3 Seed bank
 | 3.1 Seed deposits |
| 3.2 Seed decay |
| 3.3 Germination |
Information concerning soil seed banks was included in the former report by Madsen, Poulsen and Streibig (1997) but as additional information was found useful for the modelling, we have included a short section in this report as well.
3.1 Seed deposits
The seed bank should literally be interpreted as a bank into which the species deposit their annual seed production. The size of the seed production is determined by the biomass production of the species at harvest time multiplied with an individual harvest index. Data on seed production are often presented as seeds per plant and not as a harvest index, but in this model we use biomass of a species rather than number of plants. Consequently we link biomass production to seed production through a harvest index.
Harvest index
Harvest index (HI) is usually a terminology used for crops and denotes the harvested part of the crop e.g. for cereal grains, relative to the total above ground biomass production. Few data are available on harvest index of weed species. Rasmussen (1993) studied seed production in Chenopodium album as a function of plant biomass. The seed production multiplied with an average seed weight from the literature of 1.2 mg (Korsmo, 1981) resulted in a HI of 40-50% for this species. Norris (1996) studied seed production in Capsella bursa-pastoris of two populations. In this species, HI varied from 10 to 28% depending on population. Andersson (1994) studied seed production and seed weight as a function of herbicide treatment with MCPA. We have recalculated HI from these results in Table 3.1.:
Table 3.1
Harvest index for selected weed species. Recalculated after results obtained by Andersson (1994)
| Species | HI |
| Polygonum convolvulus | 52-64% |
| Chamomilla. recutita | 1-6% |
| Chenopodium album | 22-48% |
| Galium aparine | 15-27% |
| Thlaspi arvense | 5-30% |
Predation
Before the seeds enter the seed bank many will be predated by insects, birds and small mammals. It is difficult to give an exact number for this predation, but up to 50-90% seed losses are possible (Statens Planteavlsforsøg, 1993).
Immediate germination
Depending on species many seeds are able to germinate immediately. Plants from seeds that germinate before the soil is cultivated for the succeeding crop will perish.
Estimated value
The proportion of viable seeds that enter the seed bank relative to the seed production is estimated for each weed and volunteer species in the model. This estimated value is based on the following: Is the species known to have pronounced seed dormancy, and are the seeds known to be a food source for the fauna?
3.2 Seed decay
The seed number in the total seed bank decreases exponentially (Wilson & Lawson, 1992). Decrease in cultivated soil is 2-3 times greater compared to undisturbed soil (Roberts and Feast, 1973). The rate of annual decline varies somewhat in different studies: Wilson and Lawson found a total yearly decline of 50% in a study, whereas Statens Planteavlsforsøg (1993) mentions an average of 33% per year. Roberts and Feast (1973) mention an average seed decay of 32% one year after burial of weed seeds. Decay rates for individual weed species varie as shown in Table 3.2.
Table 3.2
Seed bank persistence of common weeds in natural, cultivated soil. After Roberts and Feast (1973)1 and Wilson and Lawson (1992)2.
| Species | Annual decay |
| Stellaria media | 49-56%1 |
| Chenopodium album | 33-38%1 |
| Capsella bursa-pastoris | 36-52%1 |
| Poa annua | 37-50%1 |
| Viola arvensis | 36%2 |
| Lamium purpureum, Myosotis arvensis, Galium aparine | >50%2 |
In the model the exponential seed decay is based on an annual decay rate of 33%.
3.3 Germination
The number of weed seeds that germinates from the ploughing layer (the top 20 cm soil) varies considerably. Few weeds are able to germinate from lower depths and therefore the cultivation is an important factor. Roberts and Feast (1973) found that the number of seedlings were twice as great when cultivation had taken place compared to undisturbed soil. In their experiments, an average emergence of 20-25% of the seeds plants emerged as plants during the first year, and 15% during the second year after burial, and on average over the first six years after incorporation in the soil, the seeds germinated with approx. 9% annually in cultivated soil. Wilson and Lawson found 0.1-5% germination, and Hurle et al. (1988) found a germination frequency of 0.5-9%. In the model the germination frequency is assumed to be 5% for all weed species.