Matt Greenspan

Sampling Session · Part 3 of 6 How I decide where to pull cores. A grid tells you how often to sample. It doesn't tell you where. So before anyone heads to the field, I stack maps over it: elevation and slope, where water runs from and where it runs to, and yield history. Then I sample on a modified grid, usually one sample per 2 to 2.5 acres, with points placed to represent the field's real environments. The swale that holds water. The knob that dries out first. The side slope in between. Those spots behave differently in wet and dry years, so they shouldn't get averaged into one soil test. Grid density matters too. University of Georgia found 1-acre grids matched lime, P and K needs far better than 5-acre grids. But density alone isn't the whole story. Placement is. No perfect method today, so we eliminate as much error as we can: where we sample, how we pull the cores, and how consistently we do it. How are you deciding where your sample points go?

Sampling Session · Part 2 of 6: soybean cyst nematode If your beans just came off, this is the best time all year to sample for SCN. Counts run highest from late season to shortly after harvest, and SCN can take up to 30% of your yield without you ever seeing a symptom. How I'd pull an SCN sample: • Right after soybean harvest • 15 to 20 cores, 6 to 8 inches deep, for every 20 acres • Walk a zigzag across the whole area. Don't just walk the end rows. • If the rows are still visible, probe within 2 inches of the row • Different soil texture? Pull a separate sample • Testing a field for the first time? Include the spots where SCN shows up first: near the gate, fence lines, areas that flooded, spots with pH over 7 • Mix the cores in a bucket, bag it in plastic (not paper) and keep it out of the sun Iowa State calls anything over 12,000 eggs per 100 cc of soil high if beans are next. At that level, think hard about another year of corn. Check which units your lab reports, too. Per 100 cc and per 250 cc aren't the same number. Why bother this year? Mizzou has been surveying SCN in Missouri for 37 years. In their 2024–25 samples: • 80% had SCN eggs • 86% of the populations they tested reproduced on PI 88788 at half or more of what they do on a susceptible variety. In 2005, that was 16%. • 54% were HG type 1.2, meaning they can reproduce on both PI 88788 and Peking. That's now the most common type they found. PI 88788 has been the go-to resistance for 30+ years, and that key is worn out. Peking still holds up better, but it isn't a sure thing anymore either. So before you lock in 2027 seed, ask your rep one question for every variety: PI 88788 or Peking? Then rotate sources between bean years and keep corn in the rotation. MU says the same: steward Peking in rotation with PI 88788 so both stay useful until new genetics show up. Iowa State also found SCN reproduces most in hot, dry years. If your beans went through a dry stretch this summer, that's one more reason I'd pull samples now. A reality check on seed treatments: a 51 site-year study across 13 states and Ontario found none of the six nematode-protectant treatments gave significant SCN control. ILeVO and Saltro were the only two that lowered SDS severity, and they gave the more consistent yield gains. I look at them as an SDS tool, not an SCN plan. What resistance source is in your 2027 lineup? Do you know yet?

