The stakes are considerable. Soil maps built on harmonized terminology are indispensable for capturing spatial variation in soil cover, supporting land-use planning, and integrating national data into international frameworks. Because soil surveys across Europe were conducted independently, converting national legends into a common classification has become a continent-wide challenge. The World Reference Base, or WRB, does not seek to replace national systems; it serves as a reference frame for correlating soils internationally. But as the new study, published in the Spanish Journal of Soil Science, demonstrates, a straightforward one-to-one conversion of Portuguese soil units into WRB categories is simply not possible for most of the country’s classification hierarchy.

The researchers, Manuel Madeira, Samuel Guerreiro, Pedro Arsénio, and Vasco Florentino, focused on roughly 47,555 square kilometers of southern and central Portugal, spanning the Algarve, Alentejo, the Lisbon metropolitan area, and part of the Central region. The area’s climate is predominantly Mediterranean — hot, dry summers and mild, wet winters — with annual rainfall ranging from 400 to 1,000 millimeters. Geologically, it encompasses the ancient Iberian Massif with its Precambrian and Paleozoic metamorphic and igneous formations, alongside younger Meso-Cenozoic sedimentary basins. This diversity of parent materials, relief, and climate produces a wide array of soils, and it is precisely this complexity that the 1974 classification struggled to capture in terms compatible with modern diagnostics.

The Classification of the Soils of Portugal is a hierarchical system with five categorical levels: orders, suborders, groups, subgroups, and families. Its ten orders include Incipient soils, Litholic soils, Calcareous soils, Mollic soils, the evocatively named Barros — shrinking and swelling clay soils — Argilluviated soils, Podzolized soils, Halomorphic soils, Hydromorphic soils, and Hydromorphic organic soils. The system was conceived before sufficient data and modern processing tools existed, and its definitions lack the quantitative diagnostic criteria and thresholds that the WRB demands. The 1:25,000 soil map built on it relies on scarce, non-georeferenced reference soil profiles, and surveyors of the era applied subjective judgments about soil depth and parent material that varied from one mapper to another.

To test how well the old system maps onto the new, the team compared the classification of 186 legacy reference soil profiles under both systems. The results were striking. No CSP order achieved a perfect one-to-one match with any WRB Reference Soil Group. The strongest correspondence belonged to the Barros order, whose shrinking and swelling clay soils matched Vertisols in 83 percent of profiles. Among suborders, Colluvial soils corresponded entirely to Regosols, and Humic Litholic soils matched Umbrisols in 78 percent of cases. Everything else fractured. Lithosols, defined vaguely as soils with effective depth under 10 centimeters, matched Leptosols in only 24 percent of map units, with the remainder splitting among Cambisols, Regosols, and other groups.

Some mismatches proved scientifically revealing. The CSP’s Argilluviated soils — soils with clay-enriched subsoils — corresponded to Luvisols in only 40 percent of cases, because the criteria for defining an argillic horizon differ from those for the WRB’s argic horizon, and because lithic discontinuities in sedimentary parent materials were not considered in the old system. Podzolized soils matched true Podzols in just 23 percent of map units, since the Portuguese definition allowed soils without the characteristic ortstein layers to be included, and no oxalate-extractable aluminum and iron determinations were ever performed to confirm spodic horizons. Calcareous soils corresponded largely to Cambisols rather than Calcisols, because most lacked a true calcic horizon, their carbonates being simply inherited from the parent rock.

Recognizing that direct translation was failing, the researchers turned to soil-forming factors — the classical pillars of pedology. They framed the legacy soil profiles within landscape units defined by lithological groupings, relief classes derived from a digital terrain model, and climate classes based on temperature data from 1981 to 2010. This environmental framing dramatically improved correspondence. Landscapes of granite and related rocks proved to be 73 percent Regosols with 27 percent Cambisols, matching the near-exclusive dominance of Non-Humic Litholic soils in the old map units. Diorite landscapes were 77 percent Luvisols, mirroring the dominance of Argilluviated soils. On mafic rocks, Vertisols occupied 60 percent of the landscape, echoing the Barros order’s 60 percent share.

The environmental approach also exposed errors in the old maps. In schist landscapes, the 1:25,000 map showed only Argilluviated soils and Lithosols, yet the profile data revealed that Luvisols, Cambisols, Regosols, and Leptosols all occur there — meaning both Lithosols and Argilluviated soils were overrepresented. Similarly, in sandy coastal landscapes, Podzols turned out to be far less extensive than earlier maps suggested, occupying mainly plains while only Arenosols appear on strongly sloping terrain. Relief proved to be a powerful discriminator: Regosols and Leptosols increase from undulating to steep areas, while Planosols, Stagnosols, and Solonetz appear only in plains or gently undulating landscapes.

Beyond assigning Reference Soil Groups, the WRB system allows soils to be refined with principal and supplementary qualifiers — descriptors such as Leptic, Skeletic, Calcaric, Vertic, Eutric, or Clayic that capture specific properties. The team found that many of these could only be tentatively assigned, because the CSP never quantified soil depth, coarse fragments, abrupt textural differences, or exchangeable aluminum. Still, clear lithological patterns emerged: the Leptic qualifier tracks parent material, declining from Regosols to Cambisols to Luvisols on the same rock type; the Calcaric qualifier clusters on non-compact limestones; the Vertic qualifier on mafic rocks; and the Clayic qualifier on clayey and fluvial materials. Even the old map’s soil phases offered clues — thick soil phases reliably indicated the absence of the Leptic qualifier.

The study concludes that at least 24 of the WRB’s 32 Reference Soil Groups likely occur in Portugal, with 22 on the mainland alone, underscoring the exceptional diversity of Mediterranean soils shaped by varied parent materials, landforms, and centuries of human activity. The proposed methodology offers a viable path toward a unified national 1:100,000 soil map in the WRB system, but the authors are candid about its limits: with no funding for new fieldwork, legacy data alone cannot resolve every ambiguity, and decades of land-use change — irrigation expansion, olive groves, forest plantations, wildfires — may have altered soil profiles since the original surveys. They argue that Portugal now faces a choice: patch the 1974 classification or rebuild it with quantitative, internationally correlated criteria. Either way, the conversion exercise is more than a bureaucratic translation — it is an opportunity to reimagine the country’s soil geography for the era of digital soil mapping and EU-wide harmonization.

Subject of Research: Converting Portugal's legacy genetic soil classification to the World Reference Base for updating national soil mapping

Article Title: Converting the genetic-based national soil classification to the World Reference Base: a challenge for updating soil mapping in Portugal

Article References: Madeira, M., Guerreiro, S., Arsénio, P., & Florentino, V. (2026). Converting the genetic-based national soil classification to the World Reference Base: a challenge for updating soil mapping in Portugal. Spanish Journal of Soil Science, 16, Article 16368. https://doi.org/10.3389/sjss.2026.16368

Image Credits: AI Generated

DOI: 10.3389/sjss.2026.16368

Keywords: soil classification, World Reference Base, Portugal, soil mapping, legacy soil data, Reference Soil Groups, digital soil mapping, pedology, Mediterranean soils, soil-forming factors, Vertisols, Luvisols