RESUMO
The objective of this work was to develop a fast and practical method of weed seedbank evaluation to generate spatially distributed maps for use in site-specific weed management. Soil cores were collected at 0.20 m depth, air-dried, and then submitted to seedling growth in greenhouse. The sampling grid of 20 by 20 m was georeferenced by Global Positioning System, obtaining 73 soil cores with three replicates. During the greenhouse trial, there were two peaks of weed seedling growth: one in 119 days after water irrigation and another after KNO3 application. Weeds seedbank maps were obtained at different stages of seedling growth. The Pearson correlation was 0.99 for Brachiaria plantaginea seedbank map, 0.95 for Commelina benghalensis, and 0.85 for Cyperus rotudus generated at 119 days compared with 392 days after seedling growth in the greenhouse. The Brachiaria plantaginea seedbank map evaluated at 35 days presented correlation of 0.97 with 392 days. It was concluded that, for site-specific weed seedbank management, the evaluation of seedling growth in greenhouse until the first emergence peak is enough to generate weed seedbank maps.