.
2.4 Land Use Transfer Matrix
The land use transfer matrix method employs a two-dimensional matrix based on the conversion relationships of land use conditions over different periods. This method provides a clear representation of the spatiotemporal changes in land use (Shi, Jiang, and Yao 2018) and facilitates the calculation of gains (area increase for a specific land use type during the study period) and losses (area decrease for a specific land use type during the study period). Given the extensive classification system in the original data, this study reclassified land use/cover into nine categories: cropland, woodland, shrubs, grassland, waterbody, wetland, building, bare land, and permanent snow and ice, as per the requirements of the research. The specific reclassification principles are detailed in Table S7.
2.5 Gap Analysis
Gap analysis involves overlaying species distribution data with natural protected areas within the study area to evaluate the current conservation status and determine whether conservation priorities are being met (Zengxiang et al. 2011). Protected area data were provided by personnel from the Gaoligongshan National Nature Reserve and Longling Xiaoheishan Provincial Nature Reserve.
To effectively analyze the relationship between ecological niches and habitat quality of biodiversity, the habitat quality evaluation results from the InVEST model and species richness data from the RF model were integrated as indicators of biodiversity. First, the InVEST model results were aligned with the RF model's corresponding regions. Then, the “Raster Calculator” and “Reclassify” tools in ArcGIS were used to integrate and categorize these datasets. The integrated results were then compared with the existing boundaries of protected areas in the southern Gaoligong Mountains. Areas identified as both ecologically significant and underrepresented in protection were designated as “core areas” and prioritized for conservation efforts (Huang et al. 2020).
3 Results
3.1 Status and Potential Habitats of Rare and Endangered Plants
Our results indicate that suitable habitats for endangered plants comprise 21.74% of the total area in the southern Gaoligong Mountains. Species distribution hotspots are predominantly situated in the central and northwest regions of the study area, encompassing a combined area of 2987.38 km2. Within these suitable habitats, 14.91% of the area is categorized as high suitability, primarily distributed along ridges and at higher altitudes. Unsuitable areas are scattered across the eastern and southeastern sections of the study area.
The total potential habitat for rare and endangered plants is projected to increase. In the 2041–2060 SSP2-4.5 scenario, the hotspot area in the southern Gaoligong Mountains is predicted to expand to 4036.93 km2, an increase of 35.13% from the current period. However, the highly suitable zone is expected to decrease by 72.16% (321.37 km2) compared to the current period. Similarly, under the 2041–2060 SSP5-8.5 scenario, the highly suitable area is predicted to decrease by 270.45 km2 relative to the current period (Table S8).
Future distribution hotspots are anticipated to follow patterns similar to those observed currently, primarily concentrated in the central and northwest areas of the study area. Notably, distribution hotspots in the northwest are expected to transition from a scattered to a more aggregated state. Overall, endangered plant hotspots in the southern Gaoligong Mountains exhibit a northward shift to higher altitudes, becoming increasingly prominent and concentrated (Figure 2).
3.2 Integrated Biodiversity Pattern
According to the RF model output, habitat hotspots are mainly concentrated in the western region, covering a total area of 2987.38 km2 (Figure 3). Conversely, the InVEST model identifies habitat hotspots primarily in the central area, followed by the northwest, with a total area of 7675.18 km2. In both models, scattered distributions are observed in the eastern and southeastern areas.
The comprehensive distribution map indicates that suitable biodiversity hotspot distribution areas identified by the RF and InVEST models cover 1044.64 and 5360.75 km2, respectively, accounting for 8% and 39% of the study area. The final core area of hotspot distribution covers 14% of the total study area, encompassing 1942.74 km2.
