Recoverable Foundation

Recovery Library

Doc #85 — Animal Breeding and Genetic Diversity

Managing Livestock Genetics Under Herd Reduction, Nuclear Winter, and Long-Term Isolation

Phase: 2–4 (Critical Phase 2–3; relevant through Phase 5) | Feasibility: [A] Established (NZ has existing AI infrastructure and breeding expertise)

Unreliable — not for operational use. Produced by AI under human direction and editorial review. This document contains errors of fact, judgment, and emphasis and has not been peer-reviewed. See About the Recovery Library for methodology and limitations. © 2026 Recoverable Foundation. Licensed under CC BY-ND 4.0. This disclaimer must be included in any reproduction or redistribution.

EXECUTIVE SUMMARY

When New Zealand destocks 30–60% of its national herd under nuclear winter (Doc #3), the surviving animals become the genetic foundation for all future NZ livestock. Every culling decision is also a breeding decision. If destocking selects purely for short-term productivity — keeping only the highest-producing dairy cows, the fastest-growing lambs — NZ risks a genetic bottleneck that narrows the breeding population to animals optimised for conditions that no longer exist.

Genetic diversity is the raw material of adaptation. A genetically diverse herd can be selectively bred toward cold tolerance, disease resistance, feed efficiency under reduced pasture, or draft capability. A genetically narrow herd cannot — it has lost the alleles needed to shift direction. The loss is irreversible without importation of new genetics, which under isolation may not be possible for years or decades.

NZ enters this scenario with significant genetic infrastructure. Livestock Improvement Corporation (LIC), the country’s dominant dairy genetics provider, maintains bull teams and semen storage facilities holding millions of frozen semen straws at its Newstead facility near Hamilton.1 CRV Ambreed provides additional dairy genetics.2 The sheep and beef sector relies more heavily on natural mating, but frozen semen and some embryo storage exist through commercial studs and research institutions. The Rare Breeds Conservation Society of New Zealand (RBCS) maintains registries and breeding programmes for heritage breeds whose small populations are already at genetic risk.3

The core challenge is threefold: (1) ensuring destocking decisions preserve genetic breadth, not just current productivity; (2) maintaining the artificial insemination (AI) infrastructure that stores irreplaceable frozen genetics; and (3) managing breeding in reduced herds to minimise inbreeding over the decades of isolation ahead.

Key honest uncertainties: The actual severity and duration of nuclear winter determines how aggressively NZ must destock, and therefore how narrow the genetic bottleneck becomes. The viability of frozen semen stores depends on continued liquid nitrogen supply, which depends on industrial gas infrastructure that may or may not continue operating. The effective population size required to avoid harmful inbreeding accumulation over multiple generations is well-established in theory but depends on assumptions about generation interval, mating management, and population structure that cannot be precisely predicted.

Contents

First month:

  1. Incorporate genetic diversity criteria into destocking guidance (Doc #3). Issue supplementary instructions to farmers: do not cull solely on current production. Retain animals representing different breed types, bloodlines, and sire groups. Specific guidance in Section 4.
  2. Secure LIC Newstead and CRV Ambreed facilities. Designate semen and embryo storage sites as protected national infrastructure. Confirm liquid nitrogen supply chain and backup arrangements. These facilities hold genetics that cannot be recreated.
  3. Secure Rare Breeds Conservation Society herds and flocks. Contact RBCS to identify locations of all registered rare breed populations. Issue protection orders preventing their inclusion in general destocking.
  4. Begin national sire diversity audit. Identify which bulls and rams have sired the surviving herds. This determines effective genetic diversity — if 80% of surviving cows were sired by five bulls, diversity is already critically low regardless of cow numbers.

First season:

  1. Establish regional breed nuclei. In each major farming region, designate core breeding herds/flocks that represent maximum genetic breadth. These animals are not available for slaughter under any circumstances short of starvation.
  2. Resume AI services for dairy and begin extending AI to beef and sheep. LIC technicians should continue operating. Expand AI use as a tool for managing genetic diversity — a single stored semen straw represents genetic insurance that a living bull does not (a bull can die; a frozen straw persists).
  3. Begin structured mating plans for reduced herds. Distribute guidance on avoiding close inbreeding in small herds: rotate sires between farms, use AI where available, maintain mating records.

First year:

  1. Establish a national livestock genetic register. Record breed, sire, dam (where known), and location for all breeding animals in reduced herds. This register is the basis for all future breeding coordination. Marae-based community structures — which already coordinate shared grazing, cooperative breeding decisions, and collective record-keeping — should be integrated into this framework as regional coordination units rather than treated as separate from the national programme.
  2. Conduct liquid nitrogen audit. Determine total national LN2 supply, production capacity (air separation units), and projected consumption. Establish rationing if supply is constrained.
  3. Begin training additional AI technicians. Current LIC and CRV technicians represent a critical skill bottleneck. Train additional technicians through veterinary and farming networks.
  4. Evaluate draft horse and draught ox breeding priorities. Identify existing heavy horse breeds (Clydesdales, Percherons, draft crosses) and begin a coordinated breeding programme. This is a 5–10 year project before useful numbers of trained draft animals are available.

Years 2–5:

  1. Implement managed breeding rotations across districts. Coordinate sire movement between farms to maximise effective population size. AI from stored semen extends the genetic contribution of bulls no longer alive.
  2. Begin selection for nuclear winter adaptation. Animals that thrive under reduced pasture, colder conditions, and lower-input management are the breeding stock for the future. Selection can begin as early as the second season, once performance data under new conditions exists.
  3. Monitor inbreeding coefficients. As pedigree data accumulates in the national register, calculate inbreeding coefficients and flag matings that would produce offspring above threshold levels.

ECONOMIC JUSTIFICATION

Cost of action

Maintaining genetic diversity during destocking requires modest additional effort: supplementary guidance to farmers, coordination of sire records, protection of AI facilities, and limited additional labour for record-keeping and AI services.

