Histories of forensic botany, however, typically begin in the 1930s and with its role in solving “the crime of the century.” In 1932, famed aviator Charles Lindbergh’s 20‑month‑old son was kidnapped from his nursery in rural New Jersey and later found dead in the woods. One of the few clues was a homemade wooden ladder left beneath the child’s second-story window. When police ran out of leads, they sent the ladder to Arthur Koehler, a wood specialist at the U.S. Forest Products Laboratory in Wisconsin.

Under his microscope, Koehler identified four kinds of wood in the rungs and rails—Douglas fir, two types of pine, and birch. He also analyzed tool marks and annual growth rings in the material. Using this data, he traced some of the wood to a shipment sent to a Bronx lumberyard near the home of a carpenter named Bruno Richard Hauptmann. When investigators searched Hauptmann’s attic, they found a floorboard had been cut away. One of the ladder’s side rails matched the wood around the gap precisely, ring for ring and knot for knot. In court, Koehler’s testimony that the ladder and the attic board had once been a single piece helped convince the jury that Hauptmann had built it himself.  In 1936, Hauptmann went to the electric chair.

Since then, plant evidence has helped investigators solve cases in varied ways.

In the Bahamas, grass fragments in a suspect’s socks and shoes linked him to a golf course where a 19-year-old woman had been raped and murdered, helping to convict him and an accomplice. In Italy, bits of moss on a woman’s shoes and on the parapet and service stairs of a shopping center allowed botanists to reconstruct that she had climbed the stairs, stepped onto the parapet, and jumped, making suicide far more likely than a push or struggle.

In Denver in the 1980s, a botanist identified fragments of red cabbage, kidney beans, and onions in a murdered woman’s stomach, clearing her boyfriend, who’d taken her to a burger joint for lunch but had an alibi for the rest of the day. Years later, a serial killer confessed to her murder—and to taking her to dinner in a restaurant with a salad bar beforehand.  And when Sikh terrorists were suspected in a 1985 bombing at Tokyo’s Narita Airport, dust from cocoa shells used as filler in the dynamite helped pinpoint where the explosive had been bought, contributing to the conviction.

Yet for all its usefulness and prominence in true crime podcasts and TV shows, the field has remained niche.

When Koehler testified at the Lindbergh trial, he had just a handful of other forensic scientists to call peers. Since then, forensic sciences have exploded as a field, from the routine use of fingerprints and blood typing to the rise of crime labs, ballistics, and, of course, DNA profiling. But plant evidence has remained the province of select specialists. “There’s probably less than 10” in the world, says Mark Spencer, a forensic botanist in the United Kingdom.

Why don’t crime investigators make more use of this evidence? It’s “not because they’re bad or incompetent,” Spencer once suggested. “They’re just not used to dealing with this weird green stuff.”

IF EVER THERE WAS A PLACE for a renaissance in forensic ecology, it’s Florida. The state’s agricultural industry has given it deep botanical expertise and a reputation among specialists in the field for being something of a hot spot for forensic botany.

But ask Jason Lewis, chief of homicide in the Daytona Beach State Attorney’s Office, how many times plant evidence has come up in the roughly 1,000 murder cases he has prosecuted or supervised during the past 25 years, and the answer is exact: one.  

The 2018 case involved a 16‑year‑old girl who had been raped and killed, then left in a thicket of muscadine grapevines and live oaks. After her body was found, investigators noticed leaves in the truck of her mother’s boyfriend. The prosecution called in Hardy, who testified that the leaves’ distinctive shapes matched the plants in that hidden grove. A jury convicted the man of first-degree murder.

“I think the jury was really fascinated by it,” Lewis recalls. “It was pretty exciting to use science that we learned in seventh and eighth grade to prove some element of the crime that had occurred.”

But for every investigator who notices plant evidence, countless others do not, botanists say. Botanists lament what they call “plant blindness.” The simple fact is most of us don’t register the greenery around us. “Human beings are terrible at visually recording green things as being entities in their own right. They’re just blobs to us in many ways,” Spencer once said.

Investigators who do look for plant evidence face another obstacle. There is often almost too much of it.

At the culvert where Troy’s body had been found, Wiltshire cataloged the vegetation zone by zone: the holly growing alongside the path, whose pollen made it a marker for contact with that specific spot; the bramble and privet hedge banking one side; the poplars, willows, hawthorn, and cherry laurel on the other; the carpet of rotting leaves in the area leading to the retaining wall. “It was the whole pattern that mattered rather than the individual plants,” she later wrote. Two plastic trash bags of garden debris lay on the path as well, long since split open, the contents rotted to compost—impossible for anyone approaching the culvert to avoid.

