Secrets of the seed bank:
Tiny clues to a landscape's past and future

---

by Thomas Rosburg

In a very basic sense, there seem to be two sides to the world of nature. One is conspicuous, visible to us as a hawk waltzing with the wind high above, a noisy katydid making a chlorophyll sandwich out of big bluestem leaves, or the deep jug-o-rum from the baritone voice of bullfrogs in an oxbow wetland. As long as we open our senses, we find this side of nature all around us.

The other side of nature is all around us, too, but in a world that is much less conspicuous. It's the leaf miner tunneling through the heart of a sunflower leaf, or the tangle of fungal threads penetrating and feeding on a decaying log, or the community of dormant plant seeds tucked away in the soil beneath our feet.

Because of its lack of visibility, this side of nature is more difficult to appreciate and understand, but it is not any less important. In fact, those dormant seeds in the soil - collectively known as the seed bank - often play a crucial role in the patterns of vegetation that we see and admire growing aboveground. They also provide important clues to a landscape's history and future.

Why should we care?
In the strictest sense, a description of a plant community should include the seed bank, since seeds are living organisms and are an integral, albeit dormant, part of the vegetation.

Seed banks have both historic and futuristic connections with the above-ground plant community. A seed bank provides a memory of the vegetation on a site. Species that are no longer present as plants in the vegetation may be present as seeds in the seed bank. It is possible to look back in time and see a picture of the previous vegetation. Seed banks also contribute to the future plant life in a habitat, especially after a disturbance event that enhances germination and regeneration of seeds by opening up a space for plants.

And, in practical terms, knowledge of seed banks in general-and the seed bank on a given site-can give landowners clues to understanding a landscape's history and help in guiding it toward a healthy future.

What's a seed bank?
A seed bank is the community of viable seeds present in the soil.
Seeds are mysterious organisms even when plainly visible in the palm of our hand. It's hard to think of them for what they really are - a living baby plant - because they don't look anything like miniatures of the adult plant that produced them. That would be what we call a seedling.

However, the real offspring of a plant is the seed, a tiny breathing embryo containing the blueprint for making the adult plant. Of course, the embryo needs some protection so the mother plant builds a seed coat from a few layers of cells and adds a store of food, the endosperm, for the baby plant to draw upon as it takes its first big step in life - germination and growth into a seedling.

Much of the mystique of a seed comes from their ability to become dormant for long periods of time. Of course, seeds are not the only organisms able to perform that feat. Insects, amphibians and reptiles magically put themselves into suspended animation, usually as an adaptation to survive a harsh, stressful period such as winter. Spadefoot toads, which are found along the western border of Iowa, are champions of prolonged dormancy, spending nearly 11 months of the year buried two to three feet in the soil.

For a seed, one year of dormancy is not a much of a big deal. Many species are capable of persisting for a year in dormancy, and many more can maintain viability for many years. Seed dormancy or longevity is what makes possible the existence of a seed bank. The seeds are alive and patient, waiting for the right time to germinate.

The term seed rain refers to the process by which seeds enter the seed bank. Seeds, either born and produced on site or carried to the site by a dispersal agent, become incorporated into the soil. If seeds can remain viable for many years, and new seeds continue to "rain" down, it's easy to see where the term seed bank comes from, as the seeds accumulate over time, and form a reserve of seeds in the soil.

These patterns arise because species that successfully form persistent seed banks are the species with greatest seed longevity. It is one of two contrasting strategies plants may employ to realize success in replacing themselves and insuring maintenance of the species. A plant dispersing its seed is most successful, evolutionarily speaking, if it can place its seed in an environment suitable for germination and growth. Such safe sites, as they are called, are generally rare in natural landscapes. In order to find them, seeds must be dispersed widely in space (to increase probability of landing in a rare safe site) or widely in time (to increase the probability the seed will survive long enough for a safe site to materialize). Plants with high seed longevity have evolved the strategy of waiting patiently for the right time.

How long can seeds last?
Seed longevity is not an easy characteristic to measure. Essentially the viability (often demonstrated by germination) must be assessed for seeds of a known age. Tucking some seeds away in an envelope inside a drawer or closet and testing them several years later could be one approach.

But such a controlled setting is not nearly the same environment as seeds experience in the soil, where moisture, acids, temperature, gases, predators, and pathogens create potential threats to the viability of seeds. Longevity in an ecological context should be tested by putting seeds into a natural environment.

Some of the best information on seed longevity comes from seeds recovered in archeological settings or from experiments on buried seeds. Seeds of common weed species in Mexico have been extracted and germinated from adobe bricks aged between 150 and 200 years old. Even more impressive, seeds of Canna found inside a walnut shell necklace that was dated at 600 years old proved to be viable.