3.3 Land Use/Cover Changes
Over four periods spanning 10 years each (1980–1990, 1990–2000, 2000–2010, and 2010–2020), the single land use/cover dynamic degree was calculated. As shown in Table 1, the “building” category exhibits the most pronounced and rapid changes across the four stages, with consistently high change rates of changer. The period from 2010 to 2020 recorded the highest rate of change, reaching 12.36%. Other land use/cover types experiencing rapid alterations include cropland, waterbody, and bare land. Among these, cropland exhibited the fastest rate of change between 2010 and 2020, reaching 1.4%. Meanwhile, grassland experienced the most rapid change rate during the same period, reaching −1.93%. Waterbody showed high dynamics during the earlier periods, with rates of −2.15% between 1980 and 1990% and 3.21% between 1990 and 2000.
| Land use/cover type | 1980–1990 | 1990–2000 | 2000–2010 | 2010–2020 |
|---|---|---|---|---|
| Cropland | −0.05 | −0.04 | −0.04 | 1.40 |
| Woodland | 0.07 | −0.14 | −0.03 | 0.03 |
| Shrubs | −0.04 | −0.43 | 0.46 | 1.20 |
| Grassland | −0.02 | 0.31 | −0.12 | −1.93 |
| Waterbody | −2.15 | 3.21 | 0.00 | 1.26 |
| Wetland | 0.00 | 0.00 | 0.00 | — |
| Building | 1.30 | 2.31 | 0.94 | 12.36 |
| Bare land | −10.00 | — | 0.00 | −4.00 |
The primary changes in land use/cover types in this study involved the conversion of forests due to human disturbances. Forests include woodlands, shrubs, and grasslands, with human activities predominantly converting these areas into cropland or buildings. As shown in Figure 4, a significantly larger proportion of forests was converted to cropland compared to the smaller fraction transformed into artificial structures. Most forest areas were converted into cropland or artificial structures from 2010 to 2020, accounting for approximately 10% of the study area. In contrast, forest conversion was less pronounced during the 2000 to 2010 period (Table 2).
| Conversion type | 1980–1990 | 1990–2000 | 2000–2010 | 2010–2020 |
|---|---|---|---|---|
| Forests → Cropland | 1172.92 | 1170.14 | 6.25 | 1418.06 |
| Forests → Building | 61.81 | 77.08 | 8.33 | 179.17 |
3.4 Identification of Conservation Priorities
A comparison of existing protected area boundaries with the hotspot distributions derived from the two models reveals the following findings. The “core areas outside protected areas (PA)” are mainly situated in the western and central areas of the southern Gaoligong Mountains, totaling 1521.94 km2, which constitutes 16% of the suitable habitat area for species. The “suitable habitat area outside PA” spans 7450.86 km2 in total, of which the “RF outside PA” covers 1014.35 km2, according to the RF model, while the “InVEST outside PA” accounts for 4914.57 km2, as indicated by the InVEST model. Habitat outside the protected area comprises 76.86% of the species' suitable habitat. Within the existing protected area, the total area of overprotected zones (areas within the protected area but not identified as suitable habitat by the model) is 35.20 km2, representing 5% of the protected area (Table S9) (Figure 5).
The altitude of the study area was divided into three gradients: low (< 1000 m), medium (1000–2000 m), and high (> 2000 m) (Yi et al. 2021). The degree of protection of suitable habitats for rare plants was assessed for each gradient. The results show that suitable areas for rare plants in the high-altitude zones of the reserve account for 86.48% of the total suitable area within the reserve, which is significantly higher than the 13.52% in the medium- and low-altitude gradients (Table 3). While the protection of suitable habitats in high-altitude areas was relatively adequate, the significantly lower protection ratios in low- and medium-altitude areas suggest that there is a substantial protection gap in these altitude ranges.
