The fate of páramo plant assemblages in the sky islands of the northern Andes - Appendix S1. Vegetation and occurrence data used in this study

Aims: Assessing climate change impacts on biodiversity is a main scientific challenge, especially in the tropics, therefore, we predicted the future of plant species and communities on the unique páramo sky islands. We implemented the Spatially Explicit Species Assemblage Modelling framework, by i) calculating species’ maximum dispersal distance, ii) modelling species distributions at present up to 2100, iii) assembling models into communities. Finally, we assessed the vulnerability of sky islands based on richness and composition changes.

Location: Ecuadorian super-páramo (>4200 m)

Methods: Using species trait data, the maximum dispersal distance of 435 species was calculated. Species distribution models (SDM) were fitted to obtain current and future distribution predictions based on dispersal and bioclimatic factors. The final assemblages for present and 2100 were achieved by stacking all probabilistic SDMs and applying the probability ranking rule. The vulnerability of each sky island was evaluated by quantifying richness and composition changes.

Results: Maximum dispersal distances ranged between 0.008-6027 m/year, and across all scenarios, 70% of models showed a net loss in species distribution while 9% of all species were predicted to undergo extinction by 2100. Local richness was estimated to decrease by 56.63% on average, and composition changes in each sky island suggested a mean loss of 64.74% of their original species pool against a 12.97% gain. Finally, 5% of the sky island floras reconverted from high-elevation to low-elevation species. These numbers were usually more important for high-elevation species and the mountains Pichincha, Ilinizas and Antisana.

Conclusions: Our study is methodologically pioneer and provides novel insight on the future of páramo biodiversity. Significant losses in species distribution and changes in community richness and composition suggest drastic impacts and call for further study considering additional factors, such as land-use. Finally, we recommend focusing monitoring and conservation strategies on the northern sky islands in priority.

Methods
The dataset is issued from a compilation of VegParamo vegetation data (www.vegparamo.com) and additional occurence data obtained from different herbarium databases. The dataset includes presence-absence data for super-páramo species in Ecuador

研究目标：评估气候变化对生物多样性的影响是当前核心科学挑战之一，在热带地区尤为突出。为此，我们针对独特的帕拉莫天空岛（páramo sky islands）预测了植物物种及群落的未来存续态势。我们采用空间显性物种集合建模框架（Spatially Explicit Species Assemblage Modelling framework），具体步骤为：i）计算物种的最大扩散距离；ii）模拟当前至2100年的物种分布格局；iii）将物种分布模型整合为群落模型。最终，基于物种丰富度与群落组成变化，我们评估了各天空岛的生态脆弱性。

研究区域：厄瓜多尔超帕拉莫（super-páramo）生态系统（海拔>4200米）

研究方法：本研究借助物种性状数据，计算了435个物种的最大扩散距离。通过结合扩散因子与生物气候因子，拟合物种分布模型（Species Distribution Models, SDM）以获取当前及未来的物种分布预测结果。通过堆叠所有概率型物种分布模型并应用概率排序规则，得到当前与2100年的最终群落组成。最后，通过量化物种丰富度与群落组成变化，评估各天空岛的生态脆弱性。

研究结果：物种最大扩散距离介于0.008~6027米/年之间；在所有模拟情景中，70%的模型显示物种分布呈净缩减态势，另有9%的物种预计将在2100年前灭绝。本地物种丰富度平均下降56.63%；各天空岛的群落组成变化显示，其原有物种种库平均丧失64.74%，仅新增12.97%的物种。最终，5%的天空岛植物区系由高海拔物种群落转变为低海拔物种群落。上述变化在高海拔物种及皮钦查火山、伊利尼萨火山群与安蒂萨纳火山对应的天空岛中尤为显著。

研究结论：本研究在方法学上具有开创性，为帕拉莫生物多样性的未来态势提供了全新认知。物种分布的显著缩减以及群落丰富度与组成的剧烈变化，表明气候变化带来了严重影响，呼吁后续研究纳入土地利用等额外影响因素。最后，我们建议优先将监测与保护策略聚焦于北部天空岛。

方法：本数据集整合了VegParamo植被数据集（www.vegparamo.com）与从不同标本馆数据库获取的额外物种出现数据，包含厄瓜多尔境内超帕拉莫物种的存在-缺失数据。