Minimum viable population

Part of a series on

Conservation

The extinction vortex: once a population is small, inbreeding, drift and lost fitness lower its effective size further. Diagram by John Wilson and Richard Primack (2019).

Cannabis conservation·Grow-out priorities·Populations

Threat Categories

Extinction

Genetic Threats

Genetic erosion·Genetic contamination·Introgression·Gene flow·Hybridisation·Population bottleneck·Genetic drift·Inbreeding depression

Habitat & Environmental Threats

Habitat loss

Eradication Tracking & Analysis

Eradication·Herbicide spraying of cannabis·Herbicide spraying in Mexico·Herbicide spraying in the United States·Herbicide spraying in Colombia·Herbicide spraying in Belize·Herbicide spraying in South Africa·US eradication abroad·Yukthiya·Cannabis eradication in Kirivong·Cannabis eradication in Fiji·Campaigns by country

International Enforcement

Cannabis prohibition·US eradication abroad·INCB

Legal & Policy

Farmers' rights·Access and benefit-sharing·Intellectual property·Single Convention on Narcotic Drugs·1988 Convention·Cannabis legalisation movement

Population Assessment

Minimum viable population·Population viability analysis·Effective population size

Conservation Monitoring

Population monitoring·Genetic monitoring

Conservation Methods - Ex Situ

Ex situ conservation·Gene bank·Cryopreservation·In vitro conservation·Core collection·Seed storage

Conservation Methods - In Situ

In situ conservation·On-farm conservation·Genetic reserve·Community seed bank·Feral and ruderal populations

Genetic Conservation

Germplasm·Genetic resources·Crop wild relative·Conservation genetics

Sampling & Collection

Germplasm collecting·Germplasm sampling

Preservation Strategies

Conservation prioritisation·Complementary conservation

Recovery & Restoration

Reintroduction·Genetic rescue

Conservation Science

Landrace conservation

Portal·Category

A minimum viable population (MVP) is the smallest number of individuals a population needs in order to have a specified chance of surviving for a specified period.^: 376: 307 Both the chance and the period are chosen by whoever sets the figure. Estimates for a single species have differed more than a hundredfold.^{: 308, 311} Conservation biologists also set the minimum in terms of effective population size. The "50/500 rule" puts it at an effective size of 50 to avoid inbreeding depression in the short term and 500 to keep the capacity to evolve in the long term. Later work has argued for 100 and 1,000.^: 57–59

A meta-analysis published in 2007 pooled estimates for 212 species: 182 were vertebrates and 22 were plants.^: 164 Genetic indicators under the Convention on Biological Diversity long dealt only with agricultural species and their wild relatives, through counts such as the number of gene bank accessions. A threshold of 500 proposed in 2020 was meant for both domesticated and wild species, with breeds, landraces and varieties treated as the agricultural counterparts of wild populations.^{: 3, 5–7} For Cannabis, the Vavilov Institute in Russia set a minimum of 1,000 plants for each regrowing of a stored accession, to carry nearly all of the accession's gene pool into its seed.^: 10

Definition and criteria

M. L. Shaffer introduced the concept in 1981. He defined a minimum viable population as "the smallest isolated population having a 99% chance of remaining extant for 1000 years despite the foreseeable effects of demographic, environmental, and genetic stochasticity, and natural catastrophes". He called his choice of probability and period "tentatively and arbitrarily" made, holding that the level of risk to accept was a matter for society and science alike. In practice a period of 50 to 100 years and a 5 per cent risk of extinction became the criteria most often used.^: 308 Other studies have used a 95 per cent chance over several centuries or a 99 per cent chance over 40 generations. Richard Frankham, Corey Bradshaw and Barry Brook proposed in 2014 that every estimate report the last of these. They measured the period in generations because extinction risk scales with generations better than with years.^: 61

The minimum can be a carrying capacity, a number of adults or, in genetic terms, an effective population size. The hazards it guards against are those of small numbers: catastrophes and chance variation in births and deaths, in the environment and in the genes. They tend to reinforce one another in a feedback that M. E. Gilpin and Michael Soulé called the extinction vortex.^: 57 Where a population is still declining from a continuing cause such as habitat loss or overexploitation, Brook, Lochran Traill and Bradshaw regard its MVP as irrelevant, since its long-term persistence is already ruled out.^: 381