| Altitude gradient | Suitable area in PA (km2) | Total suitable area in the study area (km2) | Percentage of the suitable area covered by PA |
|---|---|---|---|
| Low-altitude area (< 1000 m) | 4.17 | 39.30 | 10.60% |
| Medium-altitude area (1000–2000 m) | 55.00 | 1725.79 | 3.19% |
| High-altitude area (> 2000 m) | 378.33 | 1201.01 | 31.50% |
4 Discussion
4.1 Conservation Gap in the Southern Gaoligong Mountains
The Gaoligong Mountain Nature Reserve plays a pivotal role in safeguarding the area's forested landscapes and local biodiversity. However, during the establishment of the reserve, the specific needs and distribution of rare, endemic, and endangered plant species were largely overlooked. Our analysis reveals that only 8.56% of hotspots for these species are currently encompassed by protected areas in the southern Gaoligong Mountains. Additionally, core priority areas outside the protected boundaries account for 16% of suitable habitats, representing 11% of the total area in this region. These findings indicate a significant conservation gap in the southern Gaoligong Mountains.
The lack of adequate information on endangered plants is a major contributor to this conservation gap. In this study, the prediction accuracy of the RF model relies heavily on the availability of actual species distribution points and environmental data. The rugged terrain and dense forests of the southern Gaoligong Mountains present substantial challenges for acquiring comprehensive conservation data, resulting in a scarcity of ecological information on threatened species (Yang et al. 2019). Furthermore, most rare and endangered plants have limited populations, making it difficult to obtain sufficient, accurate, and reliable species distribution points for use in simulations. The geographical distribution data used in this study were primarily derived from herbarium specimen information, with limited field survey data. This scarcity of data poses significant challenges to achieving comprehensive species representation (Ye, Zhang, and Wu 2020). Such limitations can affect simulated outcomes and contribute to a discrepancy between protected area boundaries and actual biodiversity hotspots (Yang, Zhou, et al. 2014). Addressing this gap necessitates prioritizing data supplementation, particularly in under-surveyed areas, to accurately identify key areas for future biodiversity conservation (Ye, Zhang, and Wu 2020).
Moreover, the majority of the Gaoligong Mountain Nature Reserve is located at more than 1600 m above sea level. Our results show that only 13.79% of suitable habitats for rare and endangered plant species within the reserve are located in low- and medium-altitude areas. This highlights the inadequate coverage of endangered plant habitats at these elevations. A similar situation has been observed in the Sierra Madre Mountains in Mexico, where low-altitude tropical dry forests, despite being highly biodiverse, are often neglected due to land development and agricultural activities, and existing protection measures are mainly concentrated in high-altitude areas (Valiente-Banuet et al. 2015).
The Gaoligong Mountain Nature Reserve was initially established as a provincial reserve in 1983 and was upgraded to a national reserve in 1986 (Allendorf and Yang 2013). However, the early stages of its development were characterized by a lack of cohesion and systematic planning (Xu et al. 2017). Similarly, the early planning of the Changbai Mountain Nature Reserve was also primarily focused on high-altitude forest areas, neglecting the protection needs of low-altitude forests and wetlands, which consequently failed to protect the biodiversity in these areas effectively (Yang and Xu 2003).
The Three Parallel Rivers of Yunnan Protected Area, which includes the Gaoligong Mountains, is one of the three recognized biodiversity hotspots in Yunnan province, featuring significant altitude variations and distinct vertical vegetation zones. These include tropical monsoon forests and subtropical evergreen broad-leaved forests at lower elevations, mixed coniferous and broad-leaved forests and coniferous forests at mid-elevations, and alpine shrub meadows at higher altitudes (Wu 1987). This study shows that the coverage of suitable habitats for rare and endangered plants at low and medium altitudes within protected areas is insufficient. The level of protection in these areas is significantly lower than that at higher altitudes (Table 3).
Forests in low-altitude areas are often outside the jurisdiction of protected areas, and existing reserves lack long-term, effective, and sustainable protection plans. In addition, most of the core priority areas identified by the model are rarely covered by existing protected areas, undermining their conservation effectiveness and exacerbating protection gaps. These gaps render low- and medium-altitude habitats more vulnerable to threats such as land-use change and climate impacts, which in turn pose a greater risk to rare and endangered species. To address these issues, we recommended expanding the boundaries of the Gaoligong Mountain National Nature Reserve to ensure comprehensive protection of its biodiversity.
