The Gold Standard of Teaching Archaeological Soil Phosphate Mapping | Sampling Grids, Depth and Human Activity

A bright red phosphate hotspot on an archaeological map looks wonderfully decisive—until someone asks how the soil was sampled, what depth the number came from and whether a farmer applied fertiliser last year. The gold standard of teaching archaeological soil phosphate mapping is to make that evidence check routine.

This guide explains what the method measures, how to plan and audit a survey, and why an enriched patch is a research lead rather than an automatic identification of a building or historic activity. Orchard Court is a fully fictional worked example; no site-specific conclusions are drawn from the real photograph on this page.

Begin with the methodological review of archaeological soil chemistry and the connected Geoarchaeology guide.

1. What a soil phosphate map can actually show

Phosphorus is a nutrient and a measurable component of soils. Past human and animal activity can change phosphorus concentrations through refuse, organic matter, dung, bones, food processing and other inputs. Surveying those concentrations may reveal areas worth investigating. Yet phosphate enrichment is not an occupation certificate, and a coloured map is not a floor plan of ancient houses.

The gold standard begins with a simple question: does a spatial or depth-related chemical pattern survive comparison with the local geological, ecological and later land-use background? Teach the chemical observation separately from the historical explanation. Students should learn to say ‘elevated measured phosphorus under this protocol’ before deciding whether to say ‘kitchen’, ‘stable’ or ‘ritual space’.

2. Ask whether the method measures the same phosphorus pool

Different analytical methods can measure total phosphorus, extractable fractions or operationally defined pools. Values from one extraction procedure need not match values from another. Reporting ‘P = 500 mg/kg’ without the digestion or extraction method makes the number difficult to compare. Converting phosphate chemical forms into a phosphorus-equivalent figure also requires care.

Begin with two hypothetical lab sheets: one reports total elemental phosphorus on dry soil, the other an extractable phosphate index from another reagent and calibration. Students should identify why they cannot simply combine the values into a single contour map. A shared unit label does not make unlike chemical measurements identical.

3. Understand background chemistry before looking for a settlement

Parent material, soil texture, organic matter, calcium and iron chemistry, pH, drainage and sorption all affect phosphorus concentration and retention. A low measured value can reflect removal or different mineralogy as well as a lack of past activity. Some elevated values are natural or linked to modern agriculture.

An archaeologist therefore needs matched background samples outside the target activity area but on comparable soils. Teach students why a control from a sandy hilltop may not fairly represent a clay-rich valley floor. The strongest surveys use several control locations, not one supposedly pristine sample selected for convenience.

4. Plan a grid around a research question

A sampling grid makes a spatial survey reproducible: the field team records a site datum, coordinates, spacing, sample depth, collected mass, horizon and visible disturbance. A tight grid may be suitable for an excavated courtyard, while broader reconnaissance needs a different scale. The grid design should follow expected feature sizes, resources and the question, not a decorative desire for a dense coloured diagram.

Ask learners to draw a 4 by 4 metre room with a door, hearth candidate, wall line and an external reference transect. Compare a regular grid with targeted sampling. Targeted sampling can clarify features but cannot be interpreted like an unbiased census of the whole floor unless the selection rule is disclosed.

5. Stratigraphic depth can overturn a tidy surface hotspot

Many phosphate surveys have historically sampled a single depth, but soil layers may preserve more than one land-use phase. Later fertilisation, landscaping and flooding can strongly alter topsoil. A result from 10 cm and one from 40 cm cannot be assumed to represent the same historical moment merely because they share an x-y location.

Teach matched depth intervals, horizon descriptions and basic stratigraphic relationships. A profile with enriched topsoil but ordinary values in a sealed archaeological horizon may suggest recent processes rather than an ancient hearth. Conversely, enrichment within a secure buried floor deserves different scrutiny from dispersed signals in ploughsoil.

