Chiara Ferré

Urban soils play a crucial role in supporting ecosystem services, yet they are often overlooked or perceived as degraded. This study provides an assessment of physical and chemical properties of soils in public green spaces in Milan, Italy—a representative medium‐sized European city—with particular attention to soil compaction. Sixty georeferenced plots were sampled across five categories of publicly accessible green areas differing in vegetation cover, management intensity, and human pressure. Measurements included bulk density (BD), penetration resistance (PR), pH, organic carbon (SOC), total nitrogen, available phosphorus (AvP) and texture. Milan's soils generally retained favourable ecological properties. BD was low (mean ± sd: 1.03 ± 0.13 g cm ⁻³ ), below thresholds limiting root growth and soil biota, and PR profiles showed no critical compaction. However, BD and PR varied across categories: values were lowest in peri‐urban forests and highest in central parks and urban green islands, reflecting vegetation, management, and recreational pressure differences. SOC was relatively high in the topsoil (mean ± sd: 3.23% ± 0.75%), supporting fertility and carbon storage, with no significant differences among categories. Conversely, pH and phosphorus varied: peri‐urban sites had acidic to sub‐acidic conditions and moderate AvP levels, while central parks and urban islands showed near‐neutral to slightly alkaline pH and elevated AvP, indicating stronger anthropogenic influence. These findings challenge the widespread assumption that urban soils are uniformly compacted and degraded. Soils in Milan's green spaces preserve properties that sustain ecosystem services, with differences reflecting management intensity, recreational use and land‐use history. The framework adopted here can be applied in other cities to guide soil protection and support sustainable urban planning.

De-sealing, or depaving, is increasingly adopted to restore soil permeability and support green infrastructure, yet its potential to recover soil functions remains insufficiently understood. This study reports one year of soil monitoring following the de-sealing of a brownfield site in Milan (Italy). It compares the evolution of pedoclimatic parameters in sealed and de-sealed soils and assesses the suitability of recycled aggregates (RAs) from demolition waste as a soil-forming material. Buried sensors continuously recorded pedoclimatic parameters, temperature, water content, and oxygen concentration, while periodic sampling was carried out to analyse soil chemical properties, bacterial community composition, and the quality of percolation water (heavy metal content). De-sealing immediately improved pedoclimatic conditions, enhancing soil aeration, water regulation, and heat exchange capacity. No significant variation was detected in soil chemical properties, apart from pH fluctuations linked to the leaching of alkaline ions from concrete-based RAs. The presence of RAs caused no adverse effects on either soil or percolation water. Bacterial community composition was strongly associated with soil organic carbon, C:N ratio, and soil water content, without showing clear temporal trends. Overall, the study demonstrates that de-sealing rapidly triggers soil functional recovery and that, when properly characterised for composition and contamination risk, RAs pose no evident threat to the surrounding environment.

In the face of mounting challenges related to limited availability of urban land and ecological degradation, emerging novel ecosystems offer unique opportunities for ecological regeneration, social redefinition of space, and alternative urban visions. This study presents the multi-layer analysis of the Goccia Forest in Milan (Italy), a wild urban woodland that has developed over sealed and polluted post-industrial land, aiming to investigate the potential of this novel ecosystem to sustain Nature-based Solutions (NbSs). Using an integrated approach (surveys on fauna, vascular flora, lichens, analysis of forest evolution, mapping of sealed surfaces, and soil characterization) the research looks at the novel ecosystem as a whole, highlighting its ecological dynamics and Ecosystem Services (ES). La Goccia Forest serves as a prime example of how the implementation of NbSs is intricately intertwined with the spontaneous regeneration of urban brownfields. The present study offers the opportunity to rethink urban policies, ensuring their alignment with the demands of the population and the latest scientific knowledge.

