Dataset_sulfonamide_sorption_soils
Description
This dataset supports the systematic and critical review “Sorption of sulfonamides in soils: Retention, remobilization, transport, and limits of environmental prediction—A systematic and critical review”. It contains the reconciled formal corpus, report-level metadata, audited parameter-level extraction, classifications of parameter availability and comparability, reference subsets, paired and non-pairable environmental contrasts, and evidence matrices addressing desorption, ageing, remobilisation, transport, transformation, and environmental-risk inference. The formal corpus comprises 68 documents, including 63 empirical-core reports, four mechanistic complements, and one linked publication. The empirical database contains 1,255 primary parameter-level records and 1,195 classified Kd, Koc, and KF records. Records retain their original units, analytical families, experimental conditions, source locations, and audit status. Multiple records from the same report or environmental origin must not be interpreted as independent study-level replications. The workbook includes a README, data dictionary, report inventory, primary extraction tables, analytical subsets, environmental contrasts, evidence matrices, and quality-control checks. Local file paths, copyrighted source documents, internal workflow fields, and operational project records were removed from the public version. Associated manuscript currently under submission to the Journal of Environmental Management.
Files
Steps to reproduce
This is a literature-derived dataset; no original samples, instruments or reagents were used by the authors. Peer-reviewed articles published from 2004 to 2025 were eligible when they investigated sorption, adsorption or desorption of at least one sulfonamide in whole soil, soil profiles, columns, lysimeters or field systems. Studies of soil fractions, minerals or model sorbents were retained only as mechanistic complements supporting interpretation of whole-soil behaviour. Scopus, Web of Science Core Collection and SciELO were searched using terms for sorption, adsorption, desorption, mobility, leaching, transport, antibiotics, soil and sulfonamides. Results were limited to articles from 2004–2025; records focused exclusively on antimicrobial resistance or resistomes were excluded. Reference lists were examined for backward citation tracking. Records were exported, deduplicated by DOI and title with manual checking, screened by title and abstract, and assessed in full text. Doubtful cases and full texts were assessed by two reviewers, with disagreements resolved by a third. Included documents received source codes (SCO, WOS, SCI or SNB) and were classified as empirical-core reports, mechanistic complements or linked publications. The final corpus comprised 68 documents: 63 empirical reports, four mechanistic complements and one linked publication. Data were extracted into a report-level table and a long-format parameter table. Each parameter row represents report × environmental unit of origin × sulfonamide × condition × process × model × parameter. Fields include soil properties, compound identity, solution chemistry, experimental design, analytical method, model, value, unit, equation and source location. Original values, units and mathematical conventions were preserved. Reported, calculated, estimated and modelled values were distinguished. Missing data were not imputed, graph-only results were not silently digitised and non-detects were not converted to zero. Related reports were reconciled using authorship, location, soil identity, compounds, treatments, cross-citations and statements of sample or data reuse. Evidence families and environmental-unit codes prevented related publications or multiple rows from being counted as independent replications. Parameters were harmonised only when dimensional equivalence and model conventions were explicit. Kd, Koc, Freundlich, column-derived and dynamic parameters were kept in separate analytical families and assigned comparability tiers. Reference subsets, paired treatment–reference contrasts and non-pairable contextual subsets were generated. Quality control checked identifiers, row counts, formulas, duplicates, missing-value codes and agreement between tables and totals. The final database contained 1,255 primary rows, 1,195 classified Kd/Koc/KF records, 160 auxiliary records, 36 mechanistic records and 54 paired contrasts. Consult the README and DATA_DICTIONARY worksheets before analysis.
Institutions
- Universidade Tecnológica Federal do ParanáParaná, Curitiba