6. Protect the sample from modern phosphate contamination

Modern fertilisers, animal manure, footwear, dirty trowels and stored bags can all compromise interpretation. Sampling tools should be cleaned appropriately, packets labelled, handling recorded and suspected contaminant sources documented. Controls such as field blanks or background samples answer different questions: the first addresses introduced material; the second addresses the local environmental signal.

A credible laboratory record includes the instrument, calibration approach, soil preparation and whether results use oven-dry mass. Learners should explain how moisture and different grain-size fractions can bias an unstandardised comparison. Method details turn a surprising number into an observation other people can examine.

7. Read a chemical map as a set of measurements

Maps commonly interpolate between sample locations, but interpolation cannot discover a buried boundary that the grid never sampled. Colour choices, class breaks, unequal spacing and missing values can create visual drama. A red patch may be produced by one unusually rich sample surrounded by unsampled ground.

An effective teaching exercise displays the measured points first, then the interpolated surface. Students must mark where the map contains observed values and where it contains estimates. They should describe uncertainty and compare multiple classification schemes before drawing a wall or doorway on top of a phosphate hotspot.

8. Use an explicit fictional site: Orchard Court

Suppose a fictional Orchard Court survey uses one laboratory method and dry-soil concentrations expressed as milligrams of elemental phosphorus per kilogram of soil. At a comparable shallow horizon, a candidate courtyard hotspot contains 670 mg/kg; a nearby edge sample contains 310 mg/kg; a matched background control contains 290 mg/kg. A simple numerical difference of 380 mg/kg separates the hotspot from this one background control.

That difference is useful for describing the observations, but it is not an estimate of human-added phosphorus. The comparison has one control, no replicates and incomplete understanding of soil texture and post-depositional history. A careful student writes down those limitations before the map is allowed to suggest a particular ancient activity.

9. A second depth may change the story

Add a sample from the same map coordinate but a deeper horizon: 350 mg/kg at 40 cm, compared with 670 mg/kg in the 10 cm interval. One possibility is that a shallow concentration reflects later additions. Another is that the horizons differ in mineralogy or mobility. Without a profile, control locations and dates, the values cannot choose between these explanations.

Present the data in a table with depth and horizon labels rather than plotting a single x-y colour dot. Ask students to design the next test: several paired depths, better geological matching, dating of associated features and attention to later cultivation. The important lesson is that spatial maps and stratigraphic maps answer partly different questions.

10. Teach basic calculations without false significance

At Orchard Court, the ratio 670/290 is about 2.31. The words ‘2.31 times the matched control’s concentration’ are mathematically accurate for the observations as stated. The words ‘2.31 times the amount of ancient human activity’ are not. Concentration is influenced by processes other than human input and may not increase linearly with waste disposal.

Have learners calculate both the difference and ratio, then specify why neither resolves source attribution. Do not present a three-number class exercise as a statistically significant hotspot: a confidence interval, spatial model or hypothesis test requires enough relevant observations and a defensible sampling design.

Worked comparison: same site, different depth

11. Correct for soil and depth effects where appropriate

Some published approaches examine depth, organic content, sediment texture and phosphate retention to distinguish chemical background from unusual activity signals. A simple regression might help if its explanatory variables are measured and the data support the model. But an adjustment is not automatically valid because software can produce a colourful residual map.

Teach diagnostics: compare within soil horizons, inspect outliers, record laboratory uncertainty and evaluate whether a model’s assumptions are plausible. Students should be able to explain when stratified field comparisons are safer than an elaborate adjustment. Model sophistication cannot compensate for a poorly documented sampling campaign.

12. Interpret possible activity areas with independent traces

Concentrated phosphorus could be compatible with a refuse zone, a pen, food processing, burial or later fertilisation; each hypothesis predicts different associations. Ceramic assemblages, bone distribution, botanical remains, ash structures, site stratigraphy and phosphorus fractions can help test alternative interpretations. Even then, functional labels may remain provisional.