In this study, we conducted a comprehensive analysis of the spatial distribution of soil organic carbon stock (SOC stock) and the associated uncertainties in two soil layers (0–10 cm and 0–30 cm; SOC stock 10 and SOC stock 30, respectively), in Valchiavenna, an alpine valley located in northern Italy (450 km²). We employed the digital soil mapping (DSM) approach within different machine learning models, including multivariate adaptive regression splines (MARS), random forest (RF), support vector regression (SVR), and elastic net (ENET). Our dataset comprised soil data from 110 profiles, with SOC stock calculations for all sampling points based on bulk density (BD), whether measured or estimated, considering the presence of rock fragments. As environmental covariates for our research, we utilized environmental variables, in particular, geomorphometric parameters derived from a digital elevation model (with a 20 m pixel resolution), land cover data, and climatic maps. To evaluate the effectiveness of our models, we evaluated their capacity to predict SOC stock 10 and SOC stock 30 using the coefficient of determination (R²). The results for the SOC stock 10 were as follows: MARS 0.39, ENET 0.41, RF 0.69, and SVR 0.50. For the SOC stock 30, the corresponding R² values were: MARS 0.45, ENET 0.48, RF 0.65, and SVR 0.62. Additionally, we calculated the root-mean-squared error (RMSE), mean absolute error (MAE), the bias, and Lin’s concordance correlation coefficient (LCCC) for further assessment. To map the spatial distribution of SOC stock and address uncertainties in both soil layers, we chose the RF model, due to its better performance, as indicated by the highest R² and the lowest RMSE and MAE. The resulting SOC stock maps using the RF model demonstrated an accuracy of RMSE = 1.35 kg m⁻² for the SOC stock 10 and RMSE = 3.36 kg m⁻² for the SOC stock 30. To further evaluate and illustrate the precision of our soil maps, we conducted an uncertainty assessment and mapping by analyzing the standard deviation (SD) from 50 iterations of the best-performing RF model. This analysis effectively highlighted the high accuracy achieved in our soil maps. The maps of uncertainty demonstrated that the RF model better predicts the SOC stock 10 compared to the SOC stock 30. Predicting the correct ranges of SOC stocks was identified as the main limitation of the methodology.

In this study, we conducted a comprehensive analysis of the spatial distribution of soil organic carbon stock (SOC stock) and the associated uncertainties in two soil layers (0–10 cm and 0–30 cm; SOC stock 10 and SOC stock 30 respectively), in Valchiavenna, an alpine valley located in northern Italy . We employed the digital soil mapping (DSM) approach within different machine learning models, including multivariate adaptive regression splines (MARS), random forest (RF), support vector regression (SVR), and elastic net (ENET). Our dataset comprised soil data from 110 profiles, with SOC stock calculations for all sampling points based on bulk density (BD), whether measured or estimated, considering the presence of rock fragments. As environmental covariates for our research we utilized environmental variables, in particular geomorphometric parameters derived from a digital elevation model (with a 20 m pixel resolution), land cover data, and climat-ic maps. To evaluate the effectiveness of our models, we evaluated their capacity to predict SOC stock 10 and SOC stock 30 using the coefficient of determination (R2). The results for the SOC stock 10 were as follows: MARS 0.39, ENET 0.41, RF 0.69, and SVR 0.50. For the SOC stock 30, the corresponding R2 values were: MARS 0.45, ENET 0.48, RF 0.65, and SVR 0.62. Additionally, we calculated the Root Mean Square Error (RMSE) and Mean Absolute Error (MAE) for further assessment. To map the spatial distribution of SOC stock and address uncertainties in both soil layers, we chose the RF model, due to its better performance, as indicated by the highest R2 and the lowest RMSE and MAE. The resulting SOC stock maps using the RF model demonstrated an accuracy of RMSE = 1.35 kg.m-2 for the SOC stock 10 and RMSE= 3.36 kg.m-2 for the SOC stock 30. To further evaluate and illustrate the precision of our soil maps, we conducted an uncertainty assessment and mapping by analyzing the standard deviation (SD) from 50 iterations of the best-performing RF model. This analysis effectively highlighted the high accuracy achieved in our soil maps. The maps of uncertainty demonstrated that the RF model better predicts the SOC stock 10 compared to the SOC stock 30.