Police had seized the sneakers and jeans of Troy’s mother and father as well as the shoes of his grandmother. Wiltshire analyzed each for pollen. She took samples in the park where Danielle and Sherwain were known to buy drugs and from the sidewalk outside the post office where they collected their welfare payments to rule out those locations as the source of whatever she might find on their clothing.

In the morgue, she washed Troy’s hair and swabbed deep into his nasal passages, coaxing out whatever microscopic traces remained of the last places the baby had been. Collecting and mapping the samples took several days. Wiltshire then treated them with chemicals to strip away mud, fibers, and other debris, stained the now-isolated pollen and spores, and fixed them in gel on glass slides. Finally, she slid them under a microscope and started counting.

Pollen is particularly useful for linking people and objects to places because this reproductive grain of shrubs, flowers, and trees is wrapped in a hard casing that can survive washing machines, cling to hair for millennia, and persist in soil long after everything else has decayed. It’s also abundant. Plants produce it in staggering quantities, sometimes several hundred million grains in a season. And while to laypeople all pollen may look like yellow dust, under a microscope it reveals extraordinary variation: some species produce tiny spheres bristling with spines; others dimpled orbs or dumbbells dotted with holes. The specific mix of flora in an ecosystem leaves a signature—a botanical fingerprint unique to each place.

These pollen signatures can be hyper-precise, as Wiltshire has found repeatedly. A pollen profile from one side of a flower bed can look entirely different from one taken a few yards away. It can act as a “clandestine record of where you have been and what you have been doing,” she writes.

Plants can also answer a different question: when.

Forensic botanists have estimated how long skeletal remains have lain in the wild by gauging the age of the mosses growing on the bones or the maturity of a shrub that has pushed through the rib cage. They have determined when a body was concealed from the degree of wilting of the plants used to cover it. They have measured when a perpetrator crossed a patch of weeds based on how thoroughly the trampled plants have recovered.

In some cases, Wiltshire has spent weeks hunched over her microscope, analyzing and cataloging thousands of pollen grains, plant and fungal spores, and other microscopic traces in her samples, with just her cat for company and classical music in the background.  In Troy’s case, however, something jumped out quickly.

The grandmother’s shoes were almost bare of pollen, consistent with someone who rarely strayed off paved surfaces. And the mother’s sneakers yielded only modest traces. But Sherwain Smith’s jeans and sneakers were teeming with spores and grains.

For a man whose world, by all accounts, extended little further than the local drug dealer and the post office, he had somehow accumulated what Wiltshire called a “veritable botanic garden-worth” of pollen: aquatic plants, riverside trees, garden exotics, wasteland weeds—a range so extravagant it could only have come from one place. The split trash bags rotting on the path to the culvert yielded an almost identical profile.

To top it off, there was one fungal spore that turned up in exactly two places, and only two places, in all of Smethwick: at the culvert and on Smith’s clothing.

STUDIES HAVE FOUND THAT forensic evidence of any kind is collected in only about a third of serious crimes. The labs tasked with analyzing what little is collected are already overwhelmed—most of their work is toxicology and drug testing, with DNA analysis eating up much of the rest. In that environment, something as labor-intensive and rare as forensic botany barely gets a foothold.

There are no formal training programs in forensic botany,  and it’s hard to imagine what one could look like. The prerequisite body of knowledge is vast. Excluding algae and mosses, roughly 400,000 plant species exist in the world. New ones are discovered regularly.

Even in the United Kingdom, a fairly species-poor region, crime-scene surveys routinely turn up more than 50 distinct plant species; in more biodiverse regions, the number climbs considerably higher. Forensic training programs might touch on botanical evidence, but there is no certification system and no standardized curriculum. As Wiltshire has written, mastering this field requires years of accumulated experience that no degree program currently provides, making it categorically different from DNA analysis, which can be taught to technicians running commercial machinery guided by software. For that reason, careers in forensic ecology usually start unscripted, typically when police recruit a university biologist for a specific assignment. As Wiltshire puts it, “This kind of work is not for the beginner.”

The plants themselves don’t make it easier. Botanists traditionally identify plants by appearance, which is trickier than it sounds. A plant’s appearance can vary dramatically depending on its location and growing conditions. Leaves may shrink during a drought; their shape can vary based on humidity, elevation, and light. A plant collected in the Midwest can look entirely unlike the same species grown in the Northwest, says Heather Coyle, a forensic scientist at the University of New Haven and editor of one of the few textbooks on forensic botany. “It will look like a different plant, but it’s actually the same.”

This makes identifying plants “an overwhelming task, even for someone who has that background,” says Coyle, especially when plants found at crime scenes are damaged, fragmented, or covered by bodily fluids.

Recently, investigators have turned to plant DNA to help overcome some of these challenges. But in any species, DNA identification works by comparing a pre-determined segment of the genome—called a genetic barcode—and many plant species are too closely related to make that reliable. Two plants might be classified as “different species because the flower color is different,” says Hardy: But “the gene that I go after to identify it isn’t necessarily the gene that’s involved” in determining the color, meaning the test can point to the wrong species.

