Starting a seed bank for long-term storage comes down to controlling three things: moisture, temperature, and oxygen. Dry your seeds to a low moisture content, seal them in airtight containers, and keep them cold. That simple framework can preserve the viability of most common crop and garden seeds for years or even decades. But the details matter enormously, and getting them wrong can mean opening a container years later to find nothing but dead seed. What follows covers the practical steps, the science behind them, and the decisions you will face along the way.
Not All Seeds Can Be Stored the Same Way
Before you invest effort in drying and freezing, you need to know whether your seeds are the kind that tolerate it. Seeds fall into three broad categories based on how they respond to drying. Orthodox seeds, which include most vegetables, grains, and many wildflowers, can be dried to very low moisture levels and stored cold for years without losing viability. Recalcitrant seeds, like those from oaks, avocados, and many tropical trees, die when dried below a certain moisture threshold and cannot survive conventional cold storage. A third group, called intermediate, falls between the two and can handle some drying but generally stays viable for less than a year under standard seed bank conditions.1IOP Conference Series: Earth and Environmental Science. Seed handling of specific forest tree species: Recalcitrant and intermediate seed
The boundary between orthodox and recalcitrant isn’t always black and white. Researchers describe it using a water content threshold: orthodox seeds survive desiccation down to roughly 0.07 grams of water per gram of dry weight, while recalcitrant seeds lose viability above about 0.2 grams of water per gram of dry weight.2PubMed. Orthodoxy, recalcitrance and in-between: describing variation in seed storage characteristics using threshold responses to water loss For a home or community seed bank, the practical takeaway is straightforward: most garden vegetables, herbs, cereals, and legumes are orthodox and respond well to standard drying and cold storage. If you are trying to bank tree seeds or tropical species, look up their storage behavior first. Treating a recalcitrant seed like an orthodox one will kill it.
Why Drying Is the Most Important Step
Seeds age. Even under ideal conditions, chemical reactions slowly degrade cell membranes, proteins, and DNA. The main culprit is reactive oxygen species, which attack the fatty acids in cell membranes and damage genetic material over time.3PubMed Central. Reactive Oxygen Species as Potential Drivers of the Seed Aging Process In soybean seeds, researchers have shown that as antioxidant defenses decline during storage, lipid breakdown and electrolyte leakage both increase, essentially the seed’s internal membranes falling apart.4PubMed Central. Dynamic Changes in Membrane Lipid Metabolism and Antioxidant Defense During Soybean (Glycine max L. Merr.) Seed Aging
Removing water slows all of these reactions dramatically. When orthodox seeds dry sufficiently, the sugars and other molecules inside their cells form a glass-like solid state. This vitrification essentially locks cellular components in place, preventing the molecular movement that would otherwise cause degradation.5Plant Physiology. Glass Transitions in Soybean Seed 1: Relevance to Anhydrous Biology Maintaining that glassy state is critical. If temperature or moisture rises enough to break it, oxidation and crystallization resume and the seed deteriorates rapidly.6Comparative Biochemistry and Physiology Part A: Physiology. Cytoplasmic Vitrification and Survival of Anhydrobiotic Organisms
For practical purposes, aim to dry your seeds to a moisture content between about 3 and 7 percent. Professional genebanks typically target around 5 percent for most species. You can accomplish this at home with silica gel desiccant: place seeds in a sealed container with an equal weight of indicating silica gel (the kind that changes color when saturated), swap out the gel every day or two, and repeat until the gel stops changing color. In dry climates, air-drying seeds in a shaded, well-ventilated space for a week or two before finishing with desiccant often works well. The key is patience; rushing the process with heat above about 35°C (95°F) can damage the embryo.
