Previous Page Table of Contents Next Page

V. Seed Viability

What is seed viability?

The viability of the seed accession is a measure of how many seeds are alive and could develop into plants which will reproduce themselves, given the appropriate conditions.

Why do we test seed viability?

It is important to know that the seeds that are stored in a genebank will grow to produce plants. Therefore they must have a high viability at the start and during storage. The viability of seeds at the start of storage will also determine, within the environmental conditions, the storage life of the accession.

When should viability be determined?

Viability will need to be determined at the start of storage and at regular intervals during storage to predict the correct time for regeneration of the accession. The viability test takes from a few days to weeks or even months to give an accurate result. If possible the results should be available before the seeds are packaged and placed in the genebank so that poor quality seeds can be identified and regenerated before storage. Where the viability cannot be determined before storage, the seeds should be placed into long-term storage to ensure their safety whilst awaiting the results of the test.

How should viability be determined?

The most accurate test of viability is the germination test and this will be described here. The germination test is made under controlled conditions to find out how many seeds will germinate and produce normal seedlings which could develop into normal reproductively mature plants. The IBPGR Advisory Committee on Seed Storage recommends that for the initial germination test of species where a reasonable germination technique is available, a minimum of two replicates using 200 seeds (100 seeds per replicate) is acceptable, providing that germination is above 90%. If not, a further 200 seeds should be tested as before and the overall result for seed viability taken as the mean of the two tests. Other biochemical tests are available to test viability. These have the advantage of being quicker, but are not as accurate and require considerable skill and practice in their implementation and interpretation. These are not recommended by the IBPGR Advisory Committee on Seed Storage for general use as tests for seed viability.

Germination testing

How is germination tested?

There are many methods available which can be used to test germination. Seeds of different species have different requirements for light, substrate, temperature and water during germination. For an initial germination test the IBPGR Advisory Committee on Seed Storage recommends that a fixed sample size test should be used.

How do we choose which conditions to use in the germination test?

There is no general set of conditions which can be used to germinate seeds of all different species. Each species is different and must be treated separately. Guidelines for germination of seeds of the most common crop species and some of their wild relatives which are likely to be stored in genebanks are given in Vol. II of Ellis, Hong and Roberts (1985). Seeds of some species are more tolerant and can germinate under a wide range of conditions but complete germination will only be achieved under optimum conditions.

Why do some seeds fail to germinate in the test?

Two main reasons for the failure of seeds to germinate in suitable conditions are because they are either dead or dormant. Dead seeds can be identified because they usually soften and rot during the test as a result of attack by bacteria and fungi. Seeds which remain hard or absorb water, but remain firm and in good condition during the germination test are probably dormant. Seed dormancy is common in some crops straight after harvest (post harvest dormancy) and in many wild species related to crop plants.

How can dormant seeds be stimulated to germinate?

Special treatments are required to overcome seed dormancy. Treatments vary among species and more detailed advice can be found in Vol. II of Ellis, Hong and Roberts (1985).

How should non-dormant seeds be germinated?

The most appropriate test conditions should be selected for the test. A summary of suitable test conditions is given in Vol. II of Ellis, Hong and Roberts (1985). A fixed sample size germination test should be used as explained here.


1. Use the normal germination test procedure for the species in your genebank.

2. If there is no standard, ask the curator for advice on the best germination test conditions to use.

3. Check that the equipment and environments are available to fulfil these conditions.

4. If not, a compromise will have to be found using the best possible alternatives available.

5. A fixed sample size germination test using 200 seeds is recommended to determine viability at the start of storage.

Notes and Examples

Choosing the best conditions for germination is difficult because species and even accessions of the same species vary in their requirements. Methods for germinating seeds of major crop species are usually well-documented. Often there is no information on the methods needed for lesser known crops or wild and weed species.

The viability at the start of storage should be determined accurately to allow more accurate predictions of the storage life of the seeds. The germination test should be statistically accurate and therefore sufficient seeds should be used for the test. The sequential test described in Section VIII less seeds and is recommended for monitoring but not for the initial test for viability.


