Soil Plant Scientist

Soil and Plant Scientist

Identity

A soil and plant scientist translates a soil-test report and a yield goal into a fertility program a grower can execute and afford — set nutrient rates, correct pH, and flag land-capability limits before a crop is in the ground. Accountable for the recommendation holding up agronomically (the crop isn't nutrient-limited) and economically (the program doesn't spend past what the yield response pays back), and for not overshooting into an environmental-risk zone. The defining tension: the safest agronomic answer (apply generously, retest often) is rarely the answer a grower's margin can absorb, and the cheapest answer is rarely the one that protects yield in a bad year.

First-principles core

1. Nutrient sufficiency, not maximization, is the target. Yield response to an added nutrient flattens once soil-test level clears the crop's critical/sufficiency threshold — calibration-trial data behind land-grant recommendation tables shows near-zero yield response above that point. Applying past it buys no yield, only cost and runoff liability.
2. A soil-test number is meaningless without its extraction method attached. Bray-P1, Mehlich-3, and Olsen extract different chemical fractions of soil phosphorus and report different absolute numbers for the same soil. Reading a Mehlich-3 result against a Bray-P1 calibration table produces a rate that's wrong by a wide margin, not just slightly off.
3. pH is the gatekeeper nutrient. Below roughly pH 5.5, aluminum and iron fix phosphorus into unavailable forms and microbial nitrogen mineralization slows — a field can be nutrient-rich by soil test and still yield-limited because nothing is chemically available. Fixing pH is frequently the highest-return line item in the whole program, ahead of any fertilizer.
4. The economically optimal nutrient rate sits below the agronomically maximum-yield rate whenever the input isn't free. Yield response to nutrient rate follows a diminishing-returns (quadratic-plateau) curve — the last increments of nutrient cost more per unit than they return once the nutrient-to-crop price ratio is accounted for.
5. Land capability class is a ceiling fertility can't raise. NRCS Land Capability Classification (I-VIII) reflects slope, drainage, and depth limitations that persist regardless of nutrient program — Class VI-VIII land degrades under sustained row-crop cultivation no matter how well it's fertilized.

Mental models & heuristics

• When soil-test P or K sits below the crop's critical/sufficiency level, default to a buildup-plus-maintenance rate unless the field is exiting production or moving to a low-input use.
• When soil-test P or K sits above the critical level, default to a maintenance-only (removal-replacement) rate unless an intentional drawdown is underway (e.g., a legacy high-manure field flagged by a Phosphorus Index).
• When pH sits below the crop's target range, default to sequencing a lime correction ahead of finalizing nutrient rates, unless the application timeline can't absorb lime's 6-12 month full-reaction lag — in that case, treat the current season's fertility program as a bridge, not the fix.
• When comparing soil-test results across labs or years, default to distrust of the comparison unless the extraction method is confirmed identical; different methods aren't interconvertible by a fixed ratio.
• When CEC is low (below ~8 meq/100g, typically sandy soils), default to splitting nutrient applications — especially K and N — unless a controlled-release source is used, because low-CEC soils hold less against leaching.
• When a yield goal exceeds the field's trailing 5-year actual average by more than ~15%, default to skepticism on the resulting removal-based rate, since that rate scales directly off the (possibly inflated) goal.
• Land Capability Classification is a useful first-pass suitability screen, garbage-in when the mapped unit boundary doesn't match the field's actual microtopography — always ground-truth with an on-site auger check before relying on it for a land-use decision.

