Soils regulate the water cycle, sustain food production systems, play a key role in climate change mitigation, and are relevant to addressing the most important human challenges. Monitoring soil functions and translating the knowledge into policy depends on spatially explicit, representative, and accurate soil data. Despite decades of soil data collection, open soil data repositories remain patchy, inconsistent, and insufficient for the scales at which decisions are required. Open data presents an opportunity that the soil science community has been slow to embrace. This narrative review synthesizes the barriers to soil data sharing and the mechanisms proposed to address them. While some barriers reflect legitimate interests, the main legal constraints reported include ambiguous licensing frameworks, data privacy regulations, and national sovereignty considerations. Cultural sharing hesitation is sustained by misaligned incentive systems, competitive advantage logic, fear of misuse, and insufficient recognition of data contributors. Institutional fragmentation and gaps in data governance compound all other barriers. We propose a conceptually grounded framework of enabling solutions that encompass legal reform and cultural transformation, policy-driven incentive redesign, harmonized technical standards, FAIR-aligned infrastructure investment, capacity strengthening, and the development of federated Soil Information Systems (SIS). We argue that unlocking the groundbreaking potential of open access soil data requires a coordinated, multi-stakeholder paradigm shift.
Potentially toxic elements in agricultural soils can pose environmental concerns, particularly in tropical floodplains where rice cultivation interacts with geomorphic, hydrological, and edaphic processes. We evaluated total concentrations and environmental associations of Fe, Mn, Cr, Cu, Ni, Zn, Pb, and As in surface soils from three rice-growing areas of the Yuna River basin, Dominican Republic, with primary emphasis on Arenoso. Concentrations were measured by energy-dispersive X-ray fluorescence, and screening-level contamination and ecological-risk indices were calculated for seven elements using operational subsurface-reference concentrations. Multivariate analyses summarized covariation, while regression models evaluated standardized geomorphic, hydrological, and textural predictors with locality as a fixed blocking factor; random-intercept models were used as a sensitivity analysis. Arenoso showed generally low to moderate index-based contamination and low potential ecological risk under the principal reference scenario. Negative clay coefficients for Ni, Zn, and As were consistent in direction across locality specifications, but none of the 32 primary environmental coefficients remained significant after Benjamini–Hochberg false-discovery-rate correction, and locality–clay analyses indicated that the negative slopes were strongest in La Mata. The negative Ni–elevation association obtained with the random-intercept specification changed sign in the fixed-blocking model. Ordinary kriging showed clear predictive improvement over a leave-one-out mean predictor only for Cu; Cr and Zn showed modest skill, whereas Mn, Ni, and As showed no improvement. These results indicate element-specific and locality-dependent patterns and support cautious interpretation of spatial and index-based evidence in this heterogeneous tropical floodplain setting.
Soil microbial communities play a central role in nutrient cycling and the sustainability of pasture-based agricultural systems. The relative influence of grassland management systems, soil pH, soil organic matter (OM), and botanical composition on soil microbiome structure in temperate grazing systems remains unclear. In this study, shotgun metagenomic sequencing was used to characterise microbial communities within the upper 10 cm of soil collected at a single time point in May 2024 from a managed grassland farm systems experiment comprising perennial ryegrass (Lolium perenne L.; PRG) monocultures receiving 200 kg N ha⁻¹ yr⁻¹ and perennial ryegrass-white clover (Trifolium repens L.; WC) swards receiving 150 or 100 kg N ha⁻¹ yr⁻¹ . Microbial alpha diversity (richness, Shannon and Simpson indices) and beta diversity showed limited differentiation among grassland management systems at both genus and species levels. In contrast, microbial community composition was significantly associated with soil pH, OM, and WC abundance. Genus-level PERMANOVA identified significant relationships with soil organic matter (OM) (R² = 0.085, adjusted p = 0.004), soil pH (R² = 0.070, adjusted p = 0.028), and WC proportion within the plant community (R² = 0.10, adjusted p = 0.026). Several genera, including Saccharomonospora, Thermopolyspora, and Defluviicoccus, were associated with soil OM, whereas fewer genera were strongly associated with soil pH. Nitrogen cycle-associated gene profiles showed limited differentiation among grassland treatments. These findings demonstrate significant associations between soil microbial community composition and environmental and botanical gradients, with limited differentiation among grassland management systems.
This study evaluated whether a global modeled soil product adds predictive value to locally calibrated digital soil mapping of topsoil soil organic carbon (SOC; 0–30 cm) in an irrigated Peruvian coastal valley. Random forest models used 135 composite samples collected in 2024–2025, terrain derivatives, temporally weighted Landsat indices, and OpenLandMap-soildb predictions. Configurations were compared using internal performance assessment with 200 random 80/20 partitions and 50 repetitions of block spatial cross-validation (SpCV); no independent external validation dataset was used. The retained model (elevation, slope, focal valley wetness index, topographic position index, and OpenLandMap-soildb p84) showed weak transferable predictive ability within the valley, with SpCV R² = 0.155 (95% CI 0.077–0.221), compared with 0.189 under random partitioning. Terrain alone achieved SpCV R² = 0.090; adding p84 increased this by 0.065, whereas p84 alone performed poorly (R² = –0.058). OpenLandMap therefore offered limited supplementary information and could not replace local environmental covariates or field observations. Predictions were displayed on a 30 m grid, but p84 retained its effective 120 m support and the composite observations represented management areas up to 3 ha. Quantile regression forest intervals achieved empirical coverage of 0.88 at the nominal 90% level under both internal resampling schemes, with a median mapped width of 1.51% SOC. The area of applicability included 97.5% of mapped pixels but assessed covariate similarity, not prediction reliability. The main contribution is an evaluation protocol showing why single-partition model rankings are unstable; the map supports screening and sampling prioritization, not field-level decisions.
