An Introduction to Wetland Seed Banks
By Douglas A. DeBerry and James E. Perry
One of the most important structural compo- ture on wetland seed banks in natural and cre- nents of wetland ecosystems is the seed bank. ated or restored systems, and defines the role Seed banks are present in nearly all ecosystems, seed banks play in created and restored wetland and can be defined as “[an] aggregation of management. ungerminated seed potentially capable of replac- ing adult plants that may be annuals, dying a natural or unnatural death, or perennials, sus- Seeds and Seed Ecology ceptible to death by disease, disturbance, or con- Regeneration of wetland plant communities oc- sumption by animals including man” (Baker curs by sexual reproduction through either the 1989). They are a critical component in the es- development of seeds, or by asexual reproduc- tablishment and development of vegetation tion through clonal propagation by rhizomes or communities in wetlands (van der Valk 1981). other vegetative organs. A plant may employ the Practical principles concerning the “behavior” of former (e.g. annuals), the latter (e.g. submerged seeds in the soil in different wetland types may aquatics), or both (e.g. herbaceous perennials) be derived from past as a principal reproduc- research. The pur- tive strategy (Fenner pose of this and tech- 1985). Much of the lit- nical report number erature on seed bank re- 00-4 is to present to search deals with the reader a general sexually produced overview of seed bank propagules in an- ecology and the role giosperms (flowering seed banks play in plants) (Leck 1989). wetland succession processes. In this re- Seeds port we introduce the concepts of seed Seeds of angiosperms physiology and ecol- develop from a fertilized ogy as applied to seed ovule and consist of the Figure 1. Diagram of a dicot and monocot seeds. The storage in the soil. following: 1) the embryo embryos in these seeds consist of epicotyl, hypocotyl, We conclude with a (includes the epicotyl, radicle, and cotyledon. Left: A dicot seed (bean) in which short introduction to hypocotyl, radicle, and one cotyledon has been removed to reveal remaining parts our current under- cotyledon) which even- of embryo. Right: A monocot seed (corn) has only one standing of wetland tually grows into the cotyledon, but a large endosperm. Modified from Keeton seed banks. Report young plant; 2) the en- and Gould (1993). Endosperm of mature dicot is indistin- number 00-4 dis- dosperm (internal food guishable from cotyledon. cusses recent litera- source for developing 1 seed); and 3) the seed coat (Raven et al. 1992, ploit a particular dispersal vector, such as high figure 1). Although all mature seeds contain an surface to volume ratio in small seeds (e.g. many embryo, even if only poorly developed, and most grasses and sedges), or the presence of append- are surrounded by a seed coat, the extent to ages such as plumes, hairs, or wings to increase which the endosperm and perisperm persist air resistance for wind dispersal (e.g. sea side varies among species. In some cases, the seed goldenrod). Adaptations that facilitate dispersal coat is in rudimentary form and the seed is con- by animals include burrs, hooks, or adhesive tained within the pericarp (fruit coat) derived substances, which increase the opportunities for from the ovary wall. In this condition, the pri- a seed to cling to an animal (ectozoochory, e.g. mary dispersal unit is not a seed, but a fruit marsh beggar ticks). Certain seeds develop spe- (Fenner 1985, Bewley and Black 1994). In ma- cial seed coats that increase buoyancy for water ture dicots (e.g. beans) the endosperm is stored transport (e.g. arrow arum). Special oil bodies in the cotyledon and is indistinguishable. (elaiosomes) are also present on some seeds as an attractant for insects, such as ants, which are exploited in dispersal (myrmecochory). Many Dispersal mechanism plant species have developed large, nutritious fruits (e.g. persimmons) which attract frugivores Growth of the seed embryo is usually delayed (fruit-eating animals). The seeds within the fruit while the seed matures and is dispersed (Baskin are not susceptible to the digestive acids of an and Baskin 1989). Several dispersal mecha- animal’s stomach and are, therefore, viable when nisms are available to seed plants, including eliminated from the animal that ingests the fruit wind (anemonochory), water (hydrochory), and (endozoochory). In some species, the foliage of animals (zoochory - birds in particular) (Fenner the plant may serve as an attractant, alleviating 1985). Wind dispersed seeds of various species the necessity for fruit development (e.g. flower- have morphological adaptations designed to ex- ing and silky dogwoods) (Willson 1992, Raven et al. 1992, Fenner 1985).
