Oak savanna restoration: to seed or not to seed?

Oak savanna restoration is a long-term and expensive activity, but the results are worth it. Because it takes so long for a restored savanna to be established, careful long-range planning is essential.

Generally, one starts with a fairly degraded (former) savanna, containing numerous open-grown oaks, but heavily crowded with invasive trees such as elm, walnut, maple, box elder, or other fire-sensitive species. Also present are probably invasive shrubs such as buckthorn, honeysuckle, prickly ash, autumn olive, etc.

Let's take a 5-10 acre parcel. Here are the restoration steps:

• Remove all the invasive trees, paying special attention to those that are crowding open-grown oaks. This step is sometimes called "daylighting the oaks."
• At the same time, cut and treat all invasive shrubs, both native and non-native.
• Burn the understory either in the spring or fall.

Some restorationists stop at this point, perhaps doing annual or biennial burns and waiting for the understory herbaceous vegetation to appear. Whether they realize it or not, those folks are assuming that there is a good "seed bank" still present and as soon as light is allowed in, those species will be "released". 

With highly degraded oak savannas, it seems unlikely that a complete complement of savanna species seeds will still be present in the soil. Some species may be present, but most will not be. Data from studies on prairies has shown that seeds of many prairie forbs and warm-season grasses, especially, do not persist very long. The main species that persist are hard-seeded types such as sedges and legumes. Since many of the savannas being restored have been degraded for over 50 years, it seems reasonable that the seed bank is greatly impoverished.

However, even if a seed bank may exist, why not plant? Experience has shown that it takes 5-10 years to get a good savanna understory established. Let's speed things up. Of course, the species planted should be those that already exist in the area (local genotypes). They will join with those arising from the seed bank, thus increasing the diversity. The only situation where seeding would not be warranted was if the site was part of a research study.

Excellent lists of savanna species are available from Brian Pruka, Steve Packard, and Brian Bader, and I have included them in my web site on oak savanna restoration.