The famous Beal experiment also provides good data on seed longevity. It was started in 1879 when 20 open bottles containing 50 seeds each of several species were buried about 20 inches below the surface. Every 5 or 10 years, a bottle from each species was dug up and the contents germinated on sterilized soil. After 50 years 25% of the species still produced some seedlings, and after 75 years 20% of the species were still viable, including mullein, sour dock and common evening primrose.

Seeds from dated herbarium sheets have also been useful, but these seeds have not been in a natural environment so these estimates should be considered a potential longevity rather than an ecological longevity. One example is a seed of lotus that was grown from a herbarium sheet 237 years old.

These estimates of longevity highlight the maximum viabilities that are known. For most species this kind of information is lacking, and for those studied most have viability less than 10 years. For example, a couple common prairie grasses - Canada wildrye and switchgrass - appear to have longevity in the range of 1 and 3 years respectively.

How are seed banks studied?
The methods of studying and identifying seed bank composition is straightforward - obtain a sample and either find and identify the seed itself, or germinate the seed and identify the seedling.

The first, called a seed assay, is less commonly done than the second, called a seedling assay. The main reason for this is that finding and identifying the tiny seeds in a sample of soil is very tedious work. Once found and identified, their viability needs to be determined - in other words is the seed alive? It can be done, but it seems much easier to let the seeds tell you if they are alive by virtue of germination, and then identify the seedling or young plant. The only drawback of the seedling assay is that only the seeds that germinate will be recognized and counted. If seeds stay dormant, maybe because the correct germination environment is missing, those species will not be identified as part of the seed bank.

Seed banks have been studied for well over a century. In fact one of the earliest published descriptions of a seed bank study was by Charles Darwin, who counted the seedlings that geminated from a sample of mud from a pond. Many types of habitats have been studied since then, including grassland pastures, agricultural lands, forests, prairies, wetlands and tundra. Collectively this research has defined several basic principles of seed bank ecology.

Clues about history
My own research on the seed banks of prairie and edge communities in the Loess Hills illustrates what seed banks can reveal about a landscape's history. The most common species in the seed bank of the prairie were species that were either absent or very sparse in the vegetation - like hoary vervain, mullein, yellow wood sorrel, horseweed, and spurge.

Because these species all grow well in disturbed settings, they are more commonly associated with grazed pastures. The study sites, which were located on public land at the time of the study in 1990, were formerly privately owned and grazed 15 to 20 years earlier. The seed bank was revealing a picture of the former community.

Since the species in the seed bank germinate more successfully when there has been a disturbance that opens up space in the plant community, the similarity between seed bank and vegetation increases as disturbance increases. In the Loess Hills study, this pattern was observed in a couple of ways. Seed bank samples were collected at two sites with a different history of management. The similarity between seed bank and vegetation was greater at the site with the more recent history of grazing, and was more dissimilar at the site that had been in public ownership longer. As recency of grazing (heavy grazing actually) increased, disturbance increased and so did similarity between seed bank and vegetation.
Even within one of the sites, a similar pattern occurred among the different community types. The greatest similarity between seed bank and vegetation was in the ecotonal community, which is the community that forms at the interface between prairie and woodland. Arguably this is a community of high disturbance since environmental stability is low as woody vegetation gradually encroaches into the prairie.

Similar patterns have been demonstrated for forests. In a study of seed banks in European forests varying in age from young (55 to 116 years old and established on formerly arable land) to old-growth forest (greater than 250 years), similarity between seed bank and vegetation decreased in the older forests. Species in the seed banks were mainly those typically found along forest edges, in earlier successional stages, or in small disturbances within the forest. This makes it clear that minimization of disturbances is imperative for successful management of old-growth forests.

Clues about exotic invaders
Many natural ecosystems are threatened by exotic invaders from other ecosystems or continents. Some invading species, with no natural predators, crowd out local species and degrade the site. Often the best measure of protection against the habitat degradation caused by exotic species is preventative action. It is much easier to eradicate a small establishing population than one that has been established for several years.

Seed bank studies can help land managers recognize exotic or problematic species that may be entering the community as seeds, not yet visible as plants, but nonetheless accumulating on the site and building the potential to become established plants. When an exotic species not yet established in the vegetation shows up in the seed bank, it warns land managers to carefully monitor the site and remove or treat establishing plants before they take over.

An example of this early warning occurred unexpectedly in a seed bank study of the Everglades that I worked on in 1995. The goal was to assess the seed banks of the wetland communities along a transect extending from the northern boundary several miles to the south. Wetland scientists had identified nutrient pollution in the form of excess phosphorus in the Everglades along the northern boundary. They had shown the phosphorus is coming from runoff from the agricultural landscape, the huge fields of sugarcane, that lie north of the Everglades.