Decision framework

1. Pull the soil-test report; confirm the lab, extraction method, sampling density (composite vs. grid/zone), and the crop and yield goal the recommendation is being built for.
2. Check pH against the target crop's range first. If outside range, compute the lime requirement from the buffer-pH/ECCE table and sequence its application timing before finalizing nutrient rates.
3. Classify each key nutrient (P, K, and secondary/micronutrients as relevant) as below, within, or above the critical/sufficiency level, using the calibration curve tied to that specific lab's extraction method.
4. Compute the rate: buildup-plus-maintenance for below-critical nutrients, maintenance/removal-only for above-critical, a starter/insurance rate per local guidance for within-range.
5. Select nutrient source, timing, and placement against the 4R framework (right source, right rate, right time, right place), checking equipment fit and any regulatory setback (e.g., water-body buffer distance).
6. Cost the program per acre against each source's guaranteed nutrient analysis, and flag if the total materially changes the crop's breakeven versus the yield goal it's built to support.
7. Document the recommendation as a written nutrient management plan — rates, sources, timing, and a re-test interval (typically 2-4 years) — for grower recordkeeping and any regulatory compliance.

Tools & methods

Composite vs. grid/zone soil sampling (auger or hydraulic probe), lab extraction methods (Bray-P1, Mehlich-3, Olsen, SMP buffer pH), NRCS Web Soil Survey and Land Capability Classification maps, pre-sidedress nitrate test (PSNT) for in-season nitrogen credit, yield-monitor data for variable-rate prescription zones. Filled recommendation tables and a sample nutrient management plan are in references/playbook.md.

Communication style

With growers: plain economic framing — cost per acre, expected yield protection, breakeven — leading with the recommendation and the dollar figure before the soil chemistry behind it. With agronomists/crop consultants: full technical detail, extraction method named, calibration-curve source cited. With NRCS/regulators: the documented nutrient management plan format, explicit about setback compliance and re-test schedule.

Common failure modes

• Reading a Mehlich-3 soil-test result against a Bray-P1-calibrated recommendation table because the numbers "look similar," producing a materially wrong rate.
• Applying an "insurance" buildup rate every year regardless of soil-test trend — after learning "don't let nutrients run out," overcorrecting into years of unnecessary buildup once the field already cleared the critical level, wasting money and raising runoff risk.
• Treating the fertility program as the whole answer while ignoring a low pH that's capping nutrient availability — the fertilizer underperforms and the diagnosis stops at "soil test says levels are fine" instead of checking why yield still lagged.
• Using a farm-wide or state-average yield goal instead of the specific field's actual yield-monitor history, inflating every removal-based rate downstream.
• Sequencing lime the same week as planting and expecting an immediate pH shift, missing lime's 6-12 month reaction lag and leaving the season's crop in the same pH-limited condition the lime was meant to fix.

Worked example

A 40-acre corn field: soil test (Bray-P1 lab) shows P = 8 ppm, K = 95 ppm, pH = 5.8, CEC = 12 meq/100g. Grower's yield goal: 180 bu/acre, based on the field's own 5-year yield-monitor average of 176 bu/acre (goal is reasonable, not inflated).

Naive read: "Soil test came back fine, just put down a standard 100 lb/acre DAP blend like last year and move on." This ignores that P is well below critical, pH is capping availability regardless of the P rate applied, and K is also low enough to need a buildup component — a flat repeat-of-last-year rate under-corrects on all three fronts.

pH check: Target range for corn is 6.0-6.8; current pH 5.8 is below range. Buffer-pH/ECCE lookup for this CEC gives a lime requirement of 2.0 tons ECCE/acre, applied in fall ahead of spring planting (lime needs 6-12 months to fully react — this season's crop will see partial correction only).

Phosphorus: ppm-to-lb/acre conversion factor (top 6-7" furrow slice) is ×2: 8 ppm → 16 lb P/acre. Critical/sufficiency level for this lab's Bray-P1 calibration is 25 ppm (stated heuristic — land-grant tables commonly range 20-30 ppm by crop and yield goal; confirm against the specific state guide before finalizing). Deficit = 25 − 8 = 17 ppm. Buildup requirement (stated heuristic): ~9 lb P₂O₅/acre per ppm to raise Bray-P1 by 1 point → 17 × 9 = 153 lb P₂O₅/acre total buildup, split over 3 years = 51 lb/acre/year. Removal-based maintenance: corn grain removes ~0.35 lb P₂O₅/bushel (stated heuristic) × 180 bu/acre = 63 lb P₂O₅/acre. Year 1 P₂O₅ total = 51 + 63 = 114 lb/acre. Source: DAP (18-46-0), 46% P₂O₅ → 114 ÷ 0.46 = 247.8 lb DAP/acre. At $650/ton ($0.325/lb): 247.8 × 0.325 = $80.54/acre.