  • National livestock genetic register: 2–5 person-years to establish; 1–2 person-years annually to maintain. Primarily data collection and coordination, leveraging existing LIC/CRV databases and farm records.4
  • AI service continuation: LIC employs approximately 300–400 staff nationally, of whom perhaps 100–150 are field AI technicians during the breeding season.5 Maintaining this service for a reduced herd requires proportionally fewer technicians — perhaps 30–60. These skills take months to years to train and are not widely held outside LIC and CRV, making retention of existing technicians the priority rather than rapid retraining.
  • Rare breed protection: Minimal cost. The breeds already exist on farms; the requirement is preventing their slaughter during destocking. Some supplementary feeding support may be needed.
  • Liquid nitrogen supply: This is the significant ongoing cost. Air separation units that produce LN2 require grid electricity and maintenance. Securing this supply is an industrial infrastructure question (see Section 6).

Cost of inaction

The cost of neglecting genetic diversity is delayed and severe. Inbreeding depression — reduced fertility, smaller body size, increased disease susceptibility, higher calf and lamb mortality — appears gradually, typically becoming measurable within 3–5 generations of unmanaged breeding in small populations.6 In cattle (generation interval ~4–5 years), this means noticeable effects within 15–25 years. In sheep (generation interval ~3–4 years), within 10–15 years.

At that point, the damage is done and cannot be reversed without importing new genetics. If NZ is still isolated, there is no remedy.

Comparative cost: Iceland’s cattle population, isolated since the 10th century, survived but with significant inbreeding — modern Icelandic cattle have measurably reduced genetic diversity compared to their Scandinavian ancestors, and this limits breeding flexibility.7 NZ’s starting population is larger, but the bottleneck could be sharper if destocking is severe and unmanaged.

Breakeven: The investment in genetic management — perhaps 5–10 person-years in the first year, declining to 2–5 person-years annually — pays for itself if it prevents even a 1–2% reduction in national herd productivity over the following decades. Given that NZ’s reduced herd will be producing food for the entire country, a 1–2% productivity loss represents on the order of 5,000–15,000 tonnes of meat and milk equivalent annually — calculated from a post-destocking herd producing perhaps 50–60% of pre-war output, with pre-war primary livestock production (meat plus dairy) on the order of 3.5–4.5 million tonnes per year.8

4. THE GENETIC BOTTLENECK PROBLEM

4.1 What a bottleneck is

A genetic bottleneck occurs when a population is sharply reduced, and the surviving individuals carry only a fraction of the genetic diversity present in the original population. The smaller the surviving population and the more closely related the survivors, the more severe the bottleneck.

NZ’s livestock face a bottleneck from two directions simultaneously:

  • Herd size reduction: Destocking from ~83 million stock units to 35–58 million (Doc #74) removes a large proportion of the breeding population.9
  • Selection bias during culling: If culling removes animals non-randomly — favouring specific breeds, bloodlines, or production traits — the surviving population is less diverse than its numbers alone would suggest.

4.2 Effective population size

The key concept is effective population size (Ne), which measures how many unrelated individuals contribute genetics to the next generation. Ne is always smaller than the census population because:

  • Not all animals breed.
  • Some sires contribute disproportionately (one bull may sire hundreds or thousands of calves through AI).
  • Females and males contribute unequally when sex ratios in the breeding population are skewed.

A common approximation: Ne = (4 × Nm × Nf) / (Nm + Nf), where Nm = number of breeding males, Nf = number of breeding females.10 When one sex is much rarer (as bulls are relative to cows), Ne is dominated by the rarer sex. If 50 bulls mate with 10,000 cows, Ne is approximately 198 — not 10,050.

The implication for NZ: Maintaining large cow numbers is necessary but not sufficient. Genetic diversity depends critically on maintaining enough genetically distinct sires — whether as living bulls, frozen semen, or both.

4.3 Minimum viable population for livestock

Conservation genetics suggests that an effective population size of at least 50 is needed to avoid severe short-term inbreeding depression, and at least 500 is needed to maintain adaptive potential over the long term (the “50/500 rule”).11 These are rough guidelines, not precise thresholds.

For NZ’s major livestock species, meeting the Ne = 500 threshold requires:

  • Dairy cattle: At least 50–100 genetically distinct sires contributing to each generation, across at least 5,000–10,000 breeding cows. Given NZ’s post-destocking dairy herd of perhaps 2–4 million cows, the cow numbers are not the constraint — sire diversity is.12 If NZ’s reduced dairy herd is bred primarily through AI from a small number of popular sires, effective population size could be far below 500 even with millions of cows.
  • Sheep: The sheep sector uses natural mating far more than AI. With perhaps 10–15 million breeding ewes post-destocking, sire numbers are less concentrated — but the same risk applies if destocking removes rams of minority breeds and bloodlines.13
  • Beef cattle: Predominantly natural mating. Post-destocking beef herd of perhaps 1.5–3 million, with smaller herds per farm. Maintaining sire diversity requires coordination between farms.14
  • Deer: Post-destocking herd of perhaps 300,000–600,000. The smallest of NZ’s major livestock populations, and the most recent in its breeding history (NZ deer farming began in the 1970s from a relatively narrow genetic base of captured wild deer plus limited imports).15 Genetic diversity in deer is already a concern.16

4.4 The sire concentration problem in dairy

NZ’s dairy genetics are highly concentrated. LIC’s national breeding programme evaluates bulls through progeny testing and genomic selection, and the top-ranked bulls are used very widely through AI. In a typical year, the top 20–30 bulls may sire 30–50% of all AI-bred calves.17 This is efficient for genetic gain under normal conditions but creates a vulnerability: if semen stores are lost or the gene pool is narrowed to only the most popular sires, genetic diversity in the dairy herd contracts rapidly.

The stored semen at LIC’s Newstead facility is the critical counterweight. It contains straws from hundreds of bulls across multiple years, including older sires whose genetics may not be commercially popular but represent valuable genetic breadth. This archive is, in genetic terms, more valuable than any individual living bull.