Choosing Containers and Sealing Methods
Once seeds are dry, the goal of packaging is simple: keep moisture and oxygen out. Every time a container lets in humid air, the seed reabsorbs water and the protective glassy state breaks down. For small-scale seed banking, glass mason jars with good rubber seals work surprisingly well. So do heat-sealed Mylar bags or aluminum-laminated pouches. Research on onion seeds, which are famously short-lived, found that vacuum-sealed bags combined with cold storage maintained the best viability after 18 months compared with other packaging methods, with aluminum laminated bags performing second-best.7Journal of Advances in Biology & Biotechnology. Evaluation of Physiological and Biochemical Parameters of Onion Seed as Influenced by Different Packaging Materials and Storage Conditions
A few practical tips on sealing:
- Add desiccant inside: A small packet of silica gel inside each sealed container acts as insurance against any residual moisture or minor leaks.
- Remove air: Vacuum sealing is ideal but not essential. Even displacing air with an oxygen absorber packet inside a Mylar bag substantially slows oxidation.
- Use opaque containers: Light accelerates some degradation reactions. If you use glass jars, store them in the dark.
- Label everything before sealing: Species, variety, collection date, collection location, and number of seeds. Once a bag is heat-sealed, you do not want to open it just to figure out what is inside.
Temperature and Where to Store Your Seeds
Cold slows chemistry. The colder the storage, the longer seeds remain viable, assuming they are dry enough. Professional genebanks typically store seeds at around -18 to -20°C (roughly 0°F), which is the temperature of a standard household chest freezer. That is a perfectly adequate base-collection temperature for a home seed bank. If you lack freezer space, a refrigerator at about 4°C (39°F) is a solid second choice and still extends seed life many times over compared with room-temperature storage.
Room-temperature storage is the weakest option, especially in warm or humid climates. Seeds stored at 28°C lose viability far faster than those kept at 10°C or below; a study on rice seeds demonstrated that cold-stored seeds maintained strong germination rates over 12 months while room-temperature seeds declined substantially over the same period.8Seed Science and Technology. Tetrazolium Test for Evaluating Viability of Stored Rice (Oryza Sativa) Seeds For communities in tropical or resource-limited settings, even low-cost approaches to reducing temperature and humidity during storage, such as hermetic containers with desiccant stored in the coolest available room, make a meaningful difference.9Experimental Agriculture. Low-cost seed storage technologies for development impact of small-scale seed saving entities in tropical climates
One caution about freezing: seeds must be dry before they go into the freezer. If there is too much moisture inside the seed, ice crystals form during freezing and rupture cells. This is the same reason recalcitrant seeds cannot be conventionally frozen. For well-dried orthodox seeds, freezing poses no risk and is the single best thing you can do for longevity.
Testing Whether Your Seeds Are Still Alive
A seed bank is only useful if the seeds in it can still germinate when you need them. Periodic viability testing tells you when a lot is declining and needs to be grown out and replaced. The simplest test is a germination trial: take a sample of seeds, place them on moist paper towels or in seedling trays, provide appropriate warmth, and count how many sprout over the species’ normal germination window. This is the gold standard, but it takes time and uses up seeds from your limited stock.
A faster alternative is the tetrazolium test. Seeds are soaked, then cut open and immersed in a tetrazolium chloride solution. Living tissue stains red, while dead tissue stays colorless. For rice seeds, a modified tetrazolium method produced results that correlated extremely closely with standard germination and field emergence.8Seed Science and Technology. Tetrazolium Test for Evaluating Viability of Stored Rice (Oryza Sativa) Seeds Across a range of grass species, the tetrazolium test and standard germination test agreed on viability for over 90 percent of samples in most species tested.10Crop Science. Can the Tetrazolium Test be Used as an Alternative to the Germination Test in Determining Seed Viability of Grass Species? The tetrazolium test has the advantage of speed, giving you results in hours rather than days or weeks, and it can reveal viability in dormant seeds that might fail to germinate in a standard test simply because they have not had their dormancy broken.
How often should you test? For a small home bank, checking a sample from each lot every three to five years is a reasonable starting point. If viability drops below about 75 to 85 percent for a given accession, it is time to grow out a new generation of seeds.
Collecting Seeds Worth Saving
The genetic diversity inside your seed bank matters as much as the storage conditions around it. If all your seeds of a given crop come from a single plant, you have stored a very narrow genetic base that may not adapt well if conditions change. For wild species, the question of how many individuals and how many populations to sample from has been studied carefully.