Wetlands Program Germination requirements School of Marine Science Virginia Institute of Marine Science Seed germination depends on both internal and College of William and Mary external factors (Bewley and Black 1994). The Gloucester Point, Virginia 23062 three most important external (environmental) (804) 684-7380 factors include the proper amounts of water and oxygen and appropriate temperature range. Too Dr. Carl Hershner, Program Director little or too much water, too little oxygen, and Dr. Kirk Havens, Editor temperatures that stay too low may lead to de- Published by: VIMS Publication Center lay or lack of germination. Some seeds from the temperate climate zone may require freezing tem- This technical report was funded, in part, by the peratures for several consecutive days to sclarify Virginia Coastal Resources Management Program (see discussion on stratification below). In ad- of the Dept. of Environmental Quality dition, some seeds have a light requirement for through Grant #NA97O20181-01 of the germination (Leck 1989). National Oceanic and Atmospheric Ad- ministration, Office of Ocean and Coastal Most mature seeds have low moisture content, Resources Management, under the containing only about 5-20% water by weight Coastal Zone Management Act of 1972, as (Fenner 1985), and must imbibe (absorb) wa- amended. ter in order to activate enzymes and initiate me- tabolism. During the early stages of germination, The views expressed herein are those of the authors respiration may be entirely anaerobic (i.e. car- and do not necessarily reflect the views of NOAA or ried out in the lack of oxygen), but as soon as any of its subagencies or DEQ. the seed coat is ruptured, the seed switches to aerobic respiration (Bewley and Black 1994). Commonwealth’s Declared Policy: “to preserve the wet- Too much water may also be a problem since if lands and to prevent their despoliation and destruction...” the seed becomes waterlogged, the amount of oxygen available will be limited and can be inad- Printed on recycled paper. equate for seedling establishment (Karssen and
2 Hilhorst 1992). In wetlands, seeds must there- often referred to as early colonizers or volunteer fore maintain adaptations to overcome the low species (Pickett and McDonnell 1989). There- oxygen levels associated with saturated soil con- fore, certain annual species may be perennially ditions. represented in the seed bank, and will become established when disturbance or some other fac- tor removes the competitive influence of existing Dormancy as an adaptation for vegetation (Fenner 1985). seedling survival Even in favorable conditions some viable seeds Seed Bank Dynamics will fail to germinate and are said to be dormant. Common causes of dormancy include physiologi- The mobility of seeds in the soil is poorly under- cal immaturity of the embryo and impermeabil- stood, but appears to follow some disturbance ity of the seed coat to water and oxygen (Baskin mechanism which may be facilitated by the ac- and Baskin 1989). Seeds with a physiological tivity of burrowing animals or shrink-swell fis- dormancy will sometimes require a complex se- suring of the soil during alternating wet and dry ries of enzymatic and biochemical changes, re- periods. Some seed migration may occur by ferred to as “after-ripening,” in order to ground water percolation, with smaller seeds germinate. In temperate climates, after-ripen- moving more rapidly than larger ones (Leck ing is triggered by the low temperatures of win- 1989). By far, human disturbance through till- ter (stratification). The after-ripening ing and other practices provides the most im- requirement, therefore, is seen as an adaptation mediate mechanism for vertical transport of that delays germination through the stressful seeds within the soil (Simpson et al 1989). For winter climate until warmer temperatures when many soil seed banks, a physical disturbance conditions are more conducive to seedling sur- mechanism must be present in order to bring vival (Raven et al. 1992). Seed dormancy may buried seeds to the surface where germination also be induced by burial in the soil, underde- may occur (Fenner 1985). velopment of the embryo, and chemical inhibi- tion, among other factors (Fenner 1985). These Methods for Studying Seed Banks dormancy mechanisms represent adaptations that adjust germination timing for the seed to The most common method of seed bank analy- coincide with favorable environmental conditions. sis is the seedling emergence technique, where soil samples are extracted and brought back to Seed Bank a greenhouse or lab to be observed under condi- tions favorable for germination. Under this The seed bank includes all viable seeds present method, germinating species are removed from on, or in, the soil or associated litter. Contribu- the soil as they are identified, and the treatment tions are derived from the seed rain, which in- continues until all seedlings have emerged from cludes dispersal from the local vegetation, but the soil. In a review of several methods for esti- may also include major contributions from spa- mating seed numbers in the soil, Gross (1990) tially and temporally distant sources (Simpson found that the seedling emergence technique un- et al. 1989). Seed losses from the soil are gen- derestimates the seed bank because, under any erally due to germination, fire, predation, decom- experimental condition, germination require- position, physiological death induced by deep ments for some species will not be met. How- burial, and physical removal due to erosion or ever, Gross also noted that this method does other disturbances (Leck 1989, Fenner 1985). provide the most complete listing of species present in the seed bank. Other methods em- In general, seeds of species forming seed banks ployed by researchers include direct counts us- must be viable for long periods of time. This ing sieves to separate seeds (Bonis et al. 1995), requires extended periods of dormancy thus en- and elutriation, a system that recognizes fine root suring viability and persistence in the seed bank production from seeds in the soil (Gross 1990). until conditions are favorable for germination As Gross points out, either of these alternative (Murdoch and Ellis 1992). Species that tend to methods gives better results of actual seed num- form seed banks are those typically found in bers in the soil; however, the methods are se- habitats appearing unpredictably in nature (e.g. verely limited by the necessity of identifying bare soil). Such species are typically annuals, species by seed alone.