1. Take a random sample of seeds from the accession.

2. Count out 200 seeds for each test.

3. Divide these seeds into at least two replicates. More replicates of fewer seeds in each can be used to fit into the equipment at your genebank.

4. In order to prevent damage to the seeds from the rapid uptake of water (imbibition damage), seeds with a low moisture content must imbibe water slowly. Any seeds known to be prone to imbibition damage should be placed in a humid atmosphere to equilibrate as a pre-treatment. This is especially important for many species of legumes.


Although several methods are available for germination testing, it is suggested that you use one of the two major methods described here because these are adaptable to most environments and in general give uniform results.

Germination using absorbent paper

Top of paper method

1. Cut the paper to the size and shape of the dishes.

Notes and Examples

Several types of absorbent paper are available. Choose paper that is relatively cheap, easily available and strong enough to withstand handling and the weight of the seeds when damp. It must be free from chemical residues that could interfere with the germination of the seeds. The most common papers used are towelling, filter paper or rice straw paper. Filter paper and paper towelling are more expensive. Rice straw paper is cheap but fungal spores from the plant material can cause contamination unless heat-treated before use.

The petri dish/container method has the advantage that the seeds can be observed through the transparent lid and the germinating seeds are therefore easy to count. There is the disadvantage that the dishes soon dry out and need daily watering.

Petri dishes or other containers
Filter paper, paper towelling or rice straw paper
Permanent markers
2. Place a layer of paper in each dish. If the paper is too thin use a double layer.

3. Label the top and bottom of each dish with the accession number, number of the replicate and date of the test.

4. Moisten the paper with water.

Notes and Examples

The thickness of paper used should be more than 2 mm when moist.

The water used must be reasonably free from acid, alkali, organic material or other impurities. It can be either tap, distilled or de-ionized water.

5. Arrange the seeds in a regular equidistant pattern on the surface of the paper.

6. Add more water if required and replace lid.

Notes and Examples

Seeds arranged in an orderly way are easier to count.
Between paper method

1. Cut the paper to a convenient size to hold one replicate of the seeds when spaced at regular intervals. If the paper is too thin use a double layer.

2. Label each sheet on the outside of the paper at one end with the accession number, replicate of the test and the date of the start of the test.

3. Moisten the paper with water.

Notes and Examples

The between absorbent paper method is cheap and easy to prepare, but the seeds cannot be observed without unrolling the paper. The use of new paper at each test has the advantage that fungal contamination cannot be carried from one test to another.
Filter paper, paper towelling or rice straw paper
Plastic bags
Permanent marker
4. Arrange the seeds at regular intervals on the paper, leaving at least two centimetres clear from the edges all round.

Notes and Examples

Leave enough room around the edges of the paper so that the paper can be folded back without damaging the seeds.
5. Cover the seeds with another sheet of paper and fold in the edges to prevent the seeds from falling out.

Notes and Examples

Folding the edges is especially important for large spherical seeds which tend to fall out because of their weight and shape.
6. Roll the paper loosely towards the end with the label.

Notes and Examples

Do not roll the paper too tightly because the compression and lack of oxygen resulting from this can cause the seedlings to develop contorted roots and shoots. These are hard to separate from abnormalities caused by genetic factors.
7. Place the rolled papers inside ventilated plastic bags or boxes.

Notes and Examples

Oxygen is essential for respiration during germination. Therefore, any containers used should be adequately ventilated.
8. Put these in an upright position in wire baskets or plastic boxes.

9. Keep the paper moist with water.

Notes and Examples

Root and shoot development is geotropic. Placing the rolled papers in an upright position allows seedling development in the natural vertical position.
Germination in sand

1. Pack clean sterile sand into pots or trays with drainage holes in the bottom.

2. Water the sand until it is moist. Do not use excess water.

Notes and Examples

Fine sand should be used. Make sure that the sand is clean by sterilizing before use. Quarry or river sand is better than shore sand, which must be washed thoroughly to remove all salt.
Permanent markers
Pots or trays
Tools for planting
3. Make holes in a regular equidistant pattern at about the same depth as the size of the seeds for each replicate of the seeds. Ideally, the distance between holes should be at least three to five times the seed diameter.

