Seedbank response to integrated crop management systems

Non-Peer ReviewedSix integrated crop management systems varying primarily in terms of tillage and herbicide inputs were maintained for four years at two locations in the Moist Mixed Grassland ecoregion in Saskatchewan. The seedbank was sampled in each plot in the spring of 1997 prior to the initiation of the experiment and the spring of 2001 following the completion of the experiment to determine if cropping system had changed seedbank composition. Redundancy Analysis (RDA) was used to directly quantify the change in weed seedbank attributable to cropping system at each location. The RDA was constrained by cropping system for the 2001 data and year for the 1997 data. Initial spatial variation observed in the 1997 weed seedbank was removed from the analysis by using plot number and replicate block as binary covariables. The majority of species increased in the seedbank in all systems studied at both locations. The seedbank from the no herbicide high tillage systems was distinguishable from the other systems by large increases in lamb’s-quarters and stinkweed at both locations. While similar species tended to increase in all the systems with herbicides, the systems with lower levels of herbicides applied had larger increases in the seedbank than those with high herbicide inputs