Genome banking and biodiversity preservation are civilisation-continuity infrastructure. They preserve the biological information required to restore crops, livestock, keystone species, ecosystem functions, and potentially extinct lineages—but they do not preserve living ecosystems by themselves. A viable Lunar Ark must therefore maintain seeds, living cultures, cryopreserved cells, reproductive material, genomic sequence, ecological records, and the technical knowledge to use them.
1. Scale of the biological emergency
The IUCN’s 2025 assessment listed 169,420 evaluated species, of which 47,187—28%—were threatened with extinction.[1][3] This is not a complete planetary census: extinction risk has been assessed for less than 5% of described species, so the true number at risk cannot be calculated precisely.
The risk is unevenly distributed:
- Amphibians: approximately 41% threatened.
- Reef-building corals: approximately 44% threatened.
- Trees: approximately 38% threatened.
- Sharks and rays: approximately 38% threatened.
- Mammals: approximately 26% threatened.
- Freshwater fishes: approximately 26% threatened.
- Birds: approximately 11.5% threatened.
The figures change as more species are assessed; an apparent increase can reflect improved assessment rather than a sudden deterioration.[7] The strategic implication is nevertheless clear: a genome archive should not be limited to species already classified as endangered. It must sample taxonomic breadth, genetic diversity, ecosystem roles, crop wild relatives, pollinators, soil organisms, decomposers, pathogens, and microbiomes.
Known biodiversity represents only part of the archive problem. Many fungi, invertebrates, microbes, and deep-ocean organisms remain poorly described. A collection containing only familiar vertebrates would preserve a catalogue, not a functioning biosphere.
2. What a genome bank preserves
A genome bank stores biological material or digital genomic information for future study, breeding, conservation, or reconstruction.
### Physical material
Priority materials include:
- Cryopreserved sperm, eggs, embryos, ovarian and testicular tissue.
- Somatic cells capable of nuclear transfer or cellular reprogramming.
- Seed, pollen, spores, and dormant propagules.
- DNA and RNA extracts.
- Living microbial cultures.
- Tissue samples containing host-associated microbiomes.
- Voucher specimens linked to precise location, date, phenotype, and ecological data.
For animals, viable cells and reproductive material are more valuable than sequence alone. A digital genome records nucleotide order but does not automatically preserve epigenetic state, chromosome architecture, developmental timing, mitochondrial inheritance, maternal effects, symbiotic microbes, or the ability to reproduce.
For plants and fungi, seeds, spores, meristems, and living cultures may preserve more usable biological information than isolated DNA. Some species are “recalcitrant”: their seeds cannot tolerate ordinary drying and freezing, requiring cryopreservation of embryos, shoot tips, pollen, or tissue culture.
### Digital genomic archives
Sequencing should produce multiple independent copies using different technologies and include:
- Nuclear, mitochondrial, and chloroplast genomes.
- Long-read assemblies and chromosome-level scaffolds.
- Population samples rather than one reference genome.
- Structural variants, copy-number variants, and sex chromosomes.
- Phenotypic, developmental, reproductive, and ecological metadata.
- Provenance, permits, chain of custody, and taxonomic authority.
- Open, documented formats with error-correcting redundancy.
Digital sequence is essential for future synthesis, but it is not equivalent to a living specimen. A genome is an instruction set embedded in a biological system; much of that system is absent from a sequence file.
3. Seed bank versus genome bank
| Capability | Seed bank | Genome bank |
|---|---|---|
| Primary material | Viable seeds, usually dried and frozen | DNA, cells, sperm, eggs, embryos, tissues, cultures, or digital sequences |
| Best suited to | Crops, wild plants, many annuals and orthodox-seeded species | Animals, endangered plants, microbes, fungi, crop wild relatives, and species without storable seeds |
| Immediate regeneration | Often possible by germination | Usually impossible from DNA alone; requires cells, gametes, embryos, or synthesis |
| Preserves genotype | Yes, if seed remains viable | Yes, but the usefulness depends on material type |
| Preserves developmental and maternal state | Partly, through the seed | Usually not in extracted DNA |
| Main limitation | Many plants have non-storable seeds; seed banks do not preserve animals or microbiomes | Cryogenic dependence, high technical complexity, and inability to recreate organisms from sequence alone |
| Operational role | Near-term agricultural and ecological restoration | Long-term genetic insurance, research, breeding, and reconstruction |
The Svalbard Global Seed Vault is an example of distributed agricultural insurance, but a seed vault cannot substitute for a genome bank. Seed collections often represent cultivars rather than full wild populations, and a single accession may contain limited genetic variation. Conversely, a genome bank cannot replace a seed bank because growing intact plants from isolated DNA remains technically infeasible for most species.
A civilisation-continuity repository should use a layered architecture:
1. Living conservation populations in protected habitats.
2. Seed, spore, and vegetative-propagule banks.
3. Cryobanks of cells and reproductive material.
4. DNA and RNA repositories.
5. Digitised genomes and phenotypes.
6. Protocols, equipment, and trained operators for recovery.
4. Storage media and longevity
No storage medium is immortal. The relevant quantity is not merely theoretical data retention, but retention under controlled conditions, resistance to radiation and moisture, error correction, readability after technological discontinuity, and the survival of decoding instructions.
### Silicon chips and semiconductor storage
Silicon-based flash memory, solid-state drives, and memory chips are compact and practical but are poor millennial archives when powered off or exposed to heat. Charge leakage and physical degradation can make unpowered data unreliable over years to decades, depending on device type and conditions. NAND flash is therefore appropriate for working copies and transport, not as the sole thousand-year repository.
Archival silicon systems can improve survival through:
- Multiple independently manufactured copies