The selective memory of weed seedbanks after 18 years of conservation tillage.
A conservation tillage study provided the opportunity to test whether tillage effects on the germinable weed seedbank would be consistent across different crop rotations and to investigate the potential residual effects of herbicide treatments terminated 12 yr earlier. Our objective was to measure the effects of tillage (moldboard plow [MP] vs. chisel plow [CP] vs. no-till [NT]), crop rotation (2-yr barley-red clover followed by 4-yr barley-canola-wheat-soybean rotation, compared to a cereal monoculture), and of a prior weed management factor (three intensity levels of herbicide use) on the density, diversity, and community structure of weed seedbanks. Species richness, evenness (Shannon's E), and diversity (Shannon's H′) of spring seedbanks varied little across treatments and over time. Total seedbank density generally increased as tillage was reduced, with some variations due to weed management in 1993 and crop rotation in 2006. Crop rotations generally had smaller seedbanks with fewer species than the monoculture. In 1993, seedbanks with minimum weed management were twice as dense as those with intensive or moderate weed management (approximately 6,000 vs. 3,000 seed m-2). By 2006, seed density averaged 6,838 seed m-2 across intensive and moderate weed management regardless of tillage, but was nearly twice as large in NT (12,188 seed m-2) compared to MP (4,770 seed m-2) and CP (7,117 seed m-2) with minimum weed management (LSD0.005 = 4488). Species with abundant seedbanks responded differently to treatments. Barnyardgrass and green foxtail had larger seedbanks in the monoculture than in the rotation. Common lambsquarters and pigweed species had large seedbanks in tilled treatments in the rotation, whereas yellow foxtail and field pennycress contributed to the large seedbanks observed in NT treatments. The latter two species were also associated with residual effects of weed management treatments (terminated 12 yr earlier) in NT. The differential seedbank response of weed species, attributed in part to contrasting weed emergence patterns and agronomic practice effects on seed rain, explained some of the weak treatment effects observed for total seedbank density and diversity. The large weed seedbanks observed in NT plots after 18 yr confirms the importance of seed rain and seedbank management for the sustainability of NT systems.