Construct a comparison matrix: the refuse hypothesis predicts discarded material; the animal enclosure hypothesis predicts suitable surface and dung indicators; modern fertilisation predicts an upper-horizon association or land-use record. A good archaeological argument survives attempts to distinguish these paths rather than merely finding one favourable observation.

13. Keep sampling and laboratory ethics in view

Soil sampling can disturb protected deposits and future research opportunities. Fieldwork should proceed with appropriate permissions, minimal disturbance, retained reference material and accurate spatial records. Students should be taught to avoid treating a heritage site as an expendable source of classroom chemicals.

The featured image on this page shows a genuine stratigraphic trench section from An Son, Vietnam. It illustrates why layering matters; it is not a photograph of phosphorus analysis at the fictional Orchard Court. Source and illustration must remain visibly separate.

14. Write a report that can be independently repeated

An effective data table includes sample identifier, coordinates, horizon, depth, mass, preparation, method, units, replicate information, blanks, analyst and measurement. The resulting map retains the raw point data, not just a smoothed illustration. Report modern land use, geology and any missing intervals so a future investigator can revisit the interpretation.

An excellent student conclusion would be: ‘one shallow location returned a higher concentration by the stated laboratory method; deeper and matched-background observations are needed to decide whether the anomaly reflects archaeological activity’. That is useful. It does not promise a vanished kitchen on the strength of three numbers.

15. Transfer to a hillside field with a different soil

Move the same lesson to a terraced hillside where samples cross a limestone-derived soil and an acidic clay, and where manure was spread recently. Even if laboratory measurements are precise, the comparison between the two soil types may be misleading. Students must redesign controls and consider whether the map mixes natural chemistry with land-use history.

The transfer objective is explicit: students should reuse the scientific logic while refusing to reuse an inappropriate numerical threshold. There is no globally valid phosphate concentration above which every site becomes an ancient settlement.

16. The teaching benchmark: reproducible data and proportionate claims

Assess four dimensions: the sampling design, method consistency, explanation of soil and depth effects, and the interpretation of anomalies. A convincing report should contain points and uncertainty, not only contours. It must name at least one plausible non-archaeological explanation and a feasible way to test it.

Students have learned the field when they can tell an archaeologist what a phosphate map can help find, what its chemical measurements cannot prove alone, and how to move from a survey hotspot to a well-controlled investigation.

A practical sequence for teaching the whole survey

• Define the archaeological question and plausible alternative explanations.
• Map a georeferenced grid and select depth/horizon intervals with matched controls.
• Document sampling tools, labels, blanks and laboratory extraction.
• Calculate concentrations and plot raw point values before interpolation.
• Compare soil texture, pH, land-use records and stratigraphic context.
• Draft a conclusion with a follow-up sampling design that could disprove it.

Reproducibility checklist

• Same chemical analysis method, dry-mass basis and explicit units for samples being compared.
• Recorded grid coordinates, depth, horizon and relevant local soil features.
• Adequate background locations and appropriate equipment/process blanks.
• Raw observations, replicate results and accessible processing decisions.
• Separate observed values from interpolated map predictions.
• A limited inference that identifies the next evidence needed.

Frequently asked questions

Sources and connected eduKate learning routes

• Crowther, Soil phosphate surveys: sampling, analysis and interpretation (Archaeological Prospection)
• Oonk, Slomp and Huisman, archaeological soil geochemistry and interpretive issues
• Geoderma, identification of archaeological depths through soil phosphorus prospection (2022)
• Historic England, Environmental Archaeology guidance
• The Gold Standard of Teaching Geoarchaeology
• The Gold Standard of Teaching Soil Micromorphology
• The Gold Standard of Archaeology
• Gold Standard master hub

The gold standard is a learner who can plan the grid, explain the laboratory number, identify the environmental alternatives and use the resulting map to ask a better archaeological question—not simply colour the highest square red.