5. BREEDING STRATEGIES FOR REDUCED HERDS

5.1 Destocking with diversity in mind

Farmers making culling decisions need guidance that balances immediate survival (keep the best-performing animals) with long-term resilience (keep a diverse genetic base). Practical guidelines:

For dairy: - Retain animals from at least 10–15 different sire lines within each herd, where possible. Farm records and LIC herd-testing data identify sire groups. - Do not cull all animals of a minority breed. If a herd contains mostly Holstein-Friesian (KiwiCross) cows but a few Jerseys or Ayrshires, keep some of each. - Retain some older cows of known different parentage, even if their current production is below average. Their genetics may be irreplaceable. - Where two cows are similarly productive, keep the one from the rarer sire group.

For sheep: - NZ’s sheep breeds include Romney (dominant), Coopworth, Perendale, Corriedale, Merino, Texel, Suffolk, and various composites. Each breed represents a distinct genetic resource adapted to different conditions. - Merino and half-bred flocks produce finer wool — critical for textile recovery (Doc #104). Do not eliminate these in favour of meat breeds. - Hill-country breeds (Romney, Perendale) are adapted to harsh conditions that may be more prevalent under nuclear winter. Retain these even if lowland breeds are currently more productive. - Retain rams from multiple bloodlines. A minimum of 3–5 unrelated rams per flock of 500+ ewes.

For beef: - Angus and Hereford dominate NZ beef genetics. Retain animals of both breeds plus any minority breeds (Murray Grey, Shorthorn, Simmental, Charolais crosses). - Dual-purpose breeds (Shorthorn, some older Hereford strains) that produce both reasonable beef and some milking ability may become more valuable under recovery conditions.

For deer: - The NZ deer herd is predominantly Red Deer (Cervus elaphus) with some Wapiti (Cervus canadensis) genetics, mainly in Fiordland-origin herds.18 Maintain both genetic backgrounds.

5.2 Using artificial insemination strategically

AI is NZ’s most powerful tool for managing genetic diversity in reduced herds. A single bull’s stored semen — potentially hundreds to thousands of straws — can be distributed across the country, contributing genetics to herds with no other access to diverse sires. Frozen straws are not subject to accident, disease, or death the way living bulls are. And when a small herd has been using the same bull for several years, AI from an unrelated sire breaks the inbreeding cycle without needing to transport a live animal.

Limitations under recovery conditions: AI requires a full consumables chain beyond liquid nitrogen: insemination catheters and sheaths (disposable, typically imported plastics), semen extenders and cryoprotectants (glycerol, egg yolk, antibiotics — all requiring ongoing supply), and oestrus synchronisation hormones (CIDR devices, prostaglandins — manufactured overseas). NZ does not currently manufacture these consumables domestically. Without importation, existing stocks will deplete within 2–5 years of routine use, and the AI programme must shift to working from existing frozen semen without adding new collections.19 AI is also rarely used in NZ sheep; laparoscopic AI requires specialised equipment and veterinary skill.20 Natural mating with managed ram rotation is more practical for sheep. And AI does not solve the diversity problem if all stored semen comes from a few popular sires — the value depends on the breadth of the archive.

Performance gap — AI versus natural mating: In NZ dairy, synchronised AI achieves 6-week in-calf rates of approximately 70–78%; natural mating in well-managed herds achieves similar rates. The advantage of AI is not conception rate but genetic reach — one bull’s semen can serve thousands of cows across many farms, creating diversity that natural mating cannot replicate at scale. If consumables deplete and AI ceases, genetic diversity management shifts entirely to physical sire movement between farms, which is slower, logistically demanding, and creates biosecurity risks (disease transmission between herds).21

5.3 Natural mating management

For species where AI is impractical, genetic diversity depends on: rotating sires between farms every 2–3 years (coordinated through the national livestock genetic register); dividing larger flocks into mating groups each served by a different sire; and maintaining basic records of which sire was used, when, and where. NZ’s dairy sector already keeps detailed records through LIC’s MINDA database (Management Information for New Zealand Dairy Animals), which records parentage, production, and health data for most NZ dairy herds.22 The sheep and beef sectors are less systematic — Beef + Lamb NZ’s SIL (Sheep Improvement Limited) database covers stud flocks and some commercial flocks, but the majority of NZ’s 60,000+ sheep farms operate without formal genetic records. Extending even basic sire-identification record-keeping to these farms is a practical priority.

5.4 Shifting selection goals

Pre-war NZ livestock breeding optimised for traits relevant to an export economy: milk solids per cow (dairy), growth rate and carcass quality (beef, lamb), and wool weight and quality (sheep). Under recovery conditions, selection priorities shift. The performance gap is real: selecting for hardiness and stress tolerance typically reduces peak productivity by 10–30% compared to animals selected for maximum output under optimal conditions.23 A hardy cow producing 350 kg milk solids/year under nuclear winter conditions is more valuable than a high-genetic-merit cow producing 500 kg/year under normal conditions but failing to conceive or maintain body condition under feed stress.

Higher priority under nuclear winter: - Feed efficiency under reduced pasture. Animals that maintain body condition on less feed are more valuable than high-producing animals that require high feed inputs. - Cold tolerance. Animals that thrive without shelter in temperatures 3–8°C below current norms (depending on nuclear winter severity and regional variation).24 - Disease resistance. With veterinary pharmaceuticals depleting (Doc #124), animals with natural resistance to common NZ livestock diseases (facial eczema, footrot, internal parasites) become more valuable.25 - Fertility under stress. Animals that conceive and rear offspring despite nutritional stress. - Dual-purpose capability. Cattle that provide both milk and beef. Sheep that provide both meat and wool.

Lower priority: - Milk solids yield above domestic consumption requirements. - Carcass grading to export market specifications. - Fine-wool premiums (though fine wool retains some value for specific textile applications — see Doc #104).

This selection shift can begin in the first breeding season under nuclear winter. Animals that perform well under the new conditions should be preferentially retained for breeding. The danger is over-correcting — selecting so aggressively for hardiness that productive potential is lost. Maintain some high-producing genetics (especially in semen storage) even if they are not the best performers under current stress, because conditions will improve as nuclear winter eases.

