Collecting And Germinating Seeds From Soil Seed Banks

EHUF490b TERM PROJECT, GROUP 4, SPRING 2003

June 10, 2003　　

Seed banks of forest ecosystems

Seed banks can serve as gene pools by preserving populations of plants and preventing extinction (Hills, 1992).  Seed banks can enhance the negative effects of early successional or competitive species, thus contributing to forest species richness (Bossuyt, 2002). There is a lot of variation within a seed bank that depends on time and spatial features.  Temporal factors include: seed production and dispersal, habitat quality, seed germination, and environmental fluctuations.  Spatial factors include distribution of source plants, timing of dispersal, size of crop, seed quality, and suitable gaps (Hills, 1992). Seed placement in the seed bank varies with many factors. Small seeds move deeper in to the profile with the aid of rainwater, while larger seeds usually remain closer to the surface (Hills, 1992).

The abundance of seeds in the seed bank is dependent on many factors including the amount of seed production by a particular species. Early and mid-successional species tend to be well represented, but the number of viable seeds depends on the age and composition of resident vegetation (Hills, 1992). Trees do not produce seeds evenly from year to year. For instance, ponderosa pine (Pinus ponderosa) produces abundant cone crops every 4-5 years while lodgepole pines (Pinus contorta) are capable of this every 1-2 years (Schneph). This variability can be brought about by both environmental and physiological variables.

       Despite the formation of cones, there can still be factors that hamper seed bank formation, such as poor seed development and predation of the seeds by insects and animals. The latter of these is a factor, in of itself, which is dependent on external factors and is subject to flux in intensity.  Even if seed production proceeds under ideal conditions, once the seeds fall to the forest floor they are subject to another round of predation.  In order to germinate, the forest floor micro-climate conditions must be ideal for the highest likelihood of germination.  For conifer trees this means that the seeds land on bare mineral soil, with a minimum of organic matter (Schneph).

       Coniferous trees and their seeds face a number of disadvantages that influence the ability of the seed bank to regenerate the forest.  Many conifers in the inland Northwest produce seeds that are not viable after about two years once they have reached the ground (Schneph). Countering this, many shrubs have seeds with much longer seed viability windows than the trees.  For instance, redstem ceanothus (Ceanothus sanguineus) seeds are known to remain viable for decades within the seed bank (Schneph).  In addition to this, many shrubs, as opposed to most coniferous trees, are capable of reproducing vegetatively (Schneph).  This characteristic gives shrubs an advantage to reproduce and grow, out-competing and out-shading tree seedlings.

       Natural revegetation is the product of trees producing so many seeds that they are able to prevail over the above mentioned obstacles and also the careful planning, management, and timing of revegetation activities by humans (Schneph).  Activities such as prescribed burning and the use of mechanical equipment for logging purposes often clear land leaving behind soil conditions that allow for the germination of coniferous seeds.  The trick to this is finding an equilibrium between creating soil conditions needed for the seed bed and those that do not cause soil erosion and compaction.

The management of the forest can have an influence on the seed production of a site. In locations with overstory removal of plant material, the abundance of seeds in the seed bank can be higher than the surrounding intact forest (Hughes, 1991). Hughes and Fahey (1991) also found a strong correlation between the growth of seedlings and the locations of predecessors that once grew in pre-disturb sites. This shows how important disturbance and prior site conditions can be in determining the make up of a site after a disturbance event has occurred. Without disturbance little understory change usually takes place leading to semi-stable successional conditions (Hughes, 1991). After disturbance, gaps are repopulated by existing vegetation, the seed bank, seed rain, and the seedling bank (Hills, 1992). Three times as much seed can be found in disturbed forest as undisturbed ones (Hills, 1992).

Benoit et al. (1998) conducted an experiment on how best to sample from the seed bank to get an accurate picture of what lies beneath. The best technique proved to be the use of a small number of large sampling units. No fewer than 60 samples for abundant weeds worked best; no matter which method of sampling was used: systemic, stratified random, or cluster sampling techniques (Benoit, 1998). Sampling is time and labor intensive and if done improperly can lead to non-representative results. This can be caused by collecting too near parent plants rather than randomly.  This may alter the results of the sample, especially by increasing the proportion of less common plant species collected.

Once collection of seeds from the bank has occurred there are a couple techniques used to identify what is located in the seed bank. These techniques include the emergence and seed extraction methods. With the emergence method, Brown (1992) found that no tree or shrub was found, while the separation method yielded 8% of the available bank. The two methods do not produce similar estimates, which needs to be taken into account when comparing seed bank reports and when sampling the seed bank for an experiment. Seed extraction has higher variability in species number, density and diversity (Brown, 1992). If all that is needed is to test whether a seed is viable, the floatation method can be used for this purpose. Viable seeds float while nonviable ones sink.

In the forest of British Columbia, Kellman (1970) found that more viable seeds were found near the surface than deeper in the soil profile. Mladendoff (1990) found that dominant tree species are not an important component of the seed bank; though bird and ant dispersed plant seeds are more prolific. In many forest the seed bank density can range from 151 to 2157 seed/m².

Non-forest species are usually a minor component of forest seed bank. Variation in the richness of the seed bank is positively correlated with the amount of canopy cover, soil moisture and nutrients. Seed populations do not correspond well with above vegetation, but rather like that of an earlier stage of succession (Hills, 1992). However, there is no such correlation with seed bank density or the total number of species that exist aboveground (Leckie, 2000).

The above information can be used as a guideline when collecting and propagating forest species from the seedbank.  Germination of seeds from soil seed banks may be a method of obtaining a variety of species quickly and efficiently.  While seed collection directly from parent plants is the more common collection method, it is time intensive and, in the case large of tree species, is often difficult and requires tools such as a cone rake.  In addition, if several species are desired, multiple collections must take place to collect from the parent plants at their appropriate times.  Seed banks, in contrast, can store a wide variety of viable seeds and are easily harvested.  As such, collection and propagation of the seed bank may be viable alternative to traditional seed collection methods.  This method should be explored further and utlized in the restoration of forest ecosystems.