Seeding, Monitoring and Maintenance

Depending on the time of year and the species, artificial stratification of the seed may be required.  Ideally the seed would be sown in autumn shortly following collection, so that it is subject to natural stratification processes. Sowing as soon as possible after collection also removes potential human errors associated with storing the seed incorrectly. Some seed has to be stored in very particular conditions if it is going to be stored for any length of time otherwise it may not be particularly viable by the time it comes to sowing. It may be that the seed cannot be sown until spring because weed control works are required first. In this instance, professional advice should be sought to ensure that the seed is stored and stratified correctly, remembering that different species can have very different requirements.

In order to measure the success of a seeding operation, it is necessary to know how much seed has been sown over the area. Measuring out a very small amount of seed, e.g. 5 grams, and counting how much seed is in that amount, then multiplying to achieve the desired amount of seed is always going to be a lot quicker than trying to count out the exact number of seeds, as shown in the example below.


Coronet & Queenstown Hill fenced 4m2 per plot (20m2 total per site)

Coronet & Long Gully unfenced 5m2 per transect (25m2 total per site)

60 seeds per gram

8,000 seeds per 4m2
(2,000 seeds per m2 (50% viable))
= 133g per transect
= 33g/m2

15,000 seeds per 5m2 transect
(3,000 seeds per m2 (50% viable))
= 250g per transect
= 50g/m2

100 seeds per gram

4,000 seeds per 4m2
(1,000 seeds per m2 (80% viable))
= 40g per transect
= 10g/m2

7,000 seeds per 5m2 transect
(1,400 seeds per m2 (80% viable))
= 70g per transect
= 14g/m2

Example of how the number of seeds can be calculated. Note that the beech seed was not as well sorted as the coprosma and so there were fewer seeds per gram.

Seed viability should also be tested at this stage and more seed added to compensate for the percentage of non-viable seed. The rate of sowing depends on the species, the area to be covered and the amount of viable seeds available. Around 1,000 seeds per m2 is a good starting point for mānuka or kānuka, but a much lower rate can be applied when sowing with coprosma or broadleaf.

Sowing seed can be done simply by sprinkling the seed out over the prepared area, or a seed spreader can be used for some species such as beech and coprosma. A thin layer of soil can be spread over the top of the seed to enable germination and also to make the seed less obvious to birds and rodents. More recent trials have also examined ways of dispersing seed aerially via drones.

Care must be taken to avoid seeding into unnatural habits e.g. if a species usually grows best in dry, sunny areas then it shouldn’t be seeded into a dark gully.

Spreading seed by hand and using a seed spreader

Spreading seed by hand (left) and using a seed spreader (right)

Mycorrhizal Inoculation

Native plants form beneficial symbiotic relationships with mycorrhiza and so if the right mycorrhizae species are present in the soil, then the native plants will flourish. Most New Zealand native species form symbiotic relationships with a range of arbuscular mycorrhizae, but some species, such as New Zealand native beech, need a more particular ectomycorrhiza.

Any attempts to translocate beneficial mycorrhizae should be approached with extreme caution. There are a number of commercially available mycorrhizal products available, but these are usually designed for horticultural purposes, not native forest restoration projects. New Zealand native mycorrhizae will be more beneficial than shop-bought versions, however, translocating duff and fungi creates the risk of introducing undesirable, invasive exotic weed and fungal species. Exotic fungi facilitate weed invasions and can be counter-productive to native seedling growth.

Sprayed weeds


Simply sowing seed and walking away will not ensure the best chance of success for a seeding project. Weed control and animal control (and maybe even watering where possible) for the first 2 – 3 years at least will increase the chances of success remarkably. The aim is to establish a closed canopy and shade the soil to help prevent shade-intolerant weeds from invading, so the more than can be done to facilitate seedling growth in the first few years, the better. Again, this is a reason why targeting microsites, and doing this well, will be more effective than trying to restore a whole site at once.

Releasing, which is the removal of weeds from around the seedlings, will ensure more rapid growth of the seedlings because there will be less competition for light, water and nutrients. Releasing can be undertaken using specific herbicide (taking care not to spray the seedlings) or by hand pulling the competitive weeds. The type of weed management required depends on the type of weeds, the degree of infestation, and how much time, labour and funding is available.

Depending on the fencing already in place and the pest animals present, it may be desirable to install tree protectors around seedlings once they reach a certain size to provide a higher degree of protection. This may also reduce the amount of weed control required.

Another way to reduce the amount of weed control required is to mulch the area around the seedlings with sawdust (must be from untreated timber), bark chips, or even using a mat of pine needles from the surrounding area. Care must be taken not to cover any other emerging seedlings. Mulching also helps with reducing moisture loss from the soil, stabilises soil temperature and organic mulches may also add nutrients to the soil.

The more intensely a site is managed, the greater the chance of success. However, the amount of maintenance required can be reduced through careful site and species selection, adequate site preparation and weed removal prior to seeding. If species are sown into an unsuitable habitat without adequate consideration and control of risks factors such as animals and weeds, then the level of maintenance required will be a lot higher.

Thorough planning and development of a long-term restoration strategy prior to collecting or sowing any seed can ensure that the amount of maintenance required is minimal, which can make seed-based restoration strategies a viable alternative to more traditional techniques.

Seedling Monitoring


Monitoring the success of a seeding project will provide valuable insight into what did and didn’t work, and which success factors need more or less attention. This will provide site-specific insights to help guide the next phase of restoration to ensure the level of success increases year upon year.

In order to determine the amount of germination and seedling survival, it is first important to know how many seeds were broadcast into an area. Seed counting methods are discussed above. The size of the area into which the seed are broadcast must also be recorded so that the seeds per m2 seeding rate can be calculated.

It is not feasible to survey the whole area and count the number of emerging seedlings and so it is more pragmatic to use a measuring square to mark out a smaller area and count the number of seedlings within. For example, a measuring square made of plastic piping that is 500 x 500 mm in size can be used to count the number of seedlings in an area measuring 0.25 m2.

Measuring square

Measuring square made from plastic piping

The measuring square should be placed randomly within the seeded area and held in position using pegs. These pegs should be clearly marked and left in place so that the measuring square can be placed in exactly the same position for subsequent monitoring events.

The number of seedlings, species, and height should be counted and recorded along with the plot number and date. The measuring square can then be lifted and placed randomly in another location, and the process repeated. The number of plots measured depends on the size of the area seeded, the degree of variation across the site and the level of detail sought. What is important is that the same plots can be measured again later in the season to compare germination and survival rates as time progresses.

For areas sown in autumn, it might be suitable to undertake the first monitoring round in mid-summer. For areas sown in spring, it’s probably best to wait until late summer. Another round of monitoring should be undertaken just before winter, and again in spring, to determine survival rate through winter. The frequency of monitoring thereon after depends on how many seedlings have germinated and how rapidly conditions have changed.


The Wakatipu Beech Seeding Project was made possible with support from the Wakatipu Wilding Conifer Control Group Inc (WCG), the Wakatipu Reforestation Trust, Scion (NZ Forest Research Institute Ltd) and the University of Otago.

Special thanks to Jon White, Thomas Paul, Janice Lord, our team of enthusiastic volunteers and the owners of the land on which the trials were conducted.

This report was written by Hilary Lennox of Ahika Consulting Ltd with presentation by Daniel Sweeney from Total Brand Ltd.