Civilisation continuity requires two archives: living biodiversity reservoirs and information reservoirs. The current extinction crisis is already large enough that “later” is not a safe plan: the IUCN reports more than 47,000 species threatened with extinction[1], and a 2026 IUCN-led soil study found at least 1,758 soil-dependent species at risk, with another 1,722 data-deficient among 8,500+ assessed species[2].
1) How many species are at risk
- IUCN Red List baseline: >47,000 threatened species globally[1].
- Another 2026 update: a separate summary reported 49,505 species in the threatened categories CR/EN/VU, reflecting continued reassessments and additions[4].
- Species loss is ongoing: a 2026 extinction tracker reported 15 more animal species declared extinct in the latest IUCN update, and 175,909 assessed species total[3].
For planning purposes, the conservative working assumption is that tens of thousands of species already need ex situ preservation, and that the number will keep rising.
2) Seed bank vs genome bank
- Seed bank: stores viable seeds, usually cold and dry, to preserve a plant species’ ability to be regrown later.
- Genome bank: stores genetic material, often DNA, tissues, gametes, embryos, spores, cell lines, or cryopreserved living cells, for future research, restoration, or possible reconstruction.
The key difference is function:
- Seed banks preserve living reproductive units.
- Genome banks preserve genetic information and cellular material, including species that do not make storable seeds, or where seeds are unavailable, short-lived, or too genetically narrow.
A seed bank is enough for many plants. A genome bank is required for animals, fungi, microbes, recalcitrant-seeded plants, and species whose reproductive biology blocks simple seed storage.
3) DNA storage media: longevity and failure modes
### Silicon chips
- Strength: high density, rapid random access, room-temperature operation.
- Weakness: electronics fail; the data layer is only as durable as the device and format ecosystem around it.
- Civilisation-grade risk: bit rot is not the main threat; hardware obsolescence, power dependence, and interface loss are.
### Synthetic DNA
- Strength: extremely high density; encodes large datasets in tiny mass.
- Weakness: synthesis and sequencing remain slow and expensive compared with digital media; reading requires specialised equipment.
- Longevity: dry, cold, sealed DNA can remain legible for very long periods, but practical archival survival depends on encapsulation, temperature, moisture exclusion, and periodic validation.
### Crystal storage
- Strength: designed for extreme stability and resistance to heat, radiation, and time.
- Weakness: still emerging; read/write infrastructure is specialised and not yet a mature archival ecosystem.
- Best use: deep archive for “last-copy” civilization records, not high-throughput daily storage.
### Bottom line
For a lunar archive, the optimal strategy is layered redundancy:
- fast access: silicon
- medium-term dense archive: synthetic DNA
- ultra-long-term vault: crystalline or similarly robust physical media
- always: multiple copies in geographically separated vaults
4) Frozen Ark: the flagship animal genome-rescue project
The Frozen Ark was launched by scientists Bryan Clarke and Ann Clarke to preserve DNA and cells from threatened animal species before extinction. A 2015 report described the network as 22 partners across zoos, research centres, and universities, holding 48,000 samples from about 5,500 species.
Operational significance:
- It is one of the clearest proofs that genetic banking must happen before extinction.
- It preserves material for future research and recovery, not just symbolic samples.
- It is still modest relative to the scale of global biodiversity loss; even a few tens of thousands of samples cover only a small fraction of at-risk animal diversity.
5) De-extinction feasibility
De-extinction is possible in narrow cases, not as a general restoration technology.
Feasible routes:
- Back-breeding / trait recovery: works only when closely related descendants still exist and the target traits remain in the gene pool.
- Genome editing: plausible when a close living relative is available and the extinct genome is sufficiently reconstructed.
- Cloning: requires intact or near-intact cells; this is the most restrictive path.
Hard limits:
- No living cells = no straightforward clone
- Poor genome completeness = high uncertainty
- Missing epigenetics, development, microbiome, and habitat = incomplete restoration
- Ecological replacement is not resurrection; even a genetically similar organism may not restore lost function
Civilisation-grade conclusion: de-extinction should be treated as experimental triage, not a substitute for preserving species now.
6) Minimum viable genetic diversity
A preserved species is not safe if it has only one genome. Long-term survival requires enough diversity to avoid inbreeding depression and to retain adaptive potential.
Practical conservation benchmarks:
- Short-term demographic survival: roughly 50 effective breeders is a common floor for immediate inbreeding control.
- Long-term evolutionary viability: roughly 500 effective breeders has been the classic target; many modern conservation programs aim higher because climate and disease pressures are stronger than historic assumptions.
- Ex situ planning reality: for genetically healthy recovery, banking should aim to capture multiple unrelated individuals across the geographic range, not just a few representatives.
Rule of thumb for restoration banks:
- Do not bank a species with one line.
- Preserve many individuals, many lineages, both sexes, and multiple populations.
- Include rare alleles and regional adaptations, not only “typical” genotypes.
7) Civilisation-grade recommendations
- Bank living cells and reproductive material before species collapse, not after.
- Prioritise taxa with high ecological leverage: crops, pollinators, reef builders, apex predators, keystone plants, soil biota.
- Use **redundant storage media