Understanding the long-term dynamics of vegetation since 1953 in high-mountain regions

Alpine ecosystems, highly sensitive to climate change, are experiencing shifts in species ranges and community structure. These changes are driven by a complex interplay of climatic and environmental factors, land use changes, geomorphological dynamics, and species interactions, which can often lead to contrasted and sometimes unexpected dynamics. Historical records provide a valuable opportunity to capture these complexities by revealing long-term changes, opening a gateway to hypothesise about the key underlying processes. We investigated changes in the floristic composition of subalpine to nival vegetation communities by resurveying a period of 70 years. To understand vegetation patterns, we (i) resampled vegetation at plot level and remapped the areas, (ii) analysed the role of driving climate, environmental, and land use factors on vegetation distribution and vascular plant species richness, and (iii) modelled plant-plant interactions from community data. The results reveal that vegetation cover patterns were strongly influenced by local climate and soil properties. The species richness is also influenced by the livestock density and the flat morphology. It should be noted that climate change caused wetland habitats to become drier and accelerated secondary succession through upward migration and range-infilling processes. Furthermore, a trend towards eutrophication was observed. The results suggested that certain plant communities, particularly those found in snowbeds, were more vulnerable to environmental changes that have occurred over the past 70 years.

Synthesis: This study highlighted the complexity of vegetation dynamics. In addition to thermophilisation and aridisation, changes in land use affect species composition, species richness, and vegetation cover. Substrate conditions also play an important role.

Methods
Fieldwork

The study area was located within the Stilfserjoch/Stelvio National Park (46°31'19" N, 10°25'2" E, Bormio, Italy) and extends from 2000 m to 3094 m above sea level (a.s.l.). Historical data from 1953 documented plant communities in the study area using 219 plots (Giacomini and Pignatti, 1955). The plots were characterised based on elevation, inclination, aspect, and geographical location within this region. These informations facilitated the selection and repositioning of 42 original plots in 2023, using the topographical parameters derived from the 2015 digital elevation model (DEM; Regione Lombardia©). For full comparison with the original data, the survey method was based on Giacomini and Pignatti (1953), using the same plot size along with the original cover scale. Vegetation surveys were performed in July 2023. Permit for the fieldwork by: "Ente regionale per i servizi all'agricoltura e alle foreste" (ERSAF), with the number - ERSAF.2023.0003633.

The available map of the plant communities on a scale of 1:12,500 was scanned, georectified, georeferenced, and manually digitised in ArcGis Pro®. The remapping took place between July and October 2023 and the current mapping was adjusted to the scale and minimum mapping unit of Pignatti's map (Giacomini and Pignatti, 1953).

Plant communities were derived from Two-Way-SPecies_INdicator-analysis (TWINSPAN) and an Nonmetrical MultiDimensional Scaling (NMDS) utilising the package vegan (Oksanen et al., 2022) in R (RStudio Team, 2023). 

For further analyses a Principal Component Analysis (PCA) and a Spearman correlation were performed to avoid multicollinearity. Generalized Additive Models (GAM, mgcv package, Wood, 2022) were employed to examine the influence of potential environmental drivers on vascular species richness, total vegetation cover, and species composition stability, measured as Bray-Curtis dissimilarity between paired plots.

高山生态系统对气候变化高度敏感，当前正经历物种分布范围与群落结构的转变。这类变化由气候与环境因子、土地利用变化、地貌动态以及物种间相互作用的复杂相互作用所驱动，往往会产生截然不同且有时出乎意料的动态变化。历史记录能够揭示长期变化过程，为捕捉这些复杂性提供宝贵契机，同时为探究核心潜在过程搭建假说构建的路径。本研究通过对70年时间跨度的植被进行重调查，探究了亚高山至雪线植被群落的植物区系组成变化。为厘清植被格局，本研究开展了三项工作：①在样方尺度重新采样植被并对研究区域进行重新制图；②分析驱动性气候、环境及土地利用因子对植被分布和维管植物（vascular plant）物种丰富度的影响；③基于群落数据构建植物间相互作用模型。

研究结果显示，植被盖度格局深受局地气候与土壤属性影响；物种丰富度亦受牲畜密度与平缓地形的调控。值得注意的是，气候变化导致湿地生境趋于干旱，并通过物种向上迁移和分布区填充过程加速了次生演替；此外还观测到富营养化趋势。研究结果表明，部分植物群落——尤其是雪床生境中的群落——在过去70年的环境变化中表现出更强的脆弱性。

综合分析：本研究揭示了植被动态的复杂性。除了增温化（thermophilisation）与干旱化趋势外，土地利用变化同样会影响物种组成、物种丰富度与植被盖度；基质条件亦发挥着重要作用。

方法

野外工作
研究区域位于意大利博尔米奥的斯泰尔维奥国家公园（Stilfserjoch/Stelvio National Park，北纬46°31'19" N，东经10°25'2" E），海拔跨度为2000米至3094米。1953年的历史数据通过219个样方记录了该区域的植物群落（Giacomini与Pignatti，1955），这些样方基于该区域内的海拔、坡度、坡向及地理位置进行特征表征。基于2015年数字高程模型（DEM；伦巴第大区©）提取的地形参数，上述信息助力研究人员在2023年筛选并重新定位了其中42个原始样方。为与原始数据实现充分比对，本研究采用Giacomini与Pignatti（1953）的调查方法，使用相同的样方尺寸与原始盖度分级标准。植被调查于2023年7月开展。野外工作许可由“农业与林业区域服务机构（Ente regionale per i servizi all'agricoltura e alle foreste，简称ERSAF）”颁发，许可编号为ERSAF.2023.0003633。

现有1:12500比例尺的植物群落地图经扫描、地理校正、地理配准后，在ArcGIS Pro®中进行手动数字化。重新制图工作于2023年7月至10月间开展，并将当前制图调整至与Pignatti原地图（Giacomini与Pignatti，1953）一致的比例尺与最小制图单元。

植物群落分类通过双向指示种分析（Two-Way-Species Indicator Analysis，TWINSPAN）与非度量多维标度（Nonmetric Multidimensional Scaling，NMDS）完成，分析基于R语言（RStudio团队，2023）中的vegan包（Oksanen等，2022）实现。

为进一步分析，本研究通过主成分分析（Principal Component Analysis，PCA）与斯皮尔曼相关分析规避多重共线性问题。本研究采用广义可加模型（Generalized Additive Models，GAM，mgcv包，Wood，2022），探究潜在环境驱动因子对维管植物物种丰富度、总植被盖度以及以配对样方间Bray-Curtis相异度衡量的物种组成稳定性的影响。