Humanity is already in a high-loss regime: the latest IUCN update reports 175,909 assessed species, with 49,505 threatened with extinction; that is the current operational baseline for any preservation programme.[1] A separate IUCN summary still frames the global crisis as more than 47,000 species threatened, showing that the exact count shifts with each update but the order of magnitude remains the same: tens of thousands of species are already in immediate danger.[2]
1) How many species are at risk
- 49,505 threatened species on the 2026 IUCN Red List update.[1]
- That includes species across major habitat systems; one 2026 update reported 125 of 201 endemic hydrothermal vent molluscs at risk from deep-sea mining, or 62% of that group.[1]
- Another 2026 assessment reported 42% of 558 native European freshwater fish species threatened, plus 18% Near Threatened, meaning 60% were already under elevated conservation concern.[5]
- A 2026 soil-biodiversity report found 20% of assessed soil species were at risk, and a further 20% were Data Deficient, meaning extinction risk may be badly underestimated.[4]
2) Seed bank vs genome bank
- A seed bank preserves the whole propagule of plant life: seeds, typically dried and stored cold for long-term viability.
- A genome bank preserves genetic material, not necessarily whole organisms: DNA, cells, tissues, gametes, embryos, cryopreserved sperm/ova, and sometimes living cell lines.
- Seed banks are optimized for plants that make orthodox seeds; they do not cover animals, fungi, microbes, or plants with recalcitrant seeds well.
- Genome banks are broader and are the only practical preservation layer for animals and many non-seed-forming lineages.
- For civilisation continuity, the key difference is this: seed banks preserve future agriculture; genome banks preserve future reconstruction capacity for the rest of life.
3) DNA storage media longevity
### Silicon chips
- Silicon-based DNA storage is attractive for density and digital readability.
- The long-term weakness is not the silicon itself but the chemical stability of the payload and the reading ecosystem: future access requires compatible synthesis, indexing, and sequencing infrastructure.
- In practice, silicon is best treated as an information carrier, not a biological preservation medium.
### Synthetic DNA
- Synthetic DNA is the most information-dense archival medium in active research.
- Its decisive advantage is extreme compression: enormous data volumes can be encoded in tiny mass.
- Its decisive weakness is chemical decay: DNA hydrolyzes and fragments over time unless carefully dried, cooled, and protected.
- Synthetic DNA is viable for centuries to millennia only with stringent storage, redundancy, and periodic error-checking or re-encoding.
### Crystal storage
- Crystal-based storage is promising because crystalline lattices can protect information with very low molecular motion.
- The concept is strongest for very long-term archival retention because crystals can, in principle, resist heat, radiation, and diffusion better than liquid or biological media.
- The practical limitation is engineering maturity: compared with seed or cell cryobanks, crystal storage remains less operationally proven at planetary scale.
4) Frozen Ark project
The Frozen Ark is a real-world genetic ark: a network designed to collect and preserve DNA and cells from endangered species before extinction. The project was launched by British scientists Bryan Clarke and Ann Clarke, and the network reported 22 partners and 48,000 samples from about 5,500 species in the cited account. Its mission is directly relevant to civilisation backup: it is a prototype for a distributed, species-level genome vault rather than a seed bank.
5) De-extinction feasibility
- Technically feasible in limited cases, not generally.
- De-extinction is most plausible where:
- high-quality DNA exists,
- close living relatives exist,
- reproductive technology is available,
- and the extinct species’ ecological niche still exists.
- It is least plausible when:
- the genome is fragmented,
- epigenetic and developmental context is lost,
- behavior is not reconstructible,
- or the habitat no longer exists.
- The real constraint is not only genome reconstruction; it is developmental biology, surrogate gestation, microbiome transfer, and ecological replacement.
- For continuity planning, de-extinction is a secondary restoration tool, not a substitute for preservation. Saving genomes now is cheaper and far more reliable than trying to resurrect extinct taxa later.
6) Minimum viable genetic diversity
Two thresholds matter.
- The older conservation heuristic is the 50/500 rule:
- 50 effective breeders to limit short-term inbreeding depression,
- 500 effective breeders to preserve long-term evolutionary potential.
- More recent conservation genetics guidance pushes the long-term target higher:
- effective population size \(N_e\) of at least 1,000 is recommended to maintain evolutionary potential in naturally outbreeding species.
Operationally, that means:
- A species archive should aim to preserve multiple unrelated lines, not one representative genome.
- For many species, a single genome is informationally useful but biologically insufficient.
- The archive target should be enough sampled diversity to reconstruct at least 100s to 1,000s of distinct effective genetic lineages, depending on the species’ mating system and inbreeding risk.
7) Civilisation-grade priorities
- Preserve seed banks for crops and wild relatives.
- Preserve genome banks for animals, fungi, microbes, and non-seed plants.
- Preserve living cell lines, sperm, ova, embryos, and tissues where possible.
- Preserve metadata with every sample: location, date, phenotype, pedigree, pathogen status, and storage history.
- Preserve redundantly across multiple sites; one vault failure must not be terminal.
- Preserve threatened keystone species first, then phylogenetically unique lineages, then ecologically