Comparison of Seed Bank Estimation Techniques Using Six Weed Species in Two Soil Types

Abstract

Evaluation of the viable seeds in a soil, otherwise known as the seed pool or seed bank, is a crucial component of many weed dynamic and plant ecology studies. Seed bank estimation is used to predict the possibility of future weed infestations in rangelands as well as the nascent native plant diversity within them. However, there is no standardized method of reporting seed bank evaluation techniques, limiting the ability to compare across studies. After sowing known quantities of cheatgrass, Bromus tectorum (L.); brome fescue, Vulpia bromoides (L., S.F. Gray); pigweed, Amaranthus retroflexus (L.); kochia, Kochia scoparia (L. Schrad.); lambsquarters, Chenopodium album (L.); and field pepperweed, Lepidium campestre (L. R. Br.) into sterile soil, we compared two different watering regimes in two soil types to Petri plate germination of these seeds. Seed bank estimations from the emergence method were lower compared to estimations from the Petri plate germination. Top-and-bottom watering increased absolute abundance, and the rank order of abundance among species changed with watering method. Emergence levels were the same between the two soil types. The higher water availability of the top-and-bottom watering method resulted in greater seedling emergence (26.3% 6 10% SD vs. 9.1% 6 7.5% SD). Lower emergence compared to germination (62.3%624.4%) may indicate that emergence is an important postgermination barrier to seedling establishment. While emergence techniques may not accurately portray the volume of seeds in the soil, they may more accurately predict which plants can become established in field conditions. Our different species abundances between watering methods show that multiple emergence methods may need to be employed to forecast a range of future rangeland conditions from the soil seed bank.