The United Nations Sustainable Development Goals include soil conservation as an urgent climate action to be taken; alpine grassland soils, rich in organic matter and particularly vulnerable to climate change, require special attention. In the upper Adamé valley (Italy), the effects of soil and vegetation types and environmental parameters, on soil carbon (SOC) stock and other soil properties were evaluated. Sampling points were chosen starting from the main vegetation types (dominant species: Rhododendron ferrugineum, Nardus stricta, Festuca luedii, Adenostyles leucophylla and Salix herbacea). For each vegetation type, five plots were selected, within which three soil minipits were described and sampled by layers (0–10 cm, 10–20 cm and 20–40 cm), at 75 georeferenced points. The results highlighted that: (1) Podzols and Umbrisols showed high capacity to store OC in the top 40 cm (17.3 ± 8.2 kg m⁻²), whereas Fluvisols and Leptosols demonstrated a low capacity (5.0 ± 3.5 kg m⁻²); (2) topographical factors indirectly affected the variability of the SOC stock by influencing the spatial distribution of vegetation; (3) 74% of the SOC stock was sequestered in the upper 0–20 cm layer; (4) regardless of soil type, Festuca and Salix were the vegetation types that mostly contributed to the surface SOC enrichment, followed by Rhododendron, Nardus and Adenostyles. Following climate change, a shift of the Salix and Nardus vegetation types in favour of the Rhododendron type is expected in the short to medium term. As a consequence, a possible reduction in the effectiveness of the climate regulation service provided by the soils may occur.

Mountain soils provide a variety of ecosystem services (ES) that are increasingly threatened by anthropogenic impacts and climate change.We present here the results of a study on the pedodiversity, biodiversity, and soil organic carbon (SOC) sequestration of alpine soils in a pasture of the central Alps, that took into account the effect of morphology and vegetation. The study area was the Andossi plateau (350 ha) in Valchiavenna (Lombardy Region), between 1800 and 2000 m elevation.As part of the PascolAndo project, whose goal was to promote sustainable pasture management, at 160 georeferenced points we characterized soils (down to bedrock or a maximum of 50 cm) and carried out the floristic survey to define the vegetation type; soil samples were collected by horizons and analyzed for the main soil properties.The main soil types were Leptosols, Regosols, Cambisols, Umbrisols, Podzols, Gleysols and Histosols, strongly related to geomorphology and vegetation type (peat bogs, earth hummocks, and poor, calcareous, shrub and rich pastures).Histosols and Gleysols of peatland and wet areas showed the highest SOC stock with an average of 31.8±2.6 kg m-2 and 23.3±2.7 kg m-2 respectively, followed by Umbrisols, distributed all over (17.2±6.2 kg m-2), and Podzols (16.6±4.6 kg m-2), typical of poor pastures and shrub areas. The most prevalent soils on the plateau were Cambisols (average SOC stock of 13.6±2.7 kg m-2): Eutric Cambisols were primarily in rich pasture and Dystric Cambisols in poor pasture. Among poorly developed soils, Regosols stored on average 10.8±4.2 kg m-2, while Leptosols, widespread on steep slopes in calcareous pastures and shrub areas, showed the lowest SOC stock (9.6±4.9 kg m-2).Soil types, vegetation types, and SOC were first modeled using geomorphometric variables (27 variables extracted from a 4-m resolution DTM) and vegetation indices (NDVI, NDWI, NDRE, for 3 dates of the growing season, extracted from Sentinel 2 images) and then spatially estimated, allowing quantification of the pedodiversity, biodiversity, and SOC sequestration in the study area.The study showed that the plateau has a high level of biodiversity and pedodiversity, with OC-rich soils that are important for regulating the global climate. There was evidence that grazing and livestock loading affected soil ES by acting on biodiversity, soil fertility, and erosion.