A surveillance-based interpretation of commercial soybean reception reveals season-dominated organization across 16 harvest seasons

Postharvest science is built mainly on controlled experiments that resolve how management changes lots after intake. Commercial laboratories, meanwhile, continuously generate standardized physical, compositional and sanitary observations that, to our knowledge, are seldom treated as scientific evidence for population-level inference. We propose Surveillance-based Postharvest Science (SPS) as a potentially transferable interpretive framework whose first empirical application is evaluated here: reception records are read as a storage-entry state—an integrated analytical profile at the start of the postharvest chain—whose organization becomes visible when long-term archives are analysed as integrated observational systems. SPS rests on five principles (Figure 1) and four formal constructs: storage-entry state, year-conditioning, trait-specific seasonal signal and climatic readability. Using 2562 commercial soybean lots from a Mato Grosso reception network across 16 harvest seasons (2010/11–2025/26), seasonal structuring exceeded regional structuring for every trait (largest partial R 2 _season for acidity, 0.415); trait-specific seasonal signal ranked acidity (Adj. R 2 = 0.461) and ether extract (0.263) above moisture (0.163); climate–quality associations were selective (mold–rainfall retained within a pre-specified FDR focus set) while domain-level adjusted RDA remained weak; and no monotonic 16-season drift was detected, with seasons instead grouping into compositional year types. Empirical inference is restricted to the monitored network; transferability of the framework remains to be evaluated independently.