EXPOSED
The modal soil and plant scientist splits time between field sampling and greenhouse/plot trials on one side and literature review, statistical analysis, and report writing on the other — the second half is squarely in AI's wheelhouse, and satellite/sensor data plus predictive models are already replacing routine fertility recommendations and yield forecasting. What persists is physically getting into the field: pulling soil cores across variable terrain, diagnosing a wilting stand in person, running and troubleshooting multi-season breeding or agronomic trials. Licensure is thin (a handful of states register soil scientists; CCA and CPSS are voluntary), so there's no regulatory wall — the moat is embodied fieldwork and grower trust.
Headcount grew steadily across the period.
Median pay $63,200 → $78,850 -0.2% in real terms
This line is counted by the Bureau of Labor Statistics — the one figure on this page that isn't a judgement of ours. Headcount moves on demand, offshoring, demographics and the business cycle, and automation is one term among several, often not the loudest.
So a falling line is not evidence that AI did it, and a rising one is not evidence that it won't. Both happen in this register: some occupations resist automation and shrink anyway, others are highly automatable and keep growing. The marked year is 2020.
BLS projection, 2024–2034
+5.4%
Percentage only. The projection counts a different population from the 15,730 above — it includes self-employed workers, which for this occupation is most of them, so the two headcounts are not comparable.
Growing, and only partly exposed
The BLS expects +5.4% more of these jobs by 2034, and at 50/100 the work is only partly exposed — some tasks are automatable, the core of the job is not. Nothing here is in tension.
Different clocks. The score is what current AI could do to this work today. The projection is how many of these jobs will exist in 2034. Everything between the two — how fast employers actually adopt, whether demand grows in the meantime — is why they can point opposite ways without either being wrong.
~1,700 openings a year on average, including replacing people who leave.
ArboristBotanistScientistAgronomistPomologistSoil ExpertApiculturistCorn BreederEntomologistAgriculturistPlant BreederViticulturistCotton BreederFloriculturistHorticulturistArboriculturistPlant AnatomistPlant ScientistSoil SpecialistField AgronomistGrowth SpecialistPlant PathologistArboreal ScientistPlant Physiologist
Holding it up: task resistance . Weakest point: liability shield .
Mixed — a routine tier and a judgment tier Running a three-year variety trial or a replicated fertility strip trial — laying out plots, calibrating the plot planter, harvesting and weighing by treatment, tracking treatment drift — is not compressible, but the ANOVA, the mixed-model output, the literature synthesis, and the extension bulletin draft are, and routine soil test interpretation to a fertilizer rate is already algorithmic under most state recommendation systems, which puts the split near even rather than into resistance territory.
Some physical or field component Probe work in a wet spring field, digging and describing a soil pit to horizon, scouting for nematode or Phytophthora symptoms in a standing crop, and greenhouse work are all genuinely physical, but much of it is scheduled sampling in mapped locations that sensor networks, drone imagery, and autonomous samplers increasingly cover — that's a 12, not the 16 you'd give a field agronomist who must respond to an unpredictable diagnosis call every day.
Certification preferred, not legally required CPSS and CCA are credentials employers ask for, not permission slips, and only a small set of states (Texas, Wisconsin, Minnesota among them) register soil scientists — mostly for septic and wetland delineation work, which is the narrow slice where a signature and seal actually attach and why this isn't a 2.
Meaningful discretion Deciding to terminate a breeding line, calling a soil unsuitable for a proposed effluent application, or recommending against a $60/acre input on marginal data are real judgment calls with money attached, but they sit inside experiment station protocols, state recommendation frameworks, and peer or committee review, so almost none are made alone and irrevocably.
The verdict above describes this occupation as a whole. Almost nobody does the typical version of a job — tick what's actually in your week and see how your own mix sits.
Your task mix speaks to task resistance (12/20 here) — how much of the day's work current AI already does. That is the dimension the boxes above are about.
It cannot move the other three. Liability shield (6/20) is whether the law requires a licensed human to sign. Trust premium (9/20) is whether buyers specifically pay for a person. Judgment and accountability (11/20) is whether the role exists to own consequential calls. Those are facts about the occupation's standing, not about which tasks are in your week — a paralegal who does only trial exhibits still holds no licence. Together they are 26 of this occupation's 50 points (52%).