3 Wetland Seed Banks • Species composition in the seed bank is most often correlated with the standing vegetation Much of our knowledge of wetland seed bank in wetland systems dominated by herbaceous ecology has been developed through research car- vegetation. Forested wetlands tend to support ried out in tidal salt and freshwater marshes seed banks that do not reflect the standing and/or seasonal and temporary non-tidal wet- vegetation. lands (Leck 1989). Nearly all of these studies indicate that seed banks play an important role • Wind and water are two of the most important in the composition and succession of vegetation dispersal vectors operating in wetland communities (Thompson 1992). Several gener- systems. In some wetlands, waterfowl alizations may be drawn from the literature (Leck dispersal is also an important vector, allowing 1989): for a much broader potential geographic distribution for wetland plant species that • Wetland seed bank composition and size are utilize this mechanism. highly variable, with individual samples ranging from 0 to 59 species and 0 to 377,041 The importance of drawdown conditions as a seeds m2. Fewer seeds are found at continually recruitment alternative for species present in the inundated sites. seed bank is the most recurrent subject of re- view in the literature. van der Valk (1981) de- • Increasing water depth and salinity decreases scribes this phenomenon as an “environmental size and diversity of wetland seed banks. sieve” where the influence of water level fluctua- tion determines which species can become es- • Dominant species represented in wetland seed tablished in the wetland at any given time. banks are most often graminoids (grasses or Certain annuals may only become established grass-like plants); and seed banks are most in the wetland when sediments are exposed and often dominated by one species (15-90%). will die back upon inundation. However, if these species can set seed and establish a presence in • Woody species are often poorly represented or the seed bank, a flood event will not extirpate lacking in wetland seed banks (including the species from the wetland. A species with an forested wetlands). annual life history may, therefore, have a peren- nial component in the soil seed bank (van der • Vertical distribution of seeds is variable among Valk 1981, van der Valk and Davis 1978). wetland habitats with 80% of all viable seeds occurring in the top 4-5cm of soil for lacustrine, The unpredictability of drawdown frequency in temporary, and freshwater tidal wetlands. certain wetlands, induced by stochastic events Only 20-50% of viable seeds occur in the top such as drought, presents a problem to scien- 5cm of the soil in swamps, prairie marshes, tists who wish to make predictions on succes- and bogs (wetland systems for which deep sional sequences in wetlands. This effect is most burial is more characteristic). notable in seasonal wetlands, for which precipi- tation is the principal hydrologic component • Low oxygen levels in consistently saturated or (Poiani and Johnson 1989). flooded wetland soils may reduce deterioration and increase longevity in seeds. Another factor that may affect diversity in wet- land ecosystems is the potential for development • Inundation is the most important factor of aggressive weed species from the seed bank influencing the recruitment of vegetation from (Leck 1989). Invasive or exotic species present the seed bank in wetlands. Germination in the seed bank may proliferate during favor- response to the drawdown frequency able drawdown conditions. Once established, (periodicity of alternating flooded and dry these species may reproduce by seed and rhi- conditions) favors submersed aquatic species zome, and may competitively exclude annuals that during flooded conditions, and mudflat persist only in the seed bank (van der Valk and annuals or emergent perennials during Penderson 1989). The presence of invasive spe- drawdown conditions. cies in the seed bank is thus a threat to the main- tenance of diversity in wetland systems.