6. SEMEN STORAGE AND LIQUID NITROGEN

6.1 NZ’s semen storage infrastructure

LIC’s primary facility at Newstead, Hamilton, stores semen in liquid nitrogen dewars at -196°C. The collection holds straws from active and retired bulls — potentially representing hundreds of sires across multiple decades of NZ dairy breeding.26 CRV Ambreed maintains a separate collection.27 Smaller collections exist at Massey University, AgResearch, private stud farms (beef cattle, some sheep), and deer industry semen banks. Some frozen embryos are also stored, in smaller quantities — these represent complete genetics (both sire and dam) and are more valuable per unit but rarer and more fragile.

6.2 The liquid nitrogen dependency

Frozen semen remains viable indefinitely at -196°C — there is no theoretical degradation in liquid nitrogen storage.28 The constraint is not time but LN2 supply. If liquid nitrogen runs out, dewars begin warming; viability declines significantly above approximately -80°C and is rapidly destroyed above -130°C, at which point ice crystal formation in cells causes irreversible damage.29

LN2 is produced by air separation units (ASUs) — industrial facilities that cool air to cryogenic temperatures and separate it into nitrogen, oxygen, and argon. NZ has several ASUs operated by BOC (a Linde company) and Air Liquide, primarily at industrial sites.30

ASU operation requires: - Grid electricity — a medium-scale ASU draws 0.5–2 MW continuously for compressors and refrigeration31 - Maintenance of specialised industrial equipment: compressors (requiring lubricant and seals that may be imported), heat exchangers (corrosion-resistant alloys), molecular sieves (zeolite media with finite life, typically replaced every 3–7 years), and control electronics - Skilled operators — ASU operation and maintenance is a niche industrial skill; NZ likely has only a few dozen qualified operators nationally across BOC and Air Liquide sites

Under the baseline scenario (grid continues, 85%+ renewable), ASU operation is feasible. The question is whether these facilities are prioritised for continued operation when industrial demand for their other products (medical oxygen, industrial gases) is also competing for attention.

Contingency if LN2 supply fails: There is no practical field-level alternative. LN2 requires industrial air separation equipment; it cannot be produced without that infrastructure. If ASU operation ceases and stored LN2 is exhausted, all frozen semen and embryos are lost. This failure mode is worth prioritising specifically because it is silent and rapid: a dewar with a faulty vacuum seal can warm from -196°C to -80°C — the point at which viability begins to decline significantly — within days, with no visible warning.32

Mitigation: Maintain at least one ASU in operation specifically for cryogenic storage supply. The LN2 requirement for semen storage alone is modest — perhaps a few hundred litres per week nationally to top up dewars.33 This is a tiny fraction of ASU output. The justification for keeping an ASU running for this purpose alone is strong, though in practice the same facility would also supply medical oxygen and other critical gases.

6.3 Backup and dispersal

The concentration of NZ’s genetic archive at one or two sites (Newstead, plus CRV’s facility) creates a single-point-of-failure risk from fire, earthquake, equipment failure, or social disruption. Dispersal of duplicate samples to 3–5 regional locations — each with its own LN2-cooled dewar — significantly reduces this risk. LIC may already maintain some dispersal; the status should be confirmed immediately post-event and expanded if necessary.

7. RARE AND HERITAGE BREEDS

7.1 Why rare breeds matter disproportionately

NZ’s rare and heritage breeds — maintained by the Rare Breeds Conservation Society and individual breeders — represent genetic diversity that exists nowhere else in the mainstream breeding population. Some examples:34

Cattle: - Ayrshire: Once common in NZ, now rare. Good milk quality, hardier than Holstein-Friesian. - NZ Milking Shorthorn: Dual-purpose (milk and beef). Population critically low. - Devon: Hardy British breed, small NZ population. - Enderby Island cattle: Descended from a feral population isolated on Enderby Island (Auckland Islands) since the 1890s. Possibly the most genetically distinct cattle population in NZ. Numbering in the low hundreds.35

Sheep: - Pitt Island sheep: Descended from Merino sheep marooned on Pitt Island (Chatham Islands). Small, hardy, fine-woolled. Population very small. - Arapawa Island sheep: Feral population from Arapawa Island in the Marlborough Sounds. Fine wool, extremely hardy. - NZ Halfbred: Traditional Merino/English Leicester cross. Once the backbone of NZ wool production, now largely replaced by modern composites.

Goats: - Arapawa Island goats: Feral population, possibly descended from Old English goats brought by early whalers. Genetically distinct.

Horses: - Kaimanawa wild horses: Descended from domestic horses released in the central North Island. Hardy, adapted to harsh NZ conditions. Population managed by DOC at approximately 300–500 animals.36 - Clydesdale and Percheron drafts: Small populations maintained by enthusiast breeders. These are the foundation of any draft animal programme.

Pigs: - Kunekune: NZ’s only indigenous domestic pig breed, associated with Maori farming. Small, hardy, efficient forager. Population recovered from near-extinction in the 1970s to several thousand.37 Kunekune were traditionally raised in low-input, forage-based systems integrated with crop and garden production — a model well-suited to recovery conditions where concentrated feed is unavailable and animals must utilise diverse food sources including food waste, root crops, and rough grazing.

7.2 Protection measures

These breeds cannot withstand the loss of even a few breeding animals. Protection measures:

  • Identify and register all rare breed animals nationally. RBCS registries provide a starting point but may be incomplete.
  • Exempt rare breed animals from destocking. Their feed cost is trivial relative to the national herd; their genetic value is irreplaceable.
  • Establish backup populations. For breeds with fewer than 200 breeding animals, establish at least two geographically separated populations to protect against localised disasters (fire, disease, earthquake).
  • Collect and store semen from rare breed males where AI capability exists. Even a few dozen straws from a rare breed bull or ram represent significant genetic insurance.
  • Integrate Kunekune pigs into recovery food systems. Kunekune are efficient foragers that can utilise food waste, root crops, and rough grazing that other livestock cannot. They are well-suited to smallholder and community food production under recovery conditions.