Shaffer named five ways of finding an MVP: experiments, biogeographic patterns, theoretical models, simulation models and genetic considerations. Simulation became the most used. A stochastic model projects the population forward by Monte Carlo methods, with chance events in births, deaths and the environment, together with inbreeding and catastrophes. The starting size is changed until the model finds the smallest population that lasts the chosen period with the chosen probability.^: 308 The simulation is a population viability analysis (PVA). An MVP and a PVA ask different questions. An MVP fixes the acceptable risk and asks how large a population must be. A PVA estimates the risk a population faces at its present size under the threats acting on it.^: 376

The 50/500 rule

Clarkia pulchella at Dry Falls State Park, Washington. In a field study that Frankham and colleagues cite, a 5 per cent rise in inbreeding raised the share of its populations that died out from 25 to 69 per cent.

In separate papers of 1980, I. R. Franklin and Soulé proposed that an effective population size of 50 was enough to prevent inbreeding depression in the short term. Franklin also put the size needed to keep the ability to evolve indefinitely at 500.^: 57 The figure of 50 came from animal breeding, where an increase in inbreeding of 1 per cent a generation, the rate at an effective size of 50, was considered tolerable. The figure of 500 came from the balance between the additive genetic variation that mutation adds to a quantitative trait and the share that genetic drift removes in each generation. Curtis Flather and colleagues call the two values goals for keeping populations genetically diverse, with "little direct connection with extinction risk".^: 308

In 2014 Frankham, Bradshaw and Brook argued that both figures should be at least doubled. Franklin and Soulé had relied on breeders' experience and on limited data from domestic and laboratory animals. Inbreeding depression is generally stronger in the wild than in captivity. Laboratory populations kept at effective sizes of 50 and above had shown it. By the authors' calculation, five generations at an effective size of 50 would lower total fitness in the wild by 26 per cent on average. A loss of 10 per cent over five generations, they proposed, should be the tolerable limit. Theory put that limit at an effective size of 142, so they recommended at least 100. In support they cited a study of the plant Clarkia pulchella in which a 5 per cent rise in the inbreeding coefficient raised the share of populations that died out from 25 to 69 per cent.^: 57–58

For the long term, the three authors argued that a population's ability to adapt depends on variation in fitness itself. Much of that variation lies in mildly harmful alleles held at low frequencies, which drift readily removes. By the theory and experiments they reviewed, diversity of this kind approaches the level of a very large population only at an effective size of about 1,000. They recommended at least 1,000. Populations below that size are not doomed in the short or medium term, they wrote, but their capacity to evolve wears away over time.^: 59 Jamieson and Allendorf, in a review of 2012, had kept the original values.^: 57–58

Simulations by Noelia Pérez-Pereira and colleagues, published in 2022, added genetic purging, the removal by selection of harmful recessive alleles that inbreeding exposes. Purging lowered the size needed for long-term survival. They put the minimum at an effective size of about 400 for species with moderately high reproductive rates, close to the classic 500. Species with low reproductive rates, among them many mammals, needed more than 1,000. They note that their model left out catastrophes and declines driven by causes such as habitat loss or overexploitation, which could raise the estimates.

The thresholds are effective sizes. Applying them to a count of plants or animals needs the ratio between effective and census size, which varies widely between species.^: 59–60 Under a rule of thumb of 0.1, an effective size of 500 corresponds to a census of about 5,000 adults.^: 7

The search for a general figure

A grizzly bear in Yellowstone National Park (2017). Nine published estimates of the grizzly bear's MVP, put on a common footing, ranged from 395 to 44,259.

Several studies in the 2000s pooled estimates across species. In 2003 Reed and colleagues ran PVAs for 102 species of vertebrates and found a mean standardised MVP of 7,316. They wrote that habitat for "approximately 7000 adult vertebrates" was needed for long-term persistence.^: 309 In 2006 Brook, Traill and Bradshaw fitted population models to time series of abundance for 1,198 species: 629 invertebrates, 529 vertebrates and 30 plants.^: 376 For a 90 per cent chance of lasting 100 years, the median MVP was 1,377 individuals. Models with density dependence gave median MVPs nearly two orders of magnitude below those without it. Traits commonly linked with extinction risk, such as body size, range and human impact, explained almost none of the variation in MVP between species, though the same traits predicted whether a species was on the IUCN Red List. The authors judged the estimates a "useful rule of thumb" that should not be used as precise conservation targets.^{: 375, 378–381}