Embodiment (12/20) is also a property of the work rather than the worker, but we don't tag individual tasks as physical or not, so the picker can't tell you anything about it. That's a limit of this tool, not a claim.
Did we get the list right? Tell us what's missing — the tasks are written from the outside, and you're reading this from the inside.
Foresters EXPOSED
The moves above are yours to make. This is the other half: what would have to change in the world for the occupation itself to score higher. None of it is in any one person's gift, but it is where the floor actually comes from. Scores here are not a one-way ratchet. Only two of the five dimensions — task resistance and embodiment — track what machines can do. The other three track law, what buyers will pay for, and who is answerable, and those move in both directions, often in response to the same pressure AI creates. If every lever below landed, this occupation would score around 70/100 — SAFE.
Genuine two-tier structure: routine fertility recs and yield forecasts fall to satellite/model pipelines, leaving the residual job as designing multi-season trials, diagnosing anomalous failures where model priors are wrong, and calibrating/ground-truthing the models themselves. If employers restructure roles around trial design and model validation rather than recommendation output, the remaining task mix is substantially less automatable.
Soil carbon credit verification protocols (Verra VM0042, CAR Soil Enrichment) requiring in-person, chain-of-custody physical core sampling by a qualified professional rather than remote-sensed estimation. If registries harden against model-only quantification after fraud concerns, physical sampling volume rises.
State-level mandatory licensure for soil scientists with statutory sign-off authority — the model already exists in the ~5 states (e.g., North Carolina's Soil Scientist Licensing Board, Texas, Tennessee) where a licensed soil scientist must seal septic/on-site wastewater suitability evaluations and wetland delineations. If more states adopt this, or if EPA/state agencies extend a seal requirement to nutrient management plans and CAFO permits under CWA, a personally-liable human signature attaches to work AI can otherwise draft.
Crop insurance and lender requirements: if RMA or farm credit institutions require a CCA/CPSS-certified professional to attest to yield-loss appraisals or prescription-based variable-rate plans used for actual production history claims, certification acquires legal teeth it currently lacks.
Regulatory tightening on nutrient loading — e.g., Chesapeake Bay TMDL-style nutrient caps, or EU-style nitrate directives adopted at state level — puts a named scientist's professional call between a grower and enforcement exposure, making the recommendation a consequential decision under ambiguity rather than an advisory number.
Narrow route only: independent agronomists who take no input-vendor commission sell verifiable freedom from conflict of interest — something a retailer's AI recommendation engine structurally cannot claim. Growth of fee-for-service independent consulting (visible in the Independent Professional Seed Association / independent crop consultant associations) would raise this modestly, but the buyer is price-sensitive and this will not scale far.
The limit. Realistic ceiling is mid-60s. The single largest available gain is licensure, and it is geographically fragmented — a handful of states, mostly tied to septic and wetland work, not to agronomy broadly. Absent a national licensing push (no professional body is currently mounting one at scale), the moat stays embodied fieldwork, and fieldwork is the part most exposed to sensor and autonomous-sampling hardware over a decade.
| Des Moines-West Des Moines, IA | 430 | $132,290 +68% |
| Minneapolis-St. Paul-Bloomington, MN-WI | 400 | $79,610 +1% |
| Boise City, ID | 290 | $90,610 +15% |
| Portland-Vancouver-Hillsboro, OR-WA | 250 | $94,940 +20% |
| Chicago-Naperville-Elgin, IL-IN | 240 | $80,360 +2% |
| Fargo, ND-MN | 230 | $65,320 -17% |
| Washington-Arlington-Alexandria, DC-VA-MD-WV | 230 | $95,870 +22% |
| Fort Collins-Loveland, CO | 220 | $67,330 -15% |
| Des Moines-West Des Moines, IA | 430 | $132,290 +68% |
| San Jose-Sunnyvale-Santa Clara, CA | 70 | $116,660 +48% |
| Salinas, CA | 30 | $115,910 +47% |
We have no reported case of a named organisation automating this occupation. Not one deployment, not one announcement.
That is worth saying out loud next to a score of 50. The verdict above is about what the work exposes — what current AI could do to these tasks. It is not a claim that anyone has done it. For this occupation those two things have come apart completely: the capability argument is on this page, and the evidence column is empty.
Has AI actually changed your work? One tap, anonymous, and the running tally is public. Nothing else is asked of you.
Rather than check back: get the digest and we'll tell you what changed — or watch a single occupation from its own page.