4.2 Human Activities Drive Habitat Fragmentation and Loss
The periphery of the study area is predominantly characterized by collective forests and farmland, shaped by a combination of human activities and environmental heterogeneity (Zhu and Yue 2016). Our analysis revealed that cropland currently occupies 14.93% (446 km2) of the potentially suitable habitat for rare and endangered plants. An examination of the land use transfer matrix from 1980 to 2020 indicated that approximately 10% of the forest area had been converted into cropland or developed for buildings, with cropland conversion accounting for the most significant proportion. This illustrates that cropland poses the greatest threat to the potential habitat for rare and endangered wild plants in the southern Gaoligong Mountains, primarily driven by human-induced land use changes in the region (Zhang 2008). Habitat loss due to increased anthropogenic pressures poses a significant threat to species across most protected areas in China (Shrestha et al. 2021). Notably, protected areas in the southwest region are particularly vulnerable, as they face the intersection of three critical challenges: threatened species, climate change, and human interference (Shrestha et al. 2021).
The southern Gaoligong Mountains are an environmentally sensitive region where human activities predominantly drive land use changes within the Gaoligong Mountain Nature Reserve. While the protected area above 1600 m is well preserved with intact vegetation, the landscapes below this altitude consist of a mosaic of agricultural and forestry areas. These lower-altitude regions feature relatively gentle terrain interspersed with cultivated land and farmland (Ma et al. 2022). Moreover, the abundant rainfall and humid climate in the Gaoligong Mountain region (Fan et al. 2010) create favorable conditions for agricultural development. As a result, plantations and farmlands have gradually encroached upon the boundaries of the protected area, leading to severe degradation of valley vegetation (Li, Dao, and Li 2011; Lu et al. 2017). This degradation is corroborated by our field survey findings in the surrounding areas.
The Gaoligong Mountains, oriented in a north-south direction, serve as a vital corridor facilitating the northward expansion and the southern retreat of diverse biota during periods of global climate fluctuations (Yi et al. 2021). However, the expansion of agriculture has fragmented native forests, significantly increasing the risk of extinction for rare and endangered plant species in the southern Gaoligong Mountains. To tackle this issue, future land use planning should be carefully developed and rigorously enforced to restrict agricultural expansion in key ecological areas. A robust monitoring system should also be established to track the population dynamics and habitat changes of rare and endangered species. Special attention must be given to human activities along the ecological transition zones of the reserve to ensure better protection and restoration of vegetation at medium and low altitudes.
4.3 Climate Change-Derived Habitat Shift
Rare and endangered plants are generally less resilient to climate change than more common, widespread species, making them particularly vulnerable to climatic impacts (Yang et al. 2021). Comparing scenarios from 2041 to 2060 under two representative concentration pathways, our analysis reveals an overall increasing trend in suitable habitat areas for these rare and endangered plant species in the southern Gaoligong Mountains, suggesting a potential positive effect of climate warming on regional plant growth. However, under SSP2-4.5, the highly suitable area is expected to decrease by 321.37 and 270.45 km2 under SSP5-8.5. These reductions indicate that the massive emission of greenhouse gases and extreme climate deterioration could significantly threaten plant survival and development. While elevated CO2 levels can promote plant growth (Bazzaz 1990), their effects vary across plant species (Fang et al. 2014). Moreover, persistently increasing CO2 concentrations, coupled with high temperatures and drought conditions, may exacerbate plant stress and lead to further reductions in highly suitable habitats (Teskey et al. 2015). Although the total suitable area for these species may expand, the far-reaching impact of climate change cannot be ignored. The reduction of highly suitable areas poses a serious threat to the long-term survival of species, highlighting the urgent need for biodiversity conservation.
