Ponds offer vital resources and living spaces for a wide range of both aquatic and terrestrial organisms (Fehlinger et al. 2023). These small permanent aquatic systems, regardless of their origin (artificial or natural), provide habitats to plants, amphibians, insects, and birds (Simaika et al. 2016; Bubíková and Hrivnák 2018a; Zamora-Marín et al. 2021) and often rare species that are not found in larger water bodies (Oertli et al. 2002; Bolpagni et al. 2019). Despite their ecological importance, they are among the most vulnerable and endangered habitats in the world (Dudgeon et al. 2006). In particular, aquatic vegetation is undergoing significant decline driven by human activities, including urban expansion, habitat fragmentation and destruction, pollution, eutrophication, and the spread of invasive species (Akasaka et al. 2010; Bolpagni et al. 2019, 2020; Bolpagni 2020, 2021; Fernández-Aláez et al. 2020; Du Toit et al. 2021). Ponds represent one of the most common and widespread freshwater habitats, especially in human-modified landscapes. Despite their small size, ponds are essential for biodiversity conservation (Biggs et al. 2017) since they tend to host significantly higher numbers of species, including more unique and rarer species, than other types of water bodies (Oertli et al. 2005).

Pond vegetation is typically composed of wetland plants, including emergent species that are rooted in the sediment but extend their vegetative parts above the water, and aquatic plants, which include submerged species with entirely underwater leaves, along with rooted floating and free-floating types (Ervin 2023). Beyond hosting a rich diversity of strictly aquatic flora, ponds also provide suitable conditions for surrounding semi-terrestrial plant species, namely riparian vegetation (Scheff et al. 2022). These plants occupying the transition zone between land and water interact closely with aquatic habitats and, as wetland and aquatic species, are shaped both by land use and water characteristics (Wang et al. 2021; Musisi et al. 2025). Pond identity is relevant in defining plant species richness and community composition (Cannucci et al. 2025) given the interplay of multiple filtering processes, particularly local environmental filters such as physical aspects of ponds (e.g., area, depth) and physico-chemical properties of water, which influence the diversity patterns of plant species (Hrivnák et al. 2013). Plant diversity metrics can change in relation to the type of water body which can contribute uniquely and significantly to plant diversity (Bubíková and Hrivnák 2018b; Grasel et al. 2021). The observation of the ecological patterns occurring in ponds is essential to summarize the vegetation-environment relationships. Many studies on ponds have investigated the responses of plant communities to environmental drivers (Gallego et al. 2015; Fernández-Aláez et al. 2018; Sieben et al. 2021), leading to different outcomes on the most relevant drivers of plant species composition.

Accessible datasets are key tools for providing an overview of species and habitat distribution, and they support biodiversity conservation actions by highlighting the occurrence of species of conservation interest or alien species (Santoianni et al. 2025). Given the strong relationship between plant communities and environmental characteristics, we present the PONDY dataset: Pond vegetation data and water physico-chemical parameters. PONDY combines vegetation data of ponds with water physico-chemical parameters, thus aiming to serve as a tool to evaluate the ecological status of water bodies and to analyze how environmental conditions, related to water quality, affect species presence, abundance, and cover.

The study area (Fig. 1) encompasses 115 permanent farmland ponds distributed across continental (Fig. 1a, b) and insular (Fig. 1c) areas of Italy. We selected permanent ponds ranging from 70 m² to 3 ha. In each area (Suppl. material 1: fig. S1a), we chose three zones, namely pondscapes (interconnected pond networks in a landscape) based on the extent of agricultural land use (Suppl. material 1: fig. S1b). The percentage of agricultural land was calculated within a 10 km radius using Corine Land Cover maps (ISPRA 2018). Three pondscapes, based on agricultural land-use extent, were established: low (<30% agricultural land-use extent), intermediate (30–60% agricultural land-use extent), and high (>60% agricultural land-use extent). Ponds were identified within each pondscape (Suppl. material 1: fig. S2) by extracting water bodies classified under the “Water Bodies” category (5.1.2) from the Corine Land Cover map (ISPRA 2018) using QGIS (QGIS Development Team 2023). In each pondscape, the 10 km buffer zone was divided into a 500 m × 500 m grid, which was overlaid on the extracted water bodies. From this grid, we randomly selected one pond per grid cell, using QGIS’s Random Selection tool within subsets. To overcome potential accessibility issues, the selection process was repeated three times, ensuring a minimum distance of 1 km in a straight line between selected ponds. Within each area, between 10 and 15 ponds were randomly selected (see Suppl. material 1: fig. S1c). For each pond, using the QGIS Random Points Along Lines plugin, we generated three points along the pond perimeter, setting a minimum distance of 15 m between points. In the field, in proximity of these points we positioned the plots and at each point we surveyed one aquatic plot measuring 2 m × 2 m, along with one terrestrial plot of the same size located 1 m away from the aquatic plot (Suppl. material 1: fig. S1d), thus obtaining a total of 6 plots (3 aquatic and 3 terrestrial plots) for each pond. With this sampling design, some aquatic plots resulted empty (no species recorded); these latter plots were therefore removed from the dataset due to the absence of vegetation.