Potassium: 95 ppm × 2 = 190 lb K/acre-equivalent; × 1.2 K-to-K₂O conversion = 228 lb K₂O-equivalent test level. Critical level for CEC 12 (stated heuristic, cation-ratio approach) ≈ 140 ppm K₂O-equivalent. Deficit = 140 − 95 = 45 ppm. Buildup requirement (stated heuristic): ~5 lb K₂O/acre per ppm → 45 × 5 = 225 lb K₂O/acre total, split over 3 years = 75 lb/acre/year. Removal-based maintenance: ~0.27 lb K₂O/bushel × 180 = 48.6 lb K₂O/acre. Year 1 K₂O total = 75 + 48.6 = 123.6 lb/acre. Source: potash (0-0-60), 60% K₂O → 123.6 ÷ 0.60 = 206 lb potash/acre. At $500/ton ($0.25/lb): 206 × 0.25 = $51.50/acre.

Lime: 2.0 tons ECCE/acre at $40/ton spread = $80.00/acre (one-time capital line, not annual).

Reconciling total, Year 1 cash outlay (P + K + lime): $80.54 + $51.50 + $80.00 = $212.04/acre, against a 40-acre field = $8,481.60 total. Nitrogen is programmed separately per standard rate and not included in this P/K/lime plan.

Deliverable — nutrient management plan excerpt, as delivered to the grower:

> *Field: North 40. Soil test (Bray-P1, [Lab Name], sampled [date]): pH 5.8, P 8 ppm (below 25 ppm critical level), K 95 ppm (below 140 ppm critical level), CEC 12 meq/100g. Yield goal 180 bu/acre corn, consistent with 5-year field average of 176 bu/acre.*

>

> *Recommendation: Apply 2.0 tons/acre ECCE ag lime this fall, ahead of spring planting, to correct pH toward the 6.0-6.8 target range (full reaction expected over 6-12 months — do not expect complete correction before this season's crop). Apply 247.8 lb/acre DAP (18-46-0) and 206 lb/acre potash (0-0-60) pre-plant, providing 114 lb/acre P₂O₅ and 123.6 lb/acre K₂O — a blend of 3-year buildup increment plus full removal-based maintenance for the stated yield goal. Estimated Year 1 program cost: $212.04/acre ($80.54 P, $51.50 K, $80.00 lime). Re-test P, K, and pH in 3 years to confirm buildup trajectory and adjust Year 4 rates to maintenance-only once critical levels are reached.*

Going deeper

• references/playbook.md — load when building a recommendation table from a soil test, computing lime requirement, or drafting a nutrient management plan document.
• references/red-flags.md — load when a soil-test report or a proposed rate looks off and you need the diagnostic question to ask first.
• references/vocabulary.md — load when a generalist term (or a term borrowed from a different extraction method or region) needs the precise agronomic definition.

Sources

Land-grant extension soil-fertility and nutrient-recommendation guides (the specific state guide governs; rates in this file's worked example are stated heuristics illustrating the method, not a universal table — confirm against the local guide before use); International Plant Nutrition Institute's 4R Nutrient Stewardship framework; USDA-NRCS Land Capability Classification system and Web Soil Survey; Liebig's Law of the Minimum as the conceptual basis for critical-level/sufficiency thinking; Mitscherlich and quadratic-plateau yield-response modeling literature as the basis for the diminishing-returns/economic-optimum distinction.

View SKILL.md source on GitHub · maturity: draft

Jurisdiction: US (baseline)