Pollen as Trace and Provenance Evidence

Every time a person walks through vegetation, sits on ground, or is exposed to outdoor air, pollen grains deposit on their clothing, hair, and skin. The assemblage they carry is a cumulative record of environmental contacts, weighted toward the most recent and most intense exposures. When that person moves to a new environment, a new layer of pollen adds to what is already there. Careful sampling can sometimes tease apart these temporal layers, though usually the signal is a blended composite.

The forensic argument from pollen trace has a structure. The assemblage on the exhibit is compared against a reference assemblage from the suspected location and against assemblages from the suspect's alternative locations (home, workplace, regular routes). If the exhibit assemblage matches the suspected location and does not match any of the alternatives, and if the matching taxa are not so common that chance deposition from the background air explains them, then the conclusion can be drawn that the exhibit was in the suspected environment.

In 1981, a Swissair aircraft was hijacked during a flight and the passengers held for a period before the situation was resolved. Swedish authorities developed evidence pointing to involvement by Magnus Olof Nilsson, a Swedish academic. Nilsson denied being at the relevant locations, claiming to have been in Sweden throughout. Investigative palynologist Goran Henkel was asked to examine pollen from Nilsson's clothing.

The assemblage on Nilsson's clothing included pollen types from tropical and subtropical plant genera not flowering in Sweden at that time of year and not found in Swedish ambient pollen rain. These types were consistent with the vegetation zones relevant to the hijacking itinerary. The presence of these geographic and seasonal markers, incompatible with Nilsson's claimed movements, contributed to a prosecution case that used pollen evidence to reconstruct geographic itinerary at an international scale.

What made the case significant was the demonstration that pollen assemblage comparison is not limited to local scene-to-suspect linkage. When pollen types diagnostic of specific vegetation zones are found on clothing in combination, a forensic palynologist can make geographic arguments spanning thousands of kilometres. The methodological requirement is a good reference dataset for the relevant vegetation zones.

Patricia Wiltshire, the leading UK forensic palynologist, has documented a case in which a suspect was implicated in a burglary not by footwear marks or fingerprints but by the pollen on his clothing. The suspect had concealed himself in the ceiling space above a room during the course of the break-in. The ceiling void happened to be above a room in which a houseplant produced an entomophilous pollen that was not found in the air outside the building or in any location the suspect acknowledged visiting.

The pollen from this plant, combined with the distinctive building dust and organic debris characteristic of that particular ceiling space, produced an assemblage that was effectively unique to that specific micro-environment. When the same combination was recovered from the suspect's clothing, the inference was that the clothing had been in that ceiling void. The case illustrates how fine-grained spatial specificity can be when entomophilous pollen and a defined physical context combine.

Undisturbed soils accumulate pollen in a predictable stratigraphic sequence. Deeper layers contain older pollen assemblages reflecting past vegetation; shallower layers carry more recent pollen. This sequence is what Quaternary scientists read as a vegetation history archive. In forensic contexts, any feature that disrupts this sequence becomes detectable by systematic sampling.

When a clandestine grave is dug, the soil is excavated and backfilled. This mixes the stratigraphic sequence: material from depth appears at the surface, surface material is buried with the body, and the clear age-depth relationship breaks down. A pollen profile taken through the suspected grave fill will show a disordered assemblage, with pollen types from different depth horizons mixed together. Comparison with profiles from undisturbed soil immediately outside the suspected grave provides the baseline for detecting this anomaly.

The technique works best where soils preserve pollen well (acidic, moist conditions) and where there is enough of a pollen history in the profile to generate a detectable contrast. Sandy, alkaline soils with poor preservation may not yield enough pollen at depth for the comparison to work. In these cases, supplementary methods (geophysical survey, vegetation anomaly remote sensing) are needed. But where conditions are good, pollen stratigraphy profiling has successfully identified clandestine graves that other search methods had missed.

Plants flower in predictable seasonal windows that vary by species and latitude. In temperate Europe, hazel flowers January to March, alder January to April, birch March to May, and oak April to June, with grasses May to August and Artemisia (mugwort) July to September. If a clothing exhibit carries an assemblage dominated by Corylus (hazel) catkin pollen at high concentration, it is most likely that the clothing was heavily exposed outdoors in late winter or early spring.

This seasonal inference is useful but carries caveats. Pollen persists for months on clothing, so a summer exhibit can carry spring pollen from an earlier exposure. Dried flowers, herbal products, and stored items can introduce out-of-season pollen. And flowering phenology shifts with climate, meaning that published phenological windows represent averages from which any given year may deviate by weeks. A seasonal inference is best treated as a supporting line of evidence rather than a stand-alone conclusion.

A complete forensic pollen provenance argument has several components. First, characterise the exhibit assemblage: what taxa are present, in what proportions, and what does that combination indicate about habitat and season? Second, characterise the scene control: does the scene carry the same assemblage? Third, characterise the elimination controls: do the suspect's alternative environments carry the same assemblage? Fourth, assess the rarity of the match: could background airborne deposition explain the exhibit without the suspect ever being at the scene?

The strength of the conclusion depends on how many of these steps work in the analyst's favour. A match on a single common anemophilous taxon with no elimination samples is essentially meaningless. A match on a combination of three or four low-background entomophilous or habitat-specific taxa, confirmed by scene controls, contradicted by all elimination samples, is a strong provenance argument. Most real casework falls somewhere between those extremes.