As global temperatures rise, many plant species tend to migrate to higher latitudes and elevations (Fang, Zhu, and Shi 2017; Lin et al. 2023; Yang et al. 2023). Our study corroborates this trend, predicting that the habitats of rare and endangered plants in the southern Gaoligong Mountains will also shift to higher altitudes in the future, which is consistent with findings from previous research. To support these migrations and ensure the survival of rare plants in their new habitats, we propose the establishment of ecological migration corridors spanning low-, medium-, and high-altitude areas. These corridors would help plants overcome geographical barriers and facilitate smooth transitions to suitable new habitats. Specifically, ecological corridors can provide propagation pathways for seed dispersal through mechanisms such as wind, animals, and water, enabling seeds to establish themselves in new places. Additionally, these ecological corridors would connect fragmented plant communities, allowing genetic exchange and reducing the negative impacts of habitat fragmentation.
Future climate projections for East Asia indicate increasing temperatures and precipitation levels (Dai, Cheng, and Lu 2022; Sun and Ding 2010). Over the past 60 years, one-third of China's protected areas have experienced average temperature increases exceeding 1.5°C (UNFCCC 2015). Even under the most optimistic emission scenario (SSP1-2.6), nearly one-third of China's protected areas are expected to warm by more than 1.5°C (Shrestha et al. 2021). The IPCC report on climate change indicates that a 1.5°C temperature rise could lead to a more than 50% reduction in the geographical range of 8% of plant species (Masson-Delmotte et al. 2018). Climate change is thus expected to significantly reduce the diversity of threatened species in the protected areas, compromising their conservation effectiveness (Peng, Zhang, et al. 2022). Furthermore, studies suggest that the number of species threatened by future climate change may far exceed those currently listed as threatened (Peng, Hu, et al. 2022; Peng, Luo, et al. 2022). As climate change progresses, the threat status of species is likely to shift. Therefore, it is recommended to incorporate climate change considerations into the optimization of protected areas to ensure the continued protection of rare and endangered plants.
4.4 Conservation Suggestions for the Southern Gaoligong Mountain Area
- 1.
Comprehensive and long-term monitoring: Scientists should design and implement thorough surveys to assess habitat changes and species population dynamics while also establishing long-term monitoring systems. High-resolution forest cover maps, generated using satellite remote sensing and GIS technologies, can help evaluate the potential impacts of ecotourism on the environment. Such data would provide a scientific basis for developing eco-friendly tourism models that do not exceed the ecological carrying capacity while ensuring the protection of the area's original ecological landscapes.
- 2.
Land use planning and responsible ecotourism development: Authorities should use the data provided by scientists to incorporate important habitats into land use planning, thus ensuring these areas are protected from development. Sustained financial support for monitoring projects should also be secured. Relevant policies should be formulated to encourage ecotourism while regulating its scale to prevent excessive development. Local communities should be actively engaged in ecotourism projects to enhance their economic well-being and strengthen their support for ecological protection.
In addition to protected areas, ecological redline policies also play a crucial role in biodiversity conservation. These demarcated zones often encompass habitats that are critical to numerous species, including rare and threatened ones. These policies delineate regions with vital habitats occupied by rare species, imposing development restrictions to minimize habitat fragmentation, maintain ecosystem integrity, and facilitate species migration and gene flow (Yang, Gao, and Zou 2014). At the end of 2022, China's “National Park Spatial Layout Plan” was enacted, designating 49 national park candidate areas, including the Gaoligong Mountains, which are deemed an excellent example of coniferous forest habitat from the south Hengduan Mountains (Tang et al. 2023). These candidate areas are closely linked with ecological redline zones, providing safe habitats for endangered plants. This alignment promotes in situ protection and germplasm conservation and provides valuable sites for scientific research aimed at enhancing biodiversity conservation.