Additionally, only about a quarter of the world’s plants have established barcodes, says Coyle. Since plants are often hard to identify, an unknown number of those in existing genetic databases may be mislabeled, which means any positive identification must be verified by other means.

When a forensic botanist needs to do more than identify a species—linking, say, leaves found in a car to a particular tree—they often have little choice but to build an entire database from scratch,  as investigators in Arizona discovered in 1992.

On a Sunday morning in May of that year, a dirt biker found the naked, strangled body of a woman lying in the brush near a cluster of palo verde trees west of Phoenix. A pager near the body led police to a man whose pickup had been seen in the area. The man admitted knowing the woman but swore he had not been to the crime scene.

One of the palo verde trees near the body bore a fresh scrape on a low branch, and two palo verde seed pods turned up in the bed of the man’s truck. Investigators sent them to Tim Helentjaris, a plant geneticist at the University of Arizona to ask if they could have come from the scraped tree. Using a then‑new technique called randomly amplified polymorphic DNA, or RAPD, Helentjaris compared DNA from the truck pods with pods he sampled from every palo verde at the site and dozens more around Maricopa County. He concluded that the two from the truck matched the scraped tree’s genetic fingerprint. The odds that they had come from another tree were roughly one in a million, Helentjaris testified in court. The jury convicted the man of first-degree murder.

Palo verdes have high genetic variability, which made the one-in-a-million figure possible. Other species are more uniform. Years ago, Adrian Linacre, chair in forensic DNA at Flinders University in Australia, was asked whether pine needles found on a body in Norway had come from a specific tree. He sampled several hundred pines in the area—at considerable cost, he recalls. But when he ran the needles against this genetic database, the odds that they came from that specific tree, rather than any other in the area, were only 10 to 1. Not very impressive, he admits. Doing better, however, would have meant sampling most pine trees in the country. “There’s probably 100 million pine trees in Norway,” he says.

Even when botanical evidence is gathered, it rarely makes it to court. Some cases never go to trial because police fail to identify a suspect. But prosecutors may also set plant evidence aside if they have stronger options. Frequently investigators use plant evidence “as a stepping stone to get search warrants,” says Coyle. “Then if they find a pair of shoes or a gun, they don’t really need it anymore.”

Courts require scientific evidence to be testable, peer-reviewed, and widely accepted. Forensic botany sometimes struggles on these counts. Being so small, the field lacks uniform standards for collecting, preserving, and analyzing evidence. If a judge feels the science involved lacks court precedent, they can force an admissibility hearing before the evidence can be presented, says Coyle. If the prosecutors “have other evidence, they’re not going to bother.”

In Troy’s case, Wiltshire did testify. She arrived at court uneasy. Her pollen analysis had become a cornerstone of the prosecution’s case. “It had been impressed upon me that there was a lot riding on it,” she later wrote. Her testimony helped convict Smith of manslaughter. He received an indefinite sentence in 2007, although he would be released in 2024.

Wiltshire is haunted by what was left out of the trial, however. She had found cellulose fibers—the kind in pillows and cotton blankets—in Troy’s nasal passages. That evidence had not been submitted because Wiltshire’s collection method wasn’t court-recognized. Cellulose fibers don’t ordinarily travel deep into the nasal passages, she notes—unless a baby struggles desperately for air. She has wondered whether Troy was suffocated and whether the verdict might have been murder had that evidence been heard.

For now, such questions have no good answer. But some experts believe the field’s tools are finally catching up to its ambitions.  AI algorithms, trained on databases of plant and pollen images, can already identify samples and, when paired with automated microscopes, scan slide after slide, flagging individual pollen grains without human assistance.

Equally promising is environmental DNA, or eDNA. Cheaper and faster sequencing has made it possible to process every organism in a sample at once. Since every living organism sheds cells into its surroundings, a scraping of soil carries the collective genetic signature of every plant, fungus, microbe, and animal present. In principle, the technology could allow investigators to link a suspect to, say, a burial site through the ecological signature of a pinch of dust on his cuffs without an expert laboriously identifying each species.

But this would require updating and expanding existing databases to give the algorithms something to compare. And that would require researchers willing to do the painstaking work of cataloging and money to fund them. Both are scarce. “Why would you get a grant from taxpayers’ money to sample mosses in the outback of Australia?” says Linacre.

What the field needs, he suggests, is a single galvanizing case; something dramatic enough to make the absence of that expertise impossible to ignore. He imagines a high-profile crime hinging on a tantalizing clue—a plant fragment on a glove, a smear of soil on a shoe—that could have cracked the case open. “We need that exposure,” he says. The “realization that that expertise would be very valuable, and we should have it now.”