8. HERD RECOVERY TIMELINE

8.1 Biological constraints on rebuilding

Livestock populations recover slowly because of biological generation intervals:

Species Gestation Typical offspring per birth Age at first breeding Generation interval
Cattle ~9 months 1 (twins rare) 2–3 years 4–5 years
Sheep ~5 months 1–3 1.5–2 years 3–4 years
Deer ~8 months 1 2 years 3–4 years
Goats ~5 months 1–3 1–1.5 years 2–3 years
Horses ~11 months 1 3–4 years 6–8 years
Pigs ~4 months 8–14 8–12 months 1.5–2 years

Cattle are the slowest to rebuild. A cow produces at most one calf per year. Under a moderate destocking scenario (40% reduction, from ~10.2 million to ~6.1 million38), rebuilding to pre-war numbers — even assuming no further losses — requires approximately 8–12 years under optimal conditions, longer if feed remains constrained.39 Under severe destocking (60% reduction, to ~4.1 million), recovery extends to 15–20 years or more, because the breeding population itself is smaller and annual net additions are proportionally lower.

Sheep rebuild faster due to twinning. A flock reduced by 40% can recover within 4–7 years if conditions allow; at 60% reduction, 7–12 years.

Pigs are the fastest livestock to rebuild numerically, with large litters and short generation intervals. A pig breeding programme can double its population annually under good conditions. This makes pigs a valuable early-recovery protein source.

8.2 Phases of herd recovery

Phase 2 (Years 1–3): Survival. Maintain reduced herds at levels the land can support. Focus on keeping animals alive through nuclear winter. Breeding continues but restocking is not yet the goal — excess animals would starve. Genetic management during this phase is about preserving diversity in the breeding population, not expanding numbers.

Phase 3 (Years 3–7): Cautious expansion. As nuclear winter eases and pasture recovers, begin increasing herd sizes. The rate of expansion must be slower than pasture recovery — overstocking during recovery risks the degradation spiral described in Doc #86. Increase cow/ewe numbers by 5–10% per year, monitoring pasture response.40

Phase 4 (Years 7–15): Rebuilding. Pasture approaching normal productivity. Herd numbers expanding toward pre-war levels or toward a new equilibrium based on changed economic conditions (less export, more domestic consumption). Breeding programmes shift from pure survival-trait selection toward balanced improvement.

Phase 5 (Years 15–30): New equilibrium. Herd sizes stabilised. If trade has developed (Doc #151), genetics traded from Australia or further afield can refresh NZ’s gene pool. The genetic register begun in Phase 2 provides a decades-long pedigree database for informed breeding decisions.

8.3 The draft animal question

Draft animal breeding is a 5–10 year programme. Horse biology constrains this: one foal per year, 2–3 years to maturity, 1–2 years of training. Starting from NZ’s small population of heavy breeds (estimated a few hundred Clydesdales and Percherons), building several thousand working draft horses takes 15–25 years.41

Oxen offer a faster alternative. Most cattle breeds can be trained for draft work, though conformation matters: heavier-framed beef breeds (Hereford, Angus, Shorthorn) make better working oxen than high-production dairy breeds, which are bred for body condition under milking stress rather than sustained muscular exertion. Training a responsive working ox takes approximately 6–18 months depending on the animal’s temperament and the handler’s experience.42 Oxen are slower than horses — typical working speed of 2–3 km/h versus 4–6 km/h for draft horses — and less versatile: they cannot be ridden, are poorly suited to road transport over long distances, and fatigue more quickly under sustained exertion. A single draft horse can typically plough 0.4–0.6 hectares per day; a yoke of oxen, 0.2–0.4 hectares.43 Using steers for draft also removes them from the meat supply. But under severe fuel depletion (Doc #53), an ox training programme could provide draft power within 1–2 years while the longer-term horse programme develops.

9. CRITICAL UNCERTAINTIES

Uncertainty Range of outcomes Impact on breeding strategy
Severity of destocking required 30–60% of national herd Determines bottleneck severity. At 30%, diversity is manageable. At 60%, active management is essential.
Duration of nuclear winter 2–5 years at peak severity Longer duration means more generations under stress, more selection pressure, and more opportunity for genetic drift.
Liquid nitrogen supply continuity Full supply to total loss Loss of LN2 destroys the semen archive — the highest-leverage single point of failure for genetic diversity, given that stored straws represent sires no longer alive and cannot be recreated.
Farmer compliance with diversity-aware culling Variable Farmers under feed stress will cull to survive, not to maintain genetic breadth. Guidelines must account for the conditions farmers actually face under feed stress.
Disease outbreaks in reduced herds Unknown probability Crowded, stressed animals are more disease-susceptible. An outbreak in a small population could eliminate an entire breed line.
Trade resumption timeline 5–30+ years for reliable livestock genetics trade Determines when genetic importation becomes possible. Until then, NZ’s gene pool is what it is.
Effectiveness of selection for nuclear winter traits Uncertain heritability of stress-adaptation traits Some traits (facial eczema resistance, cold tolerance) have moderate heritability.44 Others (general disease resistance, adaptation to novel feed conditions) may not respond to selection as quickly as hoped.

10. CROSS-REFERENCES

  • Doc #74 — Pastoral Farming Under Nuclear Winter: Destocking framework and carrying capacity. Breeding guidance here must integrate with Doc #74 culling decisions, including dairy herd restructuring and breed implications.
  • Doc #77 — Seed Preservation and Distribution: Parallel document for plant genetic resources; similar diversity preservation principles.
  • Doc #80 — Soil Fertility Management: Pasture productivity affects carrying capacity and herd size.
  • Doc #36 — Clothing and Footwear: Wool requirements influence sheep breed priorities.
  • Doc #100 — Harakeke Fiber Processing: Textile context for wool breed decisions.
  • Doc #124 — Pharmaceutical Rationing: Veterinary pharmaceutical depletion drives disease resistance breeding priorities.
  • Doc #53 — Fuel Allocation: Fuel constraints drive draft animal requirements.
  • Doc #8 — National Asset and Skills Census: Should include AI technician numbers and rare breed populations. Wild deer and feral goats (captured during Doc #8 inventory) are additional genetic reservoirs.
  • Doc #151 — Trans-Tasman Relations and Trade: Potential source of genetic material once trade resumes.