The seed bank study revealed that not only are nutrients flowing into the Everglades from these fields, but also several exotic plant species. These species were not causing problems yet because they had not become established as plants in the community. But the study revealed the potential invader problem and gave managers even more reason to control runoff-to avoid both farm chemicals and exotic invaders.

Clues for restoration
Another example of how ecologists can use seed bank information is in connection with ecological restoration. It is logical to think that prairies or wetlands might be reborn through the germination of seeds saved away in the seed bank.

Recent seed bank studies in Iowa and the prairie pothole region provide some insight in this matter. In the late 1980's, ISU graduate student Ann Akey and I studied the potential for prairie restoration on southern Iowa pastureland. We found many pastures that, after removal of grazing and suppression of non-native grasses by fire or atrazine, became transformed almost overnight into a recognizable prairie. When we measured the seed banks of these pastures, we found very little evidence that any of the prairie species re-establishing in the community were coming from seeds. By far weedy species, like crabgrass, foxtail, and spurge dominated the seed banks of the pastures. Whatever rebirth of prairie occurred came from resilient plants that had survived the years and years of pasture use. The key message - don't expect seed banks to provide a source of seed for prairie restorations of pastureland.

Wetlands, however, are different. In a study by wetland scientists at ISU, the seed banks of wetland basins that had been drained between 5 and 70 years were sampled. The goal was to determine if seed banks of these wetland basins had any value for wetland restoration.

They found two important patterns. One was that the number of wetland species and the number of total seeds decreased as the length of time drained increased. Second, the wetlands drained for less than 20 years had a seed bank that contained viable seeds of many wetland species and are the best candidates for restoration - seed banks in these drained basins could provide a significant role in restoration.

The differences seen in the seed banks of these two studies reveal some distinctions in how seed banks function in prairies and wetlands.

In a prairie, the seed bank plays a minor role in the maintenance of the prairie community. The vast majority of prairie plants are perennial and rely on the underground growth of roots and rhizomes to maintain their presence. The presence of short-lived species and biennial or annual plants in the seed bank does provide a mechanism of adding plant diversity to the prairie after fires occur, or after soil disturbances from digging badgers or wallowing bison create open patches.

In wetlands, germination and growth of new seedlings is easier if there is little or no water present to obstruct light availability. Therefore annual or periodic drought cycles create open areas ripe for colonization by new plant species. Many wetland species have adapted to the fluctuating water depths common in the prairie pothole region by increasing the longevity of their seeds, so that when the wetland dries up, they have seeds ready to germinate.

Thus the seed bank plays a more important role in determining the year-to-year species composition and vegetation change in wetlands than in prairies.

Conclusion
It is true that seed banks are neglected in land conservation and management, but only because they are difficult to see. If an effort to measure them can be made, then the door is opened to a greater understanding of how certain management actions will affect the quality and condition of the natural area. And as you hike through and enjoy these natural landscapes, you now can have a greater appreciation for the thousands - perhaps tens of thousands - of seeds lying beneath your feet.

Thomas Rosburg is an assistant Biology professor at Drake University in Des Moines, Iowa. He is also on the Iowa Natural Heritage Foundation's advisory board. Rosburg's current research specialty is plant ecology.

Selected Bibliography
Bossuyt, B., M. Heyn, and M. Hermy. In press. Seed bank and vegetation composition across ancient-recent forest ecotones in central Belgium. Plant Ecology.

Cheplick, G. P., ed. 1998. Population biology of grasses. Cambridge University Press, Cambridge.

Fenner, M. 1985. Soil seed banks. Pages 56-71. Seed Ecology. Chapman & Hall, New York City.

Rosburg, T. R., T. W. Jurik, and D. C. Glenn-Lewin. 1992. Seed banks of communities in Iowa's Loess Hills: ecology and potential contribution to restoration of native grassland. Pages 221-237 in R. G. Wickett, P. D.

Lewis, A. Woodliffe, and P. Pratt, eds. Proceedings of the Thirteenth North American Prairie Conference: Spirit of the Land, Our Prairie Legacy. Department of Parks and Recreation, Windsor, ON, Canada, 272 pages.

van der Valk, A. G., and C. B. Davis. 1978. The role of seed banks in the vegetation dynamics of prairie glacial marshes. Ecology 59: 322-335.

van der Valk, A., and T. R. Rosburg. 1997. Seed bank composition along a phosphorus gradient in the northern Florida Everglades. Wetlands 17(2):228-236.

Wienhold, C. E., and A. van der Valk. 1989. The impact of duration of drainage on the seed banks of northern prairie wetlands. Canadian Journal of Botany 67: 1878-1884.

For more information, e-mail Cathy Engstrom, director of communications, or call (515) 288-1846.