Potential groundwater dependent vegetation in the Mediterranean

Groundwater dependent vegetation (GDV) is essential for maintaining ecosystem functions and services, providing critical habitat and sustaining human livelihoods. A novel multicriteria framework helps to identify areas where potential groundwater dependent vegetation (pGDV) occurs in the Mediterranean biome. Globally-available datasets targeting 1) groundwater vegetation interaction; 2) soil water holding capacity; 3) topographical landscape wetness potential; 4) land use land cover and 5) hydraulic conductivity of rocks are combined in a weighted, easy-to-use index, composed of eleven thematic layers.
Input layers for the index calculation are available in the data collection: 1) pre-processed (rasterised and clipped to the Mediterranean) and 2) harmonised and reclassified. All input data was extracted globally. Either, directly from the respective studies or through the data catalogue in the Google Earth Engine. All datasets were acquired and processed in 2022 and 2023. Time series data for potential inflow dependency and Normalized Difference Vegetation Index (NDVI) were extracted for the period 2003-2021. Finally, the mean value was calculated over this period. All other data sets, however, mark a fixed point in time.
Ground truth vegetation data was used to calculate layer weightings with a Random Forest. 10 m * 10 m vegetation plots were collected in July and August 2021 and 2022 in southern Italy (Campania region) inside the 'Cilento, Vallo di Diano and Alburni National Park'. 236 vegetation plots are available, containing general information on the vegetation (habitat, species number, stratification), mean indicator values, plant life forms, leaf anatomy as well as a calculated ecohydrological potential for the presence of GDV. The potential was calculated based on the coverage of phreatophyte species and the moisture value of non-phreatophyte species.
The final pGDV maps including different weightings of the eleven thematic layers are compiled at a resolution of 500 m in WGS1984 (EPSG 4326). Finally, five pGDV classes (very low to very high potential) were derived and the share of high pGDV was calculated for level 8 HydroBASINS in the Mediterranean. Results support prioritisation of areas for essential regional high-resolution identification of GDV, to ensure sustainable groundwater management and in turn protect GDV as local biodiversity hotspots.

地下水依赖型植被（Groundwater Dependent Vegetation, GDV）是维持生态系统功能与服务的核心要素，可为生物提供关键栖息地，同时支撑人类生计。本研究构建了一套全新的多准则分析框架，可用于识别地中海生物群区内潜在地下水依赖型植被（potential Groundwater Dependent Vegetation, pGDV）的分布区域。该框架将五大类全球可用数据集整合为一个加权易用型指数，共涵盖11个专题图层，这五大类数据集分别为：1）地下水-植被相互作用数据集；2）土壤持水能力数据集；3）地形景观湿润潜力数据集；4）土地利用与土地覆被数据集；5）岩石导水率数据集。
用于指数计算的输入图层已收录于本数据集，这些图层均经过两项预处理流程：1）完成栅格化处理并裁剪至地中海区域范围；2）完成数据协调与重分类。所有输入数据均取自全球范围，获取渠道包括直接从相关研究文献中提取，或是通过谷歌地球引擎（Google Earth Engine, GEE）的数据目录获取。所有数据集均于2022年至2023年间完成采集与处理。针对潜在入流依赖性与归一化差分植被指数（Normalized Difference Vegetation Index, NDVI）的时间序列数据，提取时段为2003年至2021年，并最终计算得到该时段内的平均值。其余数据集均为固定时间点的静态数据。
本研究采用实地调查获取的植被数据，通过随机森林（Random Forest）算法计算各专题图层的权重。2021年7-8月及2022年7-8月期间，研究团队在意大利南部坎帕尼亚大区的“奇伦托、瓦洛-迪迪亚诺与阿尔布尼国家公园”内，采集了共计236个10米×10米的植被样地。这些样地数据包含植被的基础信息（生境类型、物种数量、群落分层）、平均指示值、植物生活型、叶片解剖结构，以及针对GDV存在性计算得到的生态水文潜力值。该潜力值基于潜生植物（phreatophyte）的盖度与非潜生植物的湿度值计算得出。
最终生成的pGDV地图包含11个专题图层的不同权重组合，空间分辨率为500米，坐标系采用WGS1984（EPSG 4326）。研究将pGDV划分为5个潜力等级（从极低到极高），并计算了地中海区域八级HydroBASINS单元内高潜力pGDV的占比。本研究结果可为优先划定区域以开展高分辨率GDV实地识别提供决策支撑，从而保障地下水的可持续管理，进而保护作为当地生物多样性热点区域的GDV群落。