Introduction

Over the past two decades, ecological restoration has progressed to rely on more refined techniques, to include a greater array of ecosystems, and to attempt larger and more complex problems. Despite this progress, the outcome of many restorations fails to result in ecosystems that are similar to their natural counterparts. Restored ecosystems typically have fewer species and do not accumulate species over time, as expected. A lack of available seeds or suitable microsites for seedling establishment can hinder community development. Not surprisingly, seed availability is more often reported to be the key limitation to higher richness (e.g. Pywell et al., 2002; Martin & Wilsey, 2006; Kettenring, 2006). Most restorations introduce a small subset of the species expected and often at much lower abundances than exist in unaltered sites. To do otherwise seldom has been considered necessary because dispersal has the potential to add species over time. Unfortunately, habitat fragmentation has diminished native species propagule pressure and hinders dispersal in many landscapes (Galatowitsch & van der Valk, 1996; Honnay et al., 2002; Young et al., 2005) leading to increased recognition of the importance of adequate seed introductions for restorations.

When the investment in acquiring native seed for restoration is significant, there needs to be a reasonable likelihood that conditions are suitable for seedling emergence and growth. This can be especially challenging considering that site conditions at the start of a restoration project can be radically different than what might have ever existed in an unaltered community, even after natural disturbances.