Fertility, cover crops and the soybean canopy — from the ground and from the air
Data & prepStatistical testVerificationAgent-initiatedCaveat
The analysis pipeline
Drag to pan · scroll to zoom · click any node to open it on the right
100%
Key take-aways
what the analysis concluded — click a node for the working
14 steps
Agentic analysis · Pontotoc soybean trial, 2022
Fertility, cover crops and the soybean canopy — from the ground and from the air
60 plots at Pontotoc, Mississippi: 3 fertility sources × 5 cover crops × 4 blocks, measured at soybean stages R3, R5, R6 and R7
with a ceptometer (leaf-area index), destructive sampling (five-plant dry weight) and one UAV flight per stage (canopy cover + 13 spectral indices).
An AI agent was given the seven raw workbooks and the three objectives below — nothing else — and asked to work
through them. This page is the full record of what it did, including the steps where it stopped itself.
Objective 1Do fertility source (none, synthetic fertilizer, poultry litter) and cover-crop species affect soybean canopy development — leaf-area index and above-ground biomass — across reproductive stages R3 to R7?
Objective 2How well do UAV-derived canopy cover and vegetation indices track the ground-measured LAI and biomass, and which index works best at which stage?
Objective 3Can the UAV indices detect the same treatment effects that the ground measurements detect — could a flight stand in for destructive sampling?