National Park Service, Washington Office, Biological Resource Management Division, Fort Collins, Colorado 80525, USA.
Many plant species rely on viable seeds stored in the soil for recruiting new individuals. Knowledge of soil seed banks can enhance our understanding of plant traits and vegetation change while informing biodiversity conservation and ecological restoration. Soil seed banks remain among the most difficult attributes to measure of plants and plant communities. This presentation discusses a series of studies conducted in the Mojave Desert and in ponderosa pine forest in southwestern North America with the goal of using seed biology to improve characterization of soil seed banks. Smoke, charred wood, and other germination cues were evaluated for their potential for increasing emergence of seedlings in the germination method of characterizing soil seed banks. Germination assays were also used in an attempt to refine identification of viable seeds in the extraction method of measuring soil seed banks, which involves directly separating seeds from soil. Major conclusions of this work were that: (i) some fire-related cues, such as smoke, enhanced seed detection in some North American ponderosa pine forests but at levels much lower than reported for many Australian fire-dependent ecosystems; (ii) in comparing the germination and extraction procedures, extraction detected orders of magnitude more seeds than the germination procedure; and (iii) viability of the extracted desert seeds remained uncertain, because viability evaluations and germination protocols for the desert species were not well developed. The significance of ensuring viability of extracted seeds warrants further attention when using the extraction method for measuring soil seed banks. Improvements in soil seed bank assays will likely be linked with enhanced knowledge of seed biology.