Sampling Session · Part 1 of 6 Every variable-rate map we run starts with a few dozen bags of dirt. So I'm doing a series on soil sampling, and I'm starting with how the sample actually gets pulled. Here's why it matters. The lab tests a handful of points, and software (kriging, IDW, whatever your platform uses) fills in all the ground between them. So every point is speaking for a lot of acres. If one point is off, the map is off all around it. The math can't fix a bad core. What makes a point off? Mostly how it was collected. Depth. Under no-till, P builds up near the surface. In a 24-year Iowa State trial, no-till plots on one broadcast rate tested 82 ppm P in the top 3 inches and 13 ppm at 3–6 inches. Pull 4-inch cores one year and 8-inch cores the next, and the "change" you see is your probe, not the field. Cores. One core isn't a sample. MU says 10–20 per sample. K-State says 12–15 minimum and 20–30 for better accuracy. Placement. Cell sampling spreads cores across the whole cell. Point sampling keeps them close to a point (K-State says within 5–10 ft, NC State says 15–30 ft). Both work. Switching between them muddies the trend. Bands and rows. A core that lands in an old fertilizer band reads high. K-State has a paired-core method for when you don't know where the bands are. Timing. Soil-test K moves with moisture and crop residue. Iowa State says it's more stable in spring than fall, and MU says to sample at the same time of year as last time. The tools are changing too: hand probes, hydraulic probes on trucks and UTVs, and mechanical cross-row samplers. Radicle's GeoKnife is one to watch. It's in beta this fall, and Radicle says each pass pulls about 48 cores' worth of soil. Today I use ROGO to collect, and I stick with it for consistency. It's a robotic sampling service. Rogo says its self-driving robot uses RTK-level GPS to reach each core location and tracks samples with barcodes. There's no perfect method today, so we need to eliminate as much error as possible. Whatever you use, write the method down and repeat it. My routine: 0–8 inches, same depth and same time of year every cycle. What's coming: 2. Fall SCN sampling 3. Grid size 4. Letting the landscape pick where 5. What happens at the lab 6. From report to prescription What's one thing about how your samples get pulled that you'd change?

Soybean roots nearly disappear just above the claypan. That's not a figure of speech. On Centralia claypan ground, depth to that clay ran about 12 inches on the hilltops, 4 inches on the eroded hillsides, and up to 40 inches at the bottom of the slope. Roots were thick near the surface, then almost vanished right above the claypan - then showed up again 8 to 16 inches below the top of that clay, about 20 to 30% thinner. And the water story is even simpler: 6 inches of topsoil holds about half the water of 39 inches. Corn struggled where topsoil was under 6 inches. That same field had lost about 5 inches of topsoil over roughly 150 years - as much as 12 inches on the north end. So when a yield map stays stuck on those shoulders, it isn't always begging for more P and K. You can't fertilize a 4-inch topsoil shoulder into a 40-inch toeslope. Not every low zone needs more P and K. USDA-ARS / MU

That spring K test can look fine - until July shows up. Here's the part a lot of people miss: a regular soil-test K reading only sees what's in solution and hanging on exchange sites. It doesn't see the K stuck between 2:1 clay layers. On high-smectite ground, those layers can collapse when the soil dries and trap that K for a while. NDSU split their corn trials at a smectite-to-illite ratio of 3.5. Below that, the critical mark was 150 ppm. At 3.5 or higher, it jumped to 200 ppm. On those high-smectite sites, a dry July or August still threw K-deficiency symptoms and more than 20 bu/ac of response - even when spring tests were already above that old 150 ppm bar. Our Mexico and Adco claypans are Fine, smectitic, and that clay layer is usually more than 50% clay. So a grid that "looked fine" in a wet fall can still get cranky on those eroded shoulders when July dries out. Fertilizer as a whole is a gut punch right now. DTN's latest national retail survey had anhydrous at $1,007 a ton and MAP at $987. Potash is still the reasonable one in that lineup, at $510. If a dry July is leaving K bushels on the table, this isn't the year I'd skip it just to save a little money. Not a new Missouri critical. MU and Tri-State don't split K recs by clay type. Just a reason to keep an eye on those shoulders. NDSU / Franzen

Saturation Risks: why it matters. Warm soil + standing water = N gone fast. Above 60°F, full of water 2–5 days, 50%+ of N can be lost. At 55–60°F: ~1–2% of nitrate a day. At 65–70°F: ~4–5% a day. Only nitrate is at risk. Ammonium holds.

Fall burndown sets up a clean start in spring. Henbit comes up mostly in fall and is a strong SCN host. Fall control of emerged marestail is more effective and consistent than spring burndown. Spray on a 50°F+ day. Still scout before planting.

Your N bill comes with a lime bill. MU Extension: every 100 lb of N from NH3, urea or UAN takes 72 lb ENM of lime to neutralize. Ammonium sulfate? 220. So 180 lb N/ac as NH3 ≈ 130 lb ENM/ac. Compare lime quotes by ENM, not by the ton. extension.missouri.edu/publi…