Scholarlink

<p>To better monitor and protect fragile and naturally dynamic Mediterranean coastal habitats we need to achieve a better understanding of their spatial extent and communities’ properties by applying methods that allow accurate assessments with a reasonable sampling effort. In this context, available remote sensing data that can capture multiple vegetation features in a standard and replicable way across time and space is pivotal for predictive habitat modelling. Here, we aim at providing relevant information to improve our knowledge on the relationships between plant communities’ features and their spectral properties and commonly employed vegetation indices, that can represent a valuable baseline for a reliable mapping of coastal habitats and their condition.&#xa0;Tuscany coast, central Italy.&#xa0;In this study, we analysed data from 647 vegetation plots covering six different EUNIS habitats located along the coast of Tuscany, central Italy. We assessed plots’ spectral signature and vegetation indices values leveraging freely available Sentinel-2 images. We then compared habitats’ spectral properties to identify significant differences. Finally, we modeled with the multivariate Random Forest algorithm the relationships between habitats’ spectral signature and the normalized difference vegetation index (NDVI) with the community-weighted means (CWMs) of functional traits that we selected based on their connection with the main axis of ecological variation and relationships with salt tolerance, to identify the most relevant phenological characteristics driving habitats remotely sensed spectral properties at a community level.&#xa0;Habitats, except embryonic and shifting dunes, displayed relevant differences in terms of both spectral signature and vegetation indices values. Spectral properties of habitats were mostly related to leaf thickness (LT), leaf density (LD) and percent vegetation cover.&#xa0;The results of this study provide valuable information for coastal habitats mapping and conservation, pivotal in ensuring the provision of ecosystem services, such as coastal protection or CO<sub>2</sub> sequestration.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Differences in Spectral Properties of Mediterranean Coastal Habitats and Their Relationships to Communities’ Functional Composition

  • 期刊论文

  • 发表日期：

    2026年7月31日

查看全文
  • Mariasole Calbi,
  • Michele Mugnai,
  • Lorenzo Lazzaro,
  • Simona Sarmati,
  • Matilde Gennai,
  • Bruno Foggi,
  • Hamid Gholizadeh,
  • Emanuele Fanfarillo,
  • Tiberio Fiaschi,
  • Emilia Pafumi,
  • Andrea Bertacchi,
  • Claudia Angiolini,
  • Simona Maccherini,
  • Daniele Viciani

摘要

To better monitor and protect fragile and naturally dynamic Mediterranean coastal habitats we need to achieve a better understanding of their spatial extent and communities’ properties by applying methods that allow accurate assessments with a reasonable sampling effort. In this context, available remote sensing data that can capture multiple vegetation features in a standard and replicable way across time and space is pivotal for predictive habitat modelling. Here, we aim at providing relevant information to improve our knowledge on the relationships between plant communities’ features and their spectral properties and commonly employed vegetation indices, that can represent a valuable baseline for a reliable mapping of coastal habitats and their condition. Tuscany coast, central Italy. In this study, we analysed data from 647 vegetation plots covering six different EUNIS habitats located along the coast of Tuscany, central Italy. We assessed plots’ spectral signature and vegetation indices values leveraging freely available Sentinel-2 images. We then compared habitats’ spectral properties to identify significant differences. Finally, we modeled with the multivariate Random Forest algorithm the relationships between habitats’ spectral signature and the normalized difference vegetation index (NDVI) with the community-weighted means (CWMs) of functional traits that we selected based on their connection with the main axis of ecological variation and relationships with salt tolerance, to identify the most relevant phenological characteristics driving habitats remotely sensed spectral properties at a community level. Habitats, except embryonic and shifting dunes, displayed relevant differences in terms of both spectral signature and vegetation indices values. Spectral properties of habitats were mostly related to leaf thickness (LT), leaf density (LD) and percent vegetation cover. The results of this study provide valuable information for coastal habitats mapping and conservation, pivotal in ensuring the provision of ecosystem services, such as coastal protection or CO2 sequestration.