Weed seedbank diversity and sustainability indicators for simple and more diverse cropping systems

Storkey and Neve (2018) hypothesised that weed seedbank diversity could be used as an indicator of agroecosystem sustainability, with cropping systems associated with higher weed seedbank diversity being more agronomically and environmentally sustainable than those with lower weed diversity. We evaluated their hypothesis using multiple years of empirical data collected from a long-term field experiment in Iowa, USA, and life cycle assessment models parameterised with data from that experiment. We found that diversifying a 2-year maize–soyabean rotation with two additional crop phases, oat and lucerne, to form a 4-year rotation (a) increased Simpson's diversity index for weed species in the soil seedbank, while (b) increasing maize and soyabean productivity, (c) maintaining total crop energy production and profitability, and (d) reducing herbicide-related aquatic toxicity, fossil energy consumption, greenhouse gas emissions, damage to human health due to fine particulate matter, and discharge of soil sediment, nitrogen, and phosphorus. Increased weed biomass production and greater weed species diversity that occurred within oat and lucerne phases of the 4-year rotation sequence did not result in a loss of weed control within maize and soyabean phases of the rotation. We suggest that while there was no cause-and-effect relationship between greater weed seedbank diversity and improved agroecosystem sustainability in agronomic, economic and environmental dimensions, all of these responses were driven in a parallel manner by cropping system diversification. Consequently, weed seedbank diversity may indeed be a useful agroecosystem sustainability indicator. Further evaluation of this hypothesis in more systems is warranted.