All ponds were sampled during the peak of the growing season. Specifically, ponds in the continental areas of Italy were surveyed between June and August 2020, 2021, and 2023, those in insular Italy in late April 2024. In each plot, we recorded all the occurring species, including vascular plants and macroalgae of the Characeae family. Vascular plant species nomenclature follows the Portal to the Flora of Italy (2025). Nomenclature of Characeae follows Bazzichelli and Abdelahad (2009). In the field, together with vegetation data, we also collected physico-chemical water parameters to obtain data influencing aquatic and riparian species. Multiparametric probe (Aquaread 2000-d) was used to record several parameters at the center of each aquatic plot, between 8:00 a.m. and 3:00 p.m., including water temperature (°C), depth (m), dissolved oxygen (expressed in percentage, %), pH, turbidity (NTU – Nephelometric Turbidity Unit), and electrical conductivity (µS/cm). Within each plot, a water sample was collected and immediately filtered through a 0.7 μm glass fiber filter (GF/F, Whatman) for subsequent analyses. In the laboratory, soluble reactive phosphorus (as phosphate ion, PO₄³⁻; µg/L) was measured spectrophotometrically following the method by Valderrama (1977). The remaining water was filtered using a 0.2 μm nylon membrane, and concentrations of dissolved nitrate (NO₃⁻; mg/L) and ammonium (NH₄⁺; mg/L) were quantified via ion chromatography (883 Basic IC plus, Metrohm, Herisau, Switzerland). Given the higher relevance of aquatic vegetation for conservation assessment compared to agricultural lands, we classified aquatic plots to i) Annex I habitat types of 92/43/EEC Habitats Directive and ii) EUNIS habitat types (v2025-10-03; Chytrý et al. 2020). For these classifications, we used an expert-base approach and the code implemented by Bruelheide et al. (2021) in R 4.5.0 (R Core Team 2025), respectively.

The information present in the PONDY dataset strengthens the knowledge of pond plant diversity along insular and continental Italy. This dataset, which provides comprehensive data on plant species, community composition, habitat types, and physico-chemical water parameters, is important for understanding the plant diversity hosted in these freshwater systems and studying its relationship with physico-chemical water parameters. This dataset can be the base for future studies on the relationships between plant communities and environmental conditions and can be used for developing predictive models for species distribution based on chemical parameters. Furthermore, in the future, the dataset might be expanded with functional trait data to provide an assessment of functional composition of plant communities in relation to water chemistry. Additionally, it can contribute to define conservation status of lentic systems by assessing conservation indices, similarly to the ECELS index used in Catalonia (Sala et al. 2004), based on water characteristics, land use, and vegetation status aspects.

The vegetation plot data are available in the CircumMed database (GIVD: EU-00-026 - CircumMed database https://www.givd.info/ID/EU-00-026).

SC and RB collected the data in the field with contributions from CA, EF, TF, and FM. SC, CA, RB, and SM designed the sampling plan. SC and RB identified vascular plant species with contributions from FM and TF. TF identified Charophyceae species. ADV performed the chemical analysis in the laboratory. SC and RB assembled the dataset with contribution from LS. FC performed the semi-automatic habitat EUNIS classification. SC did the analyses. SC led the writing with contributions from GB. SC prepared the figures with contributions from GB. CA and GB supervised the research. All authors critically revised the manuscript and approved the final version.

The authors declare that no competing interests exist.