FOOTNOTES

  1. LIC (Livestock Improvement Corporation) is a farmer-owned cooperative headquartered in Newstead, Hamilton, serving approximately 70–75% of NZ dairy herds with herd testing, AI, and genetic evaluation. Semen storage in liquid nitrogen dewars. Specific straw counts are commercially sensitive; the “millions of straws” figure is estimated from the scale of operations (~3–4 million inseminations per year nationally, with multi-year inventories). See https://www.lic.co.nz/ — exact figures require verification from LIC.↩︎

  2. CRV Ambreed (now CRV NZ) is the second major dairy genetics provider in NZ, serving approximately 20–25% of the dairy herd. Maintains its own bull teams and semen storage. See https://www.crv4all.co.nz/↩︎

  3. The Rare Breeds Conservation Society of New Zealand (RBCS) maintains breed registries, coordinates breeding programmes, and advocates for heritage breed conservation. Founded 1988. Breed populations cited are approximate and based on RBCS publications and general breed conservation literature. See https://www.rarebreeds.co.nz/ — specific population numbers for individual breeds require verification from RBCS.↩︎

  4. LIC operates the national herd testing database (MINDA), which records pedigree, production, and breeding data for most NZ dairy herds. This existing infrastructure is the foundation for any national livestock genetic register. Extending it to beef, sheep, and deer would require additional data collection systems. See LIC Dairy Statistics publications.↩︎

  5. LIC (Livestock Improvement Corporation) is a farmer-owned cooperative headquartered in Newstead, Hamilton, serving approximately 70–75% of NZ dairy herds with herd testing, AI, and genetic evaluation. Semen storage in liquid nitrogen dewars. Specific straw counts are commercially sensitive; the “millions of straws” figure is estimated from the scale of operations (~3–4 million inseminations per year nationally, with multi-year inventories). See https://www.lic.co.nz/ — exact figures require verification from LIC.↩︎

  6. Inbreeding depression effects: ~0.5–1% decline in conception rate per 1% increase in inbreeding coefficient, plus reduced growth rates, increased calf mortality, and disease susceptibility. See Falconer and Mackay, “Introduction to Quantitative Genetics,” 4th ed., 1996; Leroy, “Inbreeding depression in livestock species,” Animal, 8(5), 2014.↩︎

  7. Icelandic cattle, isolated since ~900 AD, show significantly reduced heterozygosity compared to other European breeds. See Adalsteinsson, “Possible changes in the frequency of the Northern European short tail gene in Icelandic sheep,” Journal of Agricultural Science, 1981; and more recent genomic analyses of Icelandic cattle.↩︎

  8. NZ primary livestock production (pre-war, approximate): dairy production approximately 21 billion litres of milk per year (milk solids ~1.8 million tonnes); meat production approximately 900,000–1,000,000 tonnes carcass weight per year (beef, lamb, venison combined). Combined milk-solid-equivalent basis approximately 3.5–4.5 million tonnes annually. Under 40–60% destocking and reduced per-animal productivity, total output might fall to 40–60% of pre-war levels, or roughly 1.5–2.5 million tonnes milk-solid-equivalent per year. 1–2% of that range = 15,000–50,000 tonnes. The 5,000–15,000 tonne figure in the text uses the lower end of output scenarios and the lower productivity-loss percentage — a conservative estimate. See Stats NZ Agricultural Production Statistics; DairyNZ Economic Survey; Beef + Lamb NZ Compendium of Farm Facts.↩︎

  9. NZ livestock numbers: approximately 10.2 million cattle (dairy and beef combined) and approximately 25–26 million sheep as of 2023–2024. Total stock units (~83 million) calculated using standard NZ stock unit equivalencies. See Stats NZ Agricultural Production Statistics and Beef + Lamb NZ Compendium of Farm Facts. Post-destocking figures derived from Doc #74 carrying capacity analysis.↩︎

  10. Effective population size formulas and their application to livestock are standard population genetics. See Falconer and Mackay (1996) cited above; also Meuwissen, T.H.E., “Maximizing the response of selection with a predefined rate of inbreeding,” Journal of Animal Science, 75, 1997.↩︎

  11. The 50/500 rule originates from Franklin, I.R. (1980) and Soule, M.E. (1980). More recent analyses suggest the long-term threshold may be higher — perhaps 1,000–5,000 — but the principle that short-term and long-term minimum population sizes differ remains well-accepted. See Frankham, R. et al., “Genetics and extinction,” Biological Conservation, 2010.↩︎

  12. Post-destocking herd estimates are derived from Doc #74’s carrying capacity analysis under nuclear winter conditions, applying species-specific destocking percentages to pre-war herd sizes from Stats NZ. These are illustrative ranges, not precise predictions — actual numbers depend on nuclear winter severity, regional pasture responses, and the timing and management of destocking decisions.↩︎

  13. Post-destocking herd estimates are derived from Doc #74’s carrying capacity analysis under nuclear winter conditions, applying species-specific destocking percentages to pre-war herd sizes from Stats NZ. These are illustrative ranges, not precise predictions — actual numbers depend on nuclear winter severity, regional pasture responses, and the timing and management of destocking decisions.↩︎

  14. Post-destocking herd estimates are derived from Doc #74’s carrying capacity analysis under nuclear winter conditions, applying species-specific destocking percentages to pre-war herd sizes from Stats NZ. These are illustrative ranges, not precise predictions — actual numbers depend on nuclear winter severity, regional pasture responses, and the timing and management of destocking decisions.↩︎

  15. NZ deer farming began in the late 1960s and 1970s, initially from captured wild deer (predominantly Red Deer descended from 19th-century liberations, plus Wapiti in Fiordland). Genetic base is narrower than cattle or sheep due to this more recent origin. See Deer Industry New Zealand publications, https://www.deernz.org/↩︎

  16. Post-destocking herd estimates are derived from Doc #74’s carrying capacity analysis under nuclear winter conditions, applying species-specific destocking percentages to pre-war herd sizes from Stats NZ. These are illustrative ranges, not precise predictions — actual numbers depend on nuclear winter severity, regional pasture responses, and the timing and management of destocking decisions.↩︎

  17. LIC (Livestock Improvement Corporation) is a farmer-owned cooperative headquartered in Newstead, Hamilton, serving approximately 70–75% of NZ dairy herds with herd testing, AI, and genetic evaluation. Semen storage in liquid nitrogen dewars. Specific straw counts are commercially sensitive; the “millions of straws” figure is estimated from the scale of operations (~3–4 million inseminations per year nationally, with multi-year inventories). See https://www.lic.co.nz/ — exact figures require verification from LIC.↩︎

