Farmland ponds as unique aquatic islands for plant diversity

Wetlands are among the world’s most vulnerable ecosystems, and in the Mediterranean region, farmland ponds are particularly exposed to the influence of surrounding agricultural land use while still supporting unique biodiversity. These small waterbodies are widely recognized as “island-like” systems; however, key functional differences exist among them. Ponds host not only strictly aquatic plants but also adjacent terrestrial species, and living at the land–water interface, these species interact with aquatic environments and, as for the aquatic species, they can be influenced by water physical and chemical properties. With this work we aimed to: i) provide an accessible collection of pond vegetation data with water physico-chemical measurements, serving as a resource for investigating the links between plant communities and their environmental settings; ii) analyze how the extent of agricultural land use and various factors at multiple spatial scales influence the diversity and structure of plant communities; iii) examine beta diversity patterns to identify levels of biological isolation among ponds and explore the relevance of island biogeography concepts to ponds within agricultural landscapes. We studied 115 ponds located in landscapes with varying agricultural land-use extents, surveying aquatic and riparian communities and water parameters. This thesis has enabled the creation of a dataset covering plant diversity and abundance, community composition, habitat types, and water chemistry parameters for ponds in Italy. Furthermore, this work has shown that plant richness decreased with increasing agricultural land use and that pond-specific features, possibly related to pond management, are more important than landscape-scale features in shaping plant diversity. It has also revealed the presence of high beta diversity among ponds, primarily driven by species turnover, making them resemble traditional island systems. Despite this, unique local processes and distinct dynamics distinguish them from true islands.