Crop rotation and tillage system effects on weed seedbanks

We characterized the size and species composition of the weed seedbank after 35 yr of continuous crop rotation and tillage system treatments at two locations in Ohio. Spring seedbanks were monitored during 1997, 1998, and 1999 in continuous corn (CCC), corn–soybean (CS), and corn–oats–hay (COH) rotations in moldboard plow (MP), chisel plow (CP), and no-tillage (NT) plots where the same herbicide was used for a given crop each growing season. There were 47 species at Wooster and 45 species at Hoytville, with 37 species occurring at both locations in all 3 yr. Crop rotation was a more important determinant of seed density than was tillage system. Seed density was highest in NT and generally declined as tillage intensity increased. Seeds accumulated near the surface (0 to 5 cm) in NT but were uniformly distributed with depth in other tillage systems. At both locations there was a significant interaction between tillage and rotation for estimates of the total seed density. Seed density was highest in NT-CCC, with 26,850 seeds m−2 at Wooster and 8,680 seeds m−2 at Hoytville. At Wooster total seed density in CCC plots was 45 and 60% lower than in COH plots for CP and MP. In NT the total seed density was 40% greater in CCC than in COH. At Hoytville total seed density in CCC plots was 72% lower than in COH plots that were CP or MP, whereas seed density was 45% higher in CCC than in COH plots that were in an NT system. There were more significant differences in seedbank density for any given species for crop rotation than for tillage treatments. Seed densities of three broadleaves (shepherd's-purse, Pennsylvania smartweed, and corn speedwell) at Wooster and four broadleaves (yellow woodsorrel, redroot pigweed, Pennsylvania smartweed, and spotted spurge) at Hoytville were more abundant in COH (140 to 630 seeds m−2) than in CS (10 to 270 seeds m−2) or CCC (< 1 to 60 seeds m−2), regardless of the tillage system. At both locations Pennsylvania smartweed seeds were more abundant in COH (260 and 630 seeds m−2) than in other rotations (10 to 20 seeds m−2). Relative importance (RI) values, based on relative density and relative frequency of each species, were lower in CS than in CCC for common lambsquarters and five other weeds at Wooster; RI of giant foxtail was 80% lower in COH than in CCC at Hoytville. The data show how species composition and abundance change in response to crop and soil management. The results can help to determine how complex plant communities are “assembled” from a pool of species by specific constraints or filters.