Research suggests that sampling about 25 to 30 individual plants per population, from a few geographically spread-out populations, captures a good share of the species’ genetic diversity.11Biological Conservation. Optimal sampling of seeds from plant populations for ex-situ conservation of genetic biodiversity, considering realistic population structure But the numbers depend on what you are trying to capture. One study of an endangered plant found that 30 individuals from a genetically diverse population was enough to capture 90 percent of common alleles, but grabbing 90 percent of all alleles, including rare ones, required collecting from 60 or more individuals.12Conservation Genetics. How much is enough? Minimum sampling intensity required to capture extant genetic diversity in ex situ seed collections: examples from the endangered plant Sibara filifolia (Brassicaceae)
Breeding system matters too. Self-pollinating species concentrate their genetic diversity between populations rather than within them, which means you need to visit more sites to capture the same amount of overall diversity. For a selfing species, researchers recommend sampling from roughly twice as many populations as you would for an outcrossing species.13PubMed Central. Capturing Genetic Diversity in Seed Collections: An Empirical Study of Two Congeners with Contrasting Mating Systems For garden crops, the practical lesson is to save seed from as many healthy, productive plants as you can manage, not just the single best one.
Record Keeping and Accession Management
A seed bank without records is just a box of mystery seeds. Even a small personal collection benefits from consistent documentation. At minimum, record the species and variety name, the date seeds were collected, where they were collected from or which seed lot they came from, the initial viability test result, the date they went into storage, and the quantity stored. If you are saving wild-collected seeds, add GPS coordinates or a site description and notes on the habitat.
Professional genebanks use dedicated database systems to manage hundreds of thousands of accessions. India’s National Genebank, for example, maintains records on over 320,000 accessions across nearly 1,200 species using specialized software designed for structured data capture and retrieval.14Computers and Electronics in Agriculture. Application note Genebank Information Management System (GBIMS) You do not need anything that complex. A spreadsheet works for a home or community bank, as long as you keep it updated and backed up. The critical discipline is consistency: every container gets a unique accession number, and every accession number links to a record.
Regeneration Before It Is Too Late
Seeds do not last forever, even under ideal conditions. At some point, viability declines enough that you need to grow out a portion of your stored seeds, harvest fresh seed, and return them to storage. This cycle is called regeneration, and it is one of the most overlooked aspects of seed banking.
Timing matters. If you wait until viability has dropped very low, you may not have enough viable seeds left to produce a genetically representative new generation. Modeling work has shown a significant interaction between when you regenerate and how many seeds you plant: waiting too long means the surviving seeds are a biased sample of the original collection, which erodes genetic diversity with each regeneration cycle.15Crop Science. Modeling Demographics and Genetic Diversity in Ex Situ Collections during Seed Storage and Regeneration The general rule is to regenerate when germination rates drop to around 75 to 85 percent. For highly heterogeneous collections, such as landraces or wild-collected lots, it is better to err on the side of regenerating early and planting more individuals to preserve the range of genetic variation.
Cross-pollinating crops present special challenges during regeneration. If you grow out your corn collection next to a neighbor’s field, pollen will flow in and contaminate your genetics. Isolation, whether by distance, timing, or physical barriers like bags or cages, is essential for maintaining the integrity of each accession.
Dealing With Recalcitrant Seeds
If your interest extends to tree species or tropical plants with seeds that cannot tolerate drying, you face a much harder storage problem. Standard seed banking does not work for these species. The main alternative being explored is cryopreservation: excising the tiny embryonic axis from the seed, partially drying it (but not too much), and plunging it directly into liquid nitrogen at -196°C.