  18. NZ deer farming began in the late 1960s and 1970s, initially from captured wild deer (predominantly Red Deer descended from 19th-century liberations, plus Wapiti in Fiordland). Genetic base is narrower than cattle or sheep due to this more recent origin. See Deer Industry New Zealand publications, https://www.deernz.org/↩︎

  19. AI consumables and domestic production: insemination sheaths and catheters are disposable single-use plastics; no known NZ domestic manufacturer. Semen extenders: Tris-citrate and similar formulations require fructose, citric acid, antibiotics, and egg yolk; these can be prepared from partially NZ-sourced ingredients but pharmaceutical-grade antibiotics require importation. CIDR (Controlled Internal Drug Release) devices contain progesterone in a silicone matrix; manufactured by MSD Animal Health internationally. Prostaglandins (PGF2α) used for oestrus synchronisation are biological products manufactured offshore. Estimated stock on hand at any given time: unknown; verification from LIC, veterinary distributors, and MPI required.↩︎

  20. Laparoscopic AI in sheep requires specialised equipment (laparoscope, insemination pipettes) and trained veterinary operators. It is used in NZ stud flocks but is not widespread in commercial flocks. Natural mating with managed ram rotation is the practical approach for most sheep genetic management. See NZSAP (NZ Society of Animal Production) proceedings for NZ-specific sheep AI data.↩︎

  21. NZ dairy 6-week in-calf rate data: national average 6-week submission rate approximately 88%, 6-week in-calf rate approximately 72–76% in recent DairyNZ monitoring. Natural mating without AI achieves similar conception rates in well-managed seasonal herds (bulls at 1:30–50 cow ratio from planned start of mating). The semen distribution advantage of AI is the key difference, not conception rate per se. Bull movement biosecurity risks: Bovine Viral Diarrhoea (BVD), Bovine Herpesvirus 1 (BHV-1, or IBR), and Johne’s Disease (MAP) are the main concerns for between-herd bull movement in NZ. Without adequate testing (which requires laboratory diagnostics that may be constrained under recovery conditions), physical bull movement between farms risks spreading these diseases. See DairyNZ Key Facts; NZVA biosecurity guidelines; MPI Animal Health data.↩︎

  22. LIC operates the national herd testing database (MINDA), which records pedigree, production, and breeding data for most NZ dairy herds. This existing infrastructure is the foundation for any national livestock genetic register. Extending it to beef, sheep, and deer would require additional data collection systems. See LIC Dairy Statistics publications.↩︎

  23. The productivity gap between hardiness-selected and production-selected livestock is well-documented in breed comparison studies. NZ’s own BW (Breeding Worth) system shows that selecting for survival and fertility traits rather than production traits reduces per-cow milk solids by approximately 10–20%. For beef cattle, frame-score and growth-rate differences between hardy and production breeds range from 15–30%. See LIC Dairy Statistics; Morris, S.T. et al., NZ Journal of Agricultural Research, various years.↩︎

  24. Nuclear winter temperature depression for the Southern Hemisphere mid-latitudes is estimated at 3–8°C depending on scenario severity and season. NZ’s maritime climate moderates the effect compared to continental interiors, but cooling of 3–5°C is plausible even in moderate scenarios. See Robock, A. et al., “Nuclear winter revisited with a modern climate model and current nuclear arsenals,” Journal of Geophysical Research, 2007; and Coupe, J. et al., “Nuclear Niño responses,” Nature Communications, 2021.↩︎

  25. Facial eczema is caused by sporidesmin, a mycotoxin from Pithomyces chartarum growing on dead pasture matter, primarily in Waikato, Bay of Plenty, and Hawke’s Bay regions during late summer and autumn. Estimated annual livestock losses in NZ pre-war: 30,000–50,000 sheep and cattle affected per major outbreak year, with treatment costs and production losses of tens of millions of dollars. Footrot (Dichelobacter nodosus) is the leading cause of lameness in NZ sheep, particularly in wet regions. Internal parasites (predominantly Haemonchus contortus, Trichostrongylus spp.) are a major burden in both sheep and cattle. See Emery, D.L. et al., “Livestock Disease and its Control in New Zealand,” various; MPI surveillance data at https://www.mpi.govt.nz/. Figures require verification from Veterinary Services.↩︎

  26. LIC (Livestock Improvement Corporation) is a farmer-owned cooperative headquartered in Newstead, Hamilton, serving approximately 70–75% of NZ dairy herds with herd testing, AI, and genetic evaluation. Semen storage in liquid nitrogen dewars. Specific straw counts are commercially sensitive; the “millions of straws” figure is estimated from the scale of operations (~3–4 million inseminations per year nationally, with multi-year inventories). See https://www.lic.co.nz/ — exact figures require verification from LIC.↩︎

  27. CRV Ambreed (now CRV NZ) is the second major dairy genetics provider in NZ, serving approximately 20–25% of the dairy herd. Maintains its own bull teams and semen storage. See https://www.crv4all.co.nz/↩︎

  28. Cryopreserved semen stored at -196°C in liquid nitrogen shows no measurable decline in viability over decades of storage. Calves have been born from semen stored for 40+ years. The limiting factor is physical integrity of the storage system, not biological degradation. See Mazur, P., “Freezing of living cells: mechanisms and implications,” American Journal of Physiology, 1984.↩︎

  29. Semen viability and temperature thresholds: viability begins declining significantly above approximately -80°C, with rapid and irreversible loss above -130°C (the glass transition temperature range of water in biological systems). Dewar failure modes: vacuum jacket failure can allow warm-up from -196°C to damaging temperatures within 24–72 hours. See Benson, C.T. et al., “Physics of Liquid Nitrogen-Cooled Biorepositories,” Biopreservation and Biobanking, 2011; FAO Animal Production and Health Paper No. 90, “Cryopreservation of Animal Genetic Resources,” 1992. Dewar failure rate data: verification from LIC operations required.↩︎

