Direct Seeding: Broadcast Native Seed at Scale

When Direct Seeding Beats Tube Planting

Raising seedlings in a nursery, hardening them off, transporting them to site, digging holes, planting, staking, and guarding each one individually is effective but expensive. At scale, the cost per hectare of tube planting can be prohibitive, particularly on large degraded landscapes where hundreds or thousands of hectares need treatment. Direct seeding, the practice of sowing seed directly into the ground where you want trees to grow, can cost a tenth to a third of tube planting per hectare and can cover ground far faster.

The trade-off is control. A nursery-raised seedling is a known quantity: it has an established root system, several months of growth behind it, and a high probability of surviving if conditions are adequate. A seed in the ground faces predation by rodents and birds, competition from established grasses, desiccation, fungal attack, and the simple probability that not every seed will germinate. Establishment rates from direct seeding typically range from five to thirty percent of seeds sown, compared to sixty to ninety percent survival for tube-planted seedlings. You compensate for this lower rate by sowing far more seeds at far lower cost.

Direct seeding works best for species with large, robust seeds that germinate readily and grow aggressively once established. Native oaks, walnuts, chestnuts, and many tropical hardwoods are good candidates. Small-seeded species that require precise conditions for germination, or species with very slow juvenile growth that cannot compete with grasses, are generally better raised in a nursery. The most effective reforestation projects often combine both approaches: direct seeding for the tough, fast-establishing species and tube planting for the sensitive, slow-growing ones.

Seed Collection and Provenance

The genetic quality of your seed matters as much as the quantity. Seed provenance, the geographic and ecological origin of the parent trees, determines whether the resulting seedlings are adapted to your site's specific conditions of climate, soil, pests, and diseases. Seed collected from trees growing within fifty kilometres of your site and at a similar elevation and aspect will almost always outperform seed from distant sources, because local populations carry generations of natural selection for local conditions.

Collect from healthy, vigorous parent trees with good form. Avoid collecting from isolated trees that may have been self-pollinated, as inbreeding produces weaker offspring. Collect from at least fifteen to twenty parent trees per species to capture adequate genetic diversity. This diversity is your insurance policy against future environmental change: somewhere in that genetic mix are individuals that will tolerate the conditions your site will experience in thirty, fifty, or a hundred years.

Timing of collection is critical. Seed must be collected when ripe but before it is dispersed or consumed by wildlife. For many species, the window is only a few weeks. Learn to recognise ripeness indicators: colour change, softening of fruit, or the moment when seed begins to fall naturally. Seed banking protocols cover collection, processing, drying, and storage in detail, and are worth studying before your first collection season.

Seed Treatment

Many native tree species have evolved dormancy mechanisms that prevent all seeds from germinating simultaneously. This is an excellent survival strategy in the wild, spreading germination risk across multiple seasons, but it is frustrating for restorationists who want predictable establishment. Seed treatment techniques break dormancy and synchronise germination.

Scarification involves physically or chemically breaking through a hard seed coat that prevents water uptake. Mechanical scarification means nicking, filing, or sanding the seed coat. Chemical scarification involves soaking seeds in concentrated sulphuric acid for a carefully timed period. Hot water treatment, pouring near-boiling water over seeds and allowing them to soak as it cools, works for many legume species and mimics the effect of bushfire passing over fallen seed.

Stratification simulates the cold winter period that many temperate species require before germination. Seeds are mixed with damp sand or vermiculite and refrigerated at one to five degrees Celsius for a species-specific period, typically four to sixteen weeks. Some species require a combination of warm stratification followed by cold stratification, mimicking the passage through summer and winter before spring germination. Research the specific requirements of each species before treatment; incorrect treatment can kill seed or fail to break dormancy.

Seed Balls and Fukuoka's Technique

Masanobu Fukuoka, the Japanese farmer and philosopher who developed natural farming, devised the seed ball technique as a way to broadcast seed across large areas without tillage or precise placement. A seed ball is a small sphere of clay, compost, and seeds, roughly the size of a marble. The clay coating protects the seed from birds, rodents, and desiccation while it sits on the soil surface waiting for rain. When rain comes, the clay softens, the compost provides a small nutrient boost, and the seed germinates into the moist soil beneath.

Fukuoka used seed balls to revegetate degraded hillsides in Japan and later in arid regions of Africa and India. The technique is elegant in its simplicity: mix seed with three to five parts clay and one part compost, add water until the mixture holds together, roll into balls, dry in the shade, and broadcast by hand or from a vehicle. Thousands of seed balls can be made in an afternoon by a group of volunteers, and broadcasting them across a site takes a fraction of the time required for tube planting.

The effectiveness of seed balls varies with climate and species. In regions with a reliable wet season, seed balls work well because the triggering rain event is predictable. In arid environments with erratic rainfall, seed balls may sit on the surface for extended periods and lose viability. For some species, the clay coating can be too thick, preventing the radicle from emerging. Experimentation with clay type, ball size, and seed-to-clay ratio is necessary for each site and species mix. Despite these limitations, seed balls remain one of the most accessible and scalable tools for community-based restoration.

Broadcast Methods and Establishment Rates

Hand broadcasting works for small sites and steep terrain where machinery cannot access. Walk systematic transects across the site, scattering seed at a consistent rate. A calibrated seed spreader, the type used for spreading fertiliser on farms, provides more even coverage than hand broadcasting and can be pulled behind a vehicle or ATV on accessible terrain. For very large sites, aerial broadcasting from helicopters or drones is increasingly cost-effective, though it requires careful calibration to achieve target seeding densities.

Site preparation before broadcasting dramatically improves establishment. Scalping or raking away existing vegetation and leaf litter to expose bare mineral soil gives seeds direct contact with the growing medium and reduces competition. On grass-dominated sites, targeted herbicide application or shallow cultivation before seeding can suppress grass long enough for tree seedlings to establish. Rolling or pressing seed into the soil surface after broadcasting improves soil-seed contact and reduces predation.

Expect establishment rates of five to fifteen percent in most conditions, with higher rates on well-prepared sites with good rainfall and lower rates on unprepared or arid sites. This means you need to sow five to twenty seeds for every tree you want to establish. At scale, this arithmetic still favours direct seeding over tube planting on cost. The trees that do establish from direct seeding are often more vigorous than tube-planted individuals because their root systems developed in place without transplant shock, and natural selection during the germination and seedling phases eliminates the weakest genotypes.