4. Prepare a label with the accession number, date of sowing and replicate of the test and place in each pot or tray.

Notes and Examples

Germination in sand is especially useful for large seeds which are too large to be germinated in petri dishes or too heavy for the between paper method.
5. Fill seeds from each replicate into the holes and cover with sand.

6. Water the sand again to cover the seeds, but do not make it too wet.

Notes and Examples

Sprinkle with water slowly, so that the seeds do not float out from the holes and become mixed. Bottom watering is better than top watering.

1. Place the prepared germination tests under the light and temperature conditions which were determined in Step 1.

2. Keep the substrate moist during the test with water, but do not over-water.

3. Run the test for a sufficient period to determine whether the seeds have germinated, or are dead or dormant. A fixed interval can be used based on previous experience, but the time taken to germinate will vary among species and accessions of the same species.

Notes and Examples

Seeds absorb most water during a rapid imbibition phase at the start of the test. Make sure that there is enough water at this stage but remember that too much water at a later stage can limit the oxygen for the seeds and also lead to rotting.

Germination using absorbent paper

1. Seeds sown on top of paper can be counted either through the transparent cover or after removing the cover of the container. For those between paper, unroll the paper carefully to avoid tearing the paper or damaging the roots of the young seedlings.

2. Count and record the number of normal seedlings in each replicate.

3. Count and record the number of abnormal seedlings in each replicate. Examples of the types of abnormal seedlings found in different species have been described by ISTA (1976b).

4. Once a seed has been counted as germinated, the resulting seedling can be discarded.

5. If the seeds are dormant, treat with the appropriate technique to stimulate germination as advised in Vol. II of Ellis, Hong and Roberts (1985) and continue the test to allow those to germinate.

Notes and Examples

Normal seedlings are those which show the capacity for continued development into normal plants if grown on a suitable substrate and under favourable conditions of water supply, temperature, light and oxygen. Abnormal seedlings are those which do not appear as if they will develop into normal plants. There are three classes of abnormality - damage, deformity and decay.
Germination in sand

1. In this case the number of seedlings emerged from the sand is usually counted because early root growth cannot be seen without disrupting the test.

2. Count and record the number of normal seedlings emerged in each replicate.

3. Count and record the number of abnormal seedlings emerged in each replicate.

4. If any fungal or insect infestation is seen during the germination test, sterilize the sand before using again.


1. For each accession, consider the total number germinated in each replicate.

2. Check that the results of the different replicates are compatible by using tolerance tables (see chapter 14 Ellis, Hong and Roberts, 1985).

3. If the results are compatible, calculate the mean percentage viability of that accession from the results of all replicates.

4. If the results are not compatible consult the curator for advice on how to proceed.

Notes and Examples

Tolerance tables can be consulted to check that the results of your work are valid because all replicates of a test should be compatible and within certain limits.

Here a problem may arise. If the results of one replicate are not compatible with the others, the curator will have to decide either to repeat the test or accept these results. It may be possible to accept the results of other replicates and ignore the one that is not compatible only if there is a good reason why one replicate differs, e.g. the dish dried out during the test, etc.


1. Consider the mean percentage viability of the accession.

2. If it is above 90%, accept the test as valid and use this value as the true viability.

3. If the result is 90% or below, repeat the test using a further 200 seeds following the same procedures.

4. Calculate the mean percentage viability from the results of the two tests and use this as the overall test result.


1. If the standard test for that species in your genebank has been followed exactly, enter the percentage viability into the inventory data file for each accession.

2. If a standard test has not been used, either enter a reference to the full germination data or state the conditions which vary from the standard test.

Notes and Examples

The conditions of the germination test can influence the number of seeds which germinate, especially if seed vigour is low. Optimum conditions allow optimum germination. Therefore, it is important to note the conditions of the test and any deviations from the standard test.

Table of standard germination test regimes for your genebank

Fill in this table for your future reference:






Test Length


Previous Page Top of Page Next Page