Traill, Bradshaw and Brook then gathered 287 published estimates for 212 species from 141 sources and standardised them to a 99 per cent chance of persisting for 40 generations. The median was 4,169 individuals, with a 95 per cent confidence interval of 3,577 to 5,129. No ecological or life-history trait they tested predicted the standardised values well. The authors concluded that an MVP depends on the context of each species or population. For most species, they expected it to exceed a few thousand individuals.^{: 159, 164} In 2010 Traill and colleagues wrote that populations need "at least 5000 adult individuals".^: 307

Flather and colleagues rejected any single figure. In data assembled by Traill and colleagues, 52 species had two or more independent estimates. After standardisation the highest estimate for a species was a median 15.4 times its lowest. Nine estimates for the grizzly bear ranged from 395 to 44,259.^: 311 When Flather and colleagues kept only the populations in Brook's data that were neither growing nor declining much, the median MVP fell from 1,181 to 355 under a 40-generation criterion.^: 311 Flather and colleagues accepted that several populations totalling thousands of individuals would probably be needed for long-term persistence. They also recalled that the passenger pigeon, numbering three to five billion in the 1800s, still went extinct.^: 314 Frankham and colleagues answered that conservation managers need general rules, because the alternative is usually unscientific decisions taken at political and bureaucratic levels.^: 57

In conservation assessment

The population-size thresholds of criterion C of the IUCN Red List derive from effective sizes, two of them explicitly from the values of 50 and 500. Frankham and colleagues proposed doubling them: from 250 to 500 for Critically Endangered, from 2,500 to 5,000 for Endangered and from 10,000 to 20,000 for Vulnerable.^{: 58, 61} The endangered-species laws of the United States, Canada and Australia use MVP concepts in decisions to list and delist species.^: 314

In 2020 Sean Hoban and colleagues proposed, for the post-2020 framework of the Convention on Biological Diversity, an indicator counting the populations within species whose effective size is above 500 against those below it. An effective size below 50 would call for immediate action. They call the thresholds minimum guidelines for risk assessment. A population above 500 may still need conservation work.^: 6 The Kunming-Montreal Global Biodiversity Framework, adopted in 2022, has among its headline genetic indicators the proportion of populations within species with an effective size above 500.

Plants, crops and landraces

The 22 plants in Traill and colleagues' meta-analysis were mosses, ferns, flowering plants and gymnosperms. Their median standardised MVP was 4,824, with a 95 per cent confidence interval of 2,512 to 15,992. The 95 mammals had a median of 3,876 and the 48 birds 3,742.^: 164 Vertebrates made up 47 per cent of all species in the MVP literature reviewed, though they are a few per cent of named species.^: 163

Hoban and colleagues found the earlier genetic indicators of the Convention on Biological Diversity restricted to agricultural species and their wild relatives. One counted the plant genetic resources held in medium- or long-term storage, in practice the accessions in gene banks. They judged it loosely tied to genetic diversity, since the holdings of a species are often the offspring of one population or a few individuals. They proposed counting only collections that are drawn from across the range, large and backed up elsewhere, for example seed from more than 50 plants in each of at least five populations.^{: 3, 5} Their second proposed indicator, the share of a species' distinct populations that survive, treats the loss of breeds, landraces or varieties in crops as the counterpart of the loss of wild populations.^: 7

For the regrowing of seed accessions, the gene bank standards of the Food and Agriculture Organization ask for as few cycles as possible and sufficiently large effective population sizes.^: 10

Cannabis

Matthew Welling and colleagues call the regrowing of stored cannabis seed especially problematic. Cannabis is wind-pollinated and dioecious. Its male and hermaphrodite plants shed large amounts of pollen. An accession grown outdoors may need to stand up to 5 km from others to avoid cross-pollination. From 1993 the Vavilov Institute and the International Hemp Association set a minimum of 1,000 plants for each regrowing of an accession. The figure was meant to carry nearly all of the accession's gene pool into its seed. It is a standard for regrowing a stored accession, set by the share of its gene pool to be kept. It is not an estimate of the size at which a population persists.^: 10

See also

• Effective population size
• Population viability analysis
• Inbreeding depression
• Population bottleneck
• Germplasm sampling
• Genetic erosion