  30. Air separation units in NZ operated by BOC (Linde) and Air Liquide at Glenbrook, Christchurch, and Auckland. A typical 35-litre dewar loses ~0.5–1 litre LN2/day to evaporation. National semen storage LN2 requirements estimated at a few hundred litres per week — figures require verification from LIC and gas suppliers.↩︎

  31. ASU power consumption varies by scale and design. Small-to-medium plants producing 50–200 tonnes/day of oxygen (with nitrogen as co-product) typically consume 0.5–2 MW. NZ’s ASU installations are in this range. See Smith, A.R. and Klosek, J., “A review of air separation technologies and their integration with energy conversion processes,” Fuel Processing Technology, 2001.↩︎

  32. Semen viability and temperature thresholds: viability begins declining significantly above approximately -80°C, with rapid and irreversible loss above -130°C (the glass transition temperature range of water in biological systems). Dewar failure modes: vacuum jacket failure can allow warm-up from -196°C to damaging temperatures within 24–72 hours. See Benson, C.T. et al., “Physics of Liquid Nitrogen-Cooled Biorepositories,” Biopreservation and Biobanking, 2011; FAO Animal Production and Health Paper No. 90, “Cryopreservation of Animal Genetic Resources,” 1992. Dewar failure rate data: verification from LIC operations required.↩︎

  33. Air separation units in NZ operated by BOC (Linde) and Air Liquide at Glenbrook, Christchurch, and Auckland. A typical 35-litre dewar loses ~0.5–1 litre LN2/day to evaporation. National semen storage LN2 requirements estimated at a few hundred litres per week — figures require verification from LIC and gas suppliers.↩︎

  34. The Rare Breeds Conservation Society of New Zealand (RBCS) maintains breed registries, coordinates breeding programmes, and advocates for heritage breed conservation. Founded 1988. Breed populations cited are approximate and based on RBCS publications and general breed conservation literature. See https://www.rarebreeds.co.nz/ — specific population numbers for individual breeds require verification from RBCS.↩︎

  35. Enderby Island cattle descended from cattle left on the island by settlers in the 1890s. The population was removed by DOC in 1991–1993 and a small number were retained on the mainland. They represent over 100 years of isolation and natural selection. Population numbers are very small (estimated low hundreds) and maintained by a few dedicated breeders. See RBCS publications and DOC historical records.↩︎

  36. Kaimanawa wild horses are managed by the Department of Conservation under the Kaimanawa Wild Horses Plan. Population is maintained at approximately 300 through periodic musters and rehoming. Heavy draft breeds (Clydesdale, Percheron) are maintained by small numbers of enthusiast breeders; total NZ populations are not precisely documented but are estimated at a few hundred each. Draft breed registries may provide more precise numbers.↩︎

  37. Kunekune pigs were near extinction in the 1970s (~50 purebred animals known). Conservation by Michael Willis and John Simister, who located survivors on Maori farms, rebuilt the population to several thousand. Small, hardy, docile, efficient forager. See RBCS publications.↩︎

  38. NZ livestock numbers: approximately 10.2 million cattle (dairy and beef combined) and approximately 25–26 million sheep as of 2023–2024. Total stock units (~83 million) calculated using standard NZ stock unit equivalencies. See Stats NZ Agricultural Production Statistics and Beef + Lamb NZ Compendium of Farm Facts. Post-destocking figures derived from Doc #74 carrying capacity analysis.↩︎

  39. Herd rebuilding calculation: ~40% of calves are heifers, ~80% survive to breeding age, annual mortality 3–5%, yielding net annual growth of ~10–15%. From a 40% reduction, recovery takes ~6–10 years of compounding under good conditions; longer under feed-constrained nuclear winter recovery.↩︎

  40. The 5–10% annual expansion rate is a conservative estimate based on biological reproduction rates minus mortality, constrained by the requirement not to outpace pasture recovery. Unconstrained cattle herds can grow at 10–15% per year (see footnote 16); the lower figure reflects deliberate restraint during the recovery period when pasture productivity is still below normal.↩︎

  41. Kaimanawa wild horses are managed by the Department of Conservation under the Kaimanawa Wild Horses Plan. Population is maintained at approximately 300 through periodic musters and rehoming. Heavy draft breeds (Clydesdale, Percheron) are maintained by small numbers of enthusiast breeders; total NZ populations are not precisely documented but are estimated at a few hundred each. Draft breed registries may provide more precise numbers.↩︎

  42. Ox training timelines: a young steer (12–18 months) requires 6–12 months to train to basic commands and collar work; developing a reliable heavy-work ox capable of sustained ploughing typically requires 12–18 months or longer. Conformation and breed suitability: heavier-framed beef breeds with calm temperament (Hereford, Shorthorn, crossbreeds) are traditionally preferred over dairy breeds. See Starkey, P., “Working Animals in Agriculture and Transport,” Technical Centre for Agricultural and Rural Cooperation (CTA), 1993; and Pearson, R.A., “Draught Animal Power,” University of Edinburgh CTVM, 2005, cited above.↩︎

  43. Draft animal performance: historical and modern agricultural references report horse ploughing rates of 0.4–0.6 ha/day and ox ploughing rates of 0.2–0.4 ha/day, depending on soil type, plough design, and animal condition. Walking speeds from Starkey, P., “Animal Draught Power,” FAO, 1988; and Pearson, R.A., “Draught Animal Power,” University of Edinburgh CTVM, 2005.↩︎

  44. Facial eczema resistance in sheep: heritability estimates of 0.3–0.5 in NZ studies, with GGT (gamma-glutamyl transferase) response used as a selection criterion. Genetic progress demonstrated over 20+ years of selection in NZ. See Towers, N.R. et al., “Facial eczema resistance — genetic aspects,” New Zealand Veterinary Journal, various years; Beef + Lamb NZ Genetics programme documentation at https://beeflambnz.com/knowledge-hub/animal-genetics. Cold tolerance heritability in livestock is less well-characterised; estimates in the range of 0.1–0.3 for coat-related and metabolic thermoregulation traits. These figures require verification from AgResearch and Massey University published work.↩︎

← Back to Library Index