The seed bank represents an important compo- vides a storage medium for plants that might nent of the overall wetland ecosystem. Stresses otherwise not survive through stressful condi- at the seed bank level imposed by flooding, low tions. In general, the function of the wetland oxygen levels, and other site-specific environmen- seed bank is to maintain genetic continuity and tal conditions in wetlands (e.g. salinity, burial, diversity at the species level, and to maintain etc.), select for species with physical and/or species diversity at the community level. There- physiological adaptations to overcome environ- fore, a comprehensive understanding of plant mental stresses for successful seed germination dynamics in wetlands should not stop at the soil and seedling establishment. The seed bank pro- surface.
Baker, H.G. 1989. The natural history of seed banks. Murdoch, A.J. and R.H. Ellis. 1992. Longevity, vi- In: Leck, M.A., V.T. Parker, and R.L. Simpson ability and dormancy. In: Fenner, M. (ed.). 1992. (eds.). 1989. Ecology of Soil Seed Banks. Aca- Seeds: The Ecology of Regeneration in Plant demic Press, Inc., San Diego, CA. pp. 9-21. Communities. CAB International, Wallingford, Oxon, UK. pp. 193-230. Baskin, J.M. and C.C. Baskin. 1989. Physiology of dormancy and germination in relation to seed Pickett, S.T.A. and M.J. McDonnell. 1989. Seed bank bank ecology. In: Leck, M.A., V.T. Parker, and R.L. dynamics in temperate deciduous forest. In: Leck, Simpson (eds.). 1989. Ecology of Soil Seed M.A., V.T. Parker, and R.L. Simpson (eds.). 1989. Banks. Academic Press, Inc., San Diego, CA. pp. Ecology of Soil Seed Banks. Academic Press, 53-66. Inc., San Diego, CA. pp. 123-147. Bewley, J. D. and M. Black. 1994. Seeds: Physiol- Poiani, K. A. and W. C. Johnson. 1989. Effect of ogy of Development and Germination. Plenum hydroperiod on seed bank composition in semi- Press, New York, NY. pp. 5-10. permanent prairie wetlands. Canadian Journal of Botany 67:856-864. Bonis, A., J. Lepart, and P. Grillas. 1995. Seed bank dynamics and coexistence of annual macrophytes Raven, P. H., R. F. Evert, and S. E. Eichhorn. 1992. in a temporary and variable habitat. Oikos Biology of Plants. Worth Publishers, New York, 74:81-92. NY. pp. 441-452. Fenner, M. 1985. Seed Ecology. Chapman and Hall, Simpson, R.L., M.A. Leck, and V.T. Parker. 1989. London, U.K. pp. 38-53. Seed banks: general concepts and methodologi- cal issues. In: Leck, M.A., V.T. Parker, and R.L. Gross, K. L. 1990. A comparison of methods for Simpson (eds.). 1989. Ecology of Soil Seed estimating seed numbers in the soil. Journal of Banks. Academic Press, Inc., San Diego, CA. pp. Ecology 78:1079-1093. 3-9. Karssen, C.M. and H.W.M. Hilhorst. 1992. Effect of Thompson, K. 1992. The functional ecology of seed chemical environment on seed germination. In: banks. In: Fenner, M. (ed.). 1992. Seeds: The Fenner, M. (ed.). 1992. Seeds: The Ecology of Ecology of Regeneration in Plant Communities. Regeneration in Plant Communities. CAB In- CAB International, Wallingford, Oxon, UK. pp. ternational, Wallingford, Oxon, UK. pp. 327-348. 231-258. Keeton, W.T. and J.L. Gould. 1993. Biological Sci- van der Valk, A. G. 1981. Succession in wetlands: a ence: Fifth Edition. W. W. Norton & Co. Ney York, Gleasonian approach. Ecology 62:688-696. NY. van der Valk, A. G. and C. B. Davis. 1978. The role Leck, M.A. 1989. Wetland seed banks. In: Leck, M.A., of seed banks in the vegetation dynamics of prai- V.T. Parker, and R.L. Simpson (eds.). 1989. Ecol- rie glacial marshes. Ecology 59:322-335. ogy of Soil Seed Banks. Academic Press, Inc., San Diego, CA. pp. 283-305.
5 van der Valk, A.G. and R.L Penderson. 1989. Seed banks and the management and restoration of natural vegetation. In: Leck, M.A., V.T. Parker, and R.L. Simpson (eds.). 1989. Ecology of Soil Seed Banks. Academic Press, Inc., San Diego, CA. pp. 329-346. Willson, M.F. 1992. The ecology of seed dispersal. In: Fenner, M. (ed.). 1992. Seeds: The Ecology of Regeneration in Plant Communities. CAB International, Wallingford, Oxon, UK. pp. 61-86.