The science here is still challenging. Work on red oak seeds illustrates the difficulty. Under slow cooling, cellular water freezes and the resulting dehydration kills the tissue. Under rapid cooling with liquid nitrogen, vitrification can be achieved if water content is above a certain threshold, but that vitrified state is highly unstable during warming, meaning the tissue can die on the way back out.16Cryobiology. State and Phase Transition Behaviors of Quercus rubra Seed Axes and Cotyledonary Tissues: Relevance to the Desiccation Sensitivity and Cryopreservation of Recalcitrant Seeds For oak species more broadly, researchers have found that the climate a species evolved in affects how its embryonic axes respond to both desiccation and cryopreservation, adding another layer of complexity.17PubMed Central. Climate drives patterns of response of recalcitrant embryonic axes of Quercus species to cryopreservation
Controlled rehydration after cryopreservation is another critical step. Seeds that are high in lipid content or retain endosperm tissue may need slow, staged imbibition rather than being plunged straight into water to avoid cellular damage.18PubMed. Cryopreservation of desiccation-tolerant seeds For a home seed banker, cryopreservation of recalcitrant seeds is likely not practical. If you need to preserve recalcitrant species, the realistic options are maintaining living collections (growing the plants themselves) or connecting with a botanic garden or research institution that has the specialized equipment.
What Happens to the Microbiome Inside Stored Seeds
Seeds are not sterile. They carry bacterial communities on their surfaces and sometimes inside their tissues, and some of those bacteria are beneficial, helping with nutrient uptake and disease resistance after germination. Seed banking disrupts these microbial communities in ways that researchers are only beginning to appreciate.
A study on soybean seeds found that the initial drying step itself caused major shifts in the bacterial community. One dominant genus dropped from over 50 percent of the community to less than 1 percent after drying, while another genus surged. Subsequent cold storage at -20°C preserved the remaining community better than storage at 4°C or room temperature, though low-abundance taxa still gradually disappeared even at the coldest temperature.19PubMed Central. Implications of Seed Vault Storage Strategies for Conservation of Seed Bacterial Microbiomes Research on a native Australian shrub told a similar story: seeds from banking facilities hosted significantly less diverse bacterial communities than freshly collected seeds, and traits associated with plant growth promotion, such as nutrient cycling and disease suppression, were less prevalent in the stored seeds.20PubMed Central. Seed banking impacts native Acacia ulicifolia seed microbiome composition and function
For most home gardeners, this is not a deal-breaker. The seeds still germinate, and soil microbes colonize seedlings quickly. But for conservation projects focused on restoring wild plant communities, the loss of co-evolved microbial partners could matter. One recommendation from the soybean study is to isolate and separately preserve microbial cultures from seeds before the drying step, so those organisms are available for reintroduction later.
Safety Duplication and Backup Strategies
Any single storage location is vulnerable. A freezer failure, a natural disaster, or simple human error can wipe out years of collecting work. Professional genebanks address this by duplicating their collections at separate sites. The Svalbard Global Seed Vault in Norway, which reached a milestone of one million samples stored, exists precisely for this purpose: it holds backup copies of seeds from genebanks around the world so that no single catastrophe can erase an entire collection.21PubMed Central. The Svalbard Global Seed Vault: 10 Years-1 Million Samples
You can apply the same principle at a smaller scale. Split your collection into two portions and store them in different locations: one in your own freezer and another in a friend’s freezer, a community space, or a separate building. Even keeping a backup jar of your most important seeds in a different room from your main collection provides some protection. The effort of dividing your collection in half during the initial sealing process is trivial compared with the effort of re-collecting everything from scratch.
Starting Simple and Scaling Up
If you are reading this and feeling overwhelmed by the technical details, here is a stripped-down version of a working home seed bank. Harvest seeds when they are fully mature and dry on the plant. Finish drying them indoors with silica gel for a week or two. Seal them in labeled glass jars or Mylar bags with a desiccant packet and an oxygen absorber. Put the containers in a chest freezer. Test a small sample every few years. Grow out and replace seeds when germination falls below about 80 percent. Store a duplicate set somewhere else.
That process, done consistently, will keep most common garden and crop seeds viable for a decade or longer. From there, if your interest grows, you can refine your drying protocols, adopt tetrazolium testing for faster viability checks, build a proper accession database, and expand into more challenging species. The communities and institutions that maintain the world’s largest seed collections started with the same basic principles. The scale is different, but the science is the same.