4.5 Research Limitations and Future Prospects
Despite using the RF and InVEST models for our conservation gap analysis, the accuracy of model predictions was constrained by multiple factors. Single species distribution models (SDMs) have certain limitations when used to predict the habitats of rare and endangered species. The data used in this study were primarily sourced from online databases and limited field surveys, which may introduce deviations from actual species distribution. In addition, the effectiveness of SDM relies on the availability of sufficient distribution data. However, rare and endangered species often have limited distribution ranges, resulting in scarce data. When sample sizes are insufficient, models may suffer from overfitting, which can undermine their ability to predict habitat suitability, especially in areas with insufficient data coverage. For example, regions with sparse distribution data produce biased suitability predictions. This study only considered 22 environmental factors related to climate and topography, excluding other key factors that may affect distribution, such as soil characteristics, dispersal barriers, and the presence of competing species. The omission of these variables may limit the comprehensiveness of the model. To address these limitations, future studies could integrate multiple modeling approaches or combine multi-source data, such as species functional traits and ecological interactions, to provide a more comprehensive assessment of suitable habitats for rare and endangered species. By incorporating diverse data sources and methodologies, the accuracy of habitat suitability predictions is likely to improve, thereby offering more robust guidance for effective conservation strategies.
5 Conclusions
Using the RF and InVEST models to assess the effectiveness of protected areas in safeguarding rare and endangered plant species in the southern Gaoligong Mountains, our results reveal that the majority of habitat areas for these plants (approximately 76.86%) remain unprotected. Currently, existing protected areas cover only 6% of potential habitat hotspots. Moreover, 59.5% of the suitable habitats for rare and endangered plant species are located in low- and medium-altitude areas, suggesting that the existing protected area network does not adequately protect local biodiversity.
Land use changes, particularly the expansion of agricultural land, are the primary drivers of habitat loss for rare and endangered plant species in the region. Despite these challenges, suitable areas for these species are expected to continue expanding over the period from 2041 to 2060. Our combined modeling approach (RF-InVEST) provides a flexible and dynamic assessment of protected area effectiveness, accounting for land use dynamics and potentially improving protected area conservation efficiency. In the face of ongoing climate change, developing dynamic and adaptive conservation strategies is crucial. Therefore, we recommended considering the integrity of vertical zonation when optimizing protected areas to ensure the long-term protection of rare and endangered plant species.
Author Contributions
Yong-Jing-Wen Yang: conceptualization, data curation, formal analysis, methodology, software; visualization, writing–original draft; writing–review and editing. Xin-Run Hu: data curation, investigation. Min Deng: conceptualization, data curation, formal analysis, methodology, writing–review and editing. Yun-Hong Tan: conceptualization, data curation, formal analysis, methodology, writing–review and editing.
Acknowledgments
The authors are grateful to Professor Richard T. Corlett and Professor Yang Bai of XTBG for their invaluable suggestions and insightful comments in improving the manuscript. They also express their appreciation to Prof. Rong Li (Kunming Institute of Botany, CAS) for generously providing data and to Lin Lin (Yunnan University), Zhongde Huang (XTBG), and Lin Wang (XTBG) for their technical assistance with the software. Their heartfelt thanks go to the dedicated staff of Gaoligong Mountain National Nature Reserve and Longling Xiaoheishan Provincial Nature Reserve for their invaluable support during the field data collection. This research was supported by the National Natural Science Foundation of China (Grant nos. 31970223 and 31972858) and the Project of the Yunnan Province Science and Technology Department (Grant nos. 202303AK140009, 202405AC350011, and 202203AP140007), the Transboundary Cooperation on Biodiversity Research and Conservation in Gaoligong Mountains (Grant no. E1ZK251) and the Project of the Southeast Asia Biodiversity Research Institute, Chinese Academy of Sciences (Grant no. Y4ZK111B01), and Yunnan Science and Technology Talent and Platform Program (202205AG070006).
Conflicts of Interest
The authors declare no conflicts of interest.
Open Research
Data Availability Statement
The data that supports the findings of this study are available in the supplementary material of this article.
