ContextCultivation and crop rotation, influenced by federal policy, prices, and precipitation, are significant sources of land-cover heterogeneity. Characterization of heterogeneity is required to identify areas and trends of stability or change.ObjectivesWe analyzed a land-cover time series within a prominent agroecosystem in the US, the Yazoo-Mississippi Delta (the Delta), as a case study of which metrics capture dynamics of landscape composition, configuration, connectivity, and context.MethodsAn assessment of land cover- from 2008 to 2021- was conducted and analyzed for potential differences among three Farm Bill eras. Twelve out of 23 metrics (including three new ones presented herein) examined were useful in characterizing land-cover heterogeneity.ResultsAlthough there was no increase in cultivated land, > 72% of the Delta experienced changes in land-cover type, and ~ 3% of the Delta was stable monoculture. Configurational metrics varied across years for soybeans, cotton, and rice, indicating prevalence of field-level changes in composition; connectivity metrics revealed isolation of upland forest and rice. The amount of corn was positively associated with the previous year’s commodity prices and negatively with precipitation whereas soybean acreage was lower in high-precipitation years and more dependent on commodity prices. Farm Bill effects were mixed among categories, whereas CRP generally declined.ConclusionsThe Delta experienced land-cover change with no net loss or gain of cultivated lands. Using 12 metrics that captured temporal shifts in spatial patterns, we characterized this agroecosystem as a shifting mosaic. Our approach may be useful for identifying areas of spatio-temporal heterogeneity or stability, with implications on resource management.