The laboratory cannot fix the sample

A bunker, hay lot, commodity bay, and TMR are not one feed. They are layers, wet pockets, fines, and faces that changed since yesterday. One convenient grab can overrepresent the pretty top, the exposed shoulder, or a single bale. The laboratory can repeat its method on what arrived in the bag. It cannot recover the parts of the lot that never entered it.

Penn State’s forage-testing guide treats the sample as the first method: one lot is one field, cutting, and harvest date — not a mixed week of bales. For small-square hay, core at least 20 widely separated bales; for large rounds, sample 10 to 12 bales through the layers, not a flake off the end. Combine cores in a clean bucket, label the structure and date, and keep wet silage cold. The practical lesson is that sampling error begins on farm, before analytical error begins in the laboratory.

Sister notes on this universe cover compare-dry-matter-not-water and a-number-without-a-unit-is-not-a-result — a clean bag still has to carry the right basis and unit. The physically-effective-fiber universe’s shake-the-bunk-not-the-memory note is the same idea at the mixer: the result represents what you sampled, not what you wished was in the pile. Educational only — not a diagnosis, prescription, or ration formula.