Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

Extinction and the Four Horsemen

15 views
Skip to first unread message

Doug Bashford

unread,
Oct 11, 2009, 12:47:58 PM10/11/09
to

� Area: Sci.Environm
���������������������������������������������������������
Msg#: 53 Date: 03-12-94 16:44
From: Alan Mcgowen Read: Yes Replied:
No
To: All Mark:

Subj: Ecocentral 10
������������������������������������������������������������������������������

Well, Alan Mcgowen said to All on 03-12-94 16:44
about: Ecocentral 10....

AM> From: Alan McGowen <alanm igc.apc.org>

AM>
AM> /* Written 1:09 pm Apr 12, 1992 by alanm in cdp:sci.environmen */
AM> /* ---------- "ECO CENTRAL 10" ---------- */
AM> ECO CENTRAL 10
AM>
AM> Extinction and the Four Horsemen
AM>
AM> Extinction can be either *deterministic*, when a resource or
AM> environmental condition essential for life is removed, or
AM> *stochastic*. Deterministic extinctions are wrought by
AM> catastrophes large and small: a volcanic eruption, an asteroid
AM> impact, a clearcut. The occurrence of the catastrophes themselves
AM> may be stochastic, but the extinctions they produce are
AM> deterministic. Depriving plants of light for many months will
AM> kill them, acidifying a lake sufficiently will kill fish,
AM> depriving a population of monkeys of their forest with a clearcut
AM> will kill the monkeys. In general, when the environment moves
AM> outside the range of the tolerances that define a species' niche,
AM> deterministic extinction occurs. The characteristic time to a
AM> deterministic extinction by catastrophe is unrelated to
AM> population size, i.e. a population of size 2N does not
AM> typically wait far longer than one of size N.
AM>
AM> *Stochastic* extinctions, on the other hand, are more probable
AM> the smaller a population is, and the characteristic time to
AM> extinction by stochastic processes increases with population
AM> size. The waiting time for extinction by *demographic
AM> stochasticity* increases extremely rapidly with N, and only very
AM> tiny populations are vulnerable to extinction by this means.
AM> However, tiny populations are quite vulnerable, and the _coupe de
AM> gras_ of extinction is frequently dealt to a remnant population
AM> by demographic fluctuations. The time to extinction increases
AM> more slowly with K for *environmental stochasticity*. Although
AM> the details are model-dependent, under fairly plausible
AM> assumptions about the behavior of the environmental fluctuations,
AM> the time to extinction increases roughly logarithmically with K
AM> [Leigh].
AM>
AM> Thus the possibility exists of determining a K for a *Minimum
AM> Viable Population* (MVP), such that if the population is smaller
AM> than MVP, extinction is rapid, while if it is greater, the chance
AM> of extinction is low and the population will be expected to
AM> persist for a long time. The greater the population is than the
AM> MVP, the lower the chances of extinction. The problem of
AM> determining an MVP is called *population vulnerability analysis.*
AM>
AM> The Four Horsemen
AM>
AM> There are four extinction processes that need to be considered
AM> when doing a population vulnerability analysis. These "four
AM> horsemen" are four positive feedback processes that drive
AM> populations towards extinction. [Gilpin]
AM>
AM> The *demographic process* occurs when chance lowerings of N and
AM> increases in var(r), make the population more vulnerable to
AM> extinction. It does not matter how the changes in N or var(r) are
AM> produced, by demographic randomness or environmental fluctuation.
AM> The point is that when N is decreased and var(r) [or var(K)]
AM> increased the population is less stable, and more likely to
AM> experience further fluctuation. The demographic process is
AM> usually the proximate cause of a stochastic extinction.
AM>
AM> The *fragmentation process* operates through the spatial
AM> distribution of a population. As habitat is fragmented (e.g. by
AM> many clearcuts, roads, or scattered developments), larger
AM> populations are carved up into smaller, more isolated ones. The
AM> smaller populations are at greater risk of extinction by the
AM> demographic process. As they go extinct, the mean distance
AM> between the small populations increases, and reestablishment by
AM> dispersal becomes less likely, i.e. effective fragmentation
AM> increases, exacerbating the fragmentation process.
AM>
AM> The fragmentation process is generally slower than the
AM> demographic process.
AM>
AM> The remaining two "horsemen" operate through genetics.
AM>
AM> The *fitness loss process* occurs as a result of *inbreeding
AM> depression*. In small populations, genetic drift is strong and
AM> removes population genetic diversity. One result of this is that
AM> gene loci become *fixed* to single alleles with no other variants
AM> in the population. Because genetic drift is quite random, the
AM> fixed loci are as likely to have a deleterious allele as the only
AM> one remaining as a beneficial one. Also, the entire population is
AM> now homozygous at the fixed locus, i.e. both copies of the gene
AM> are the same allele, rather than two different alleles (a
AM> condition called *heterozygosity*.] Both the fixation of
AM> deleterious alleles and loss of heterozygosity (which protects
AM> against the effects of deleterious recessives) take their toll on
AM> the organism. It becomes less fertile, shorter-lived, less
AM> resistant to diseases, less good at predation if a predator and
AM> less good at escaping from predators if a prey. These effects can
AM> be very large, substantially reducing r, and this reduction of r
AM> is inbreeding depression. The reduced r leads to a reduced N,
AM> and a further increase in inbreeding, and further fitness loses.
AM>
AM> The fitness loss process is generally slower than the
AM> fragmentation process.
AM>
AM> The *adaptation loss process* is another result of the loss of
AM> genetic diversity due to genetic drift in smaller populations.
AM> The rate of adaptation is proportional to genetic diversity, a
AM> result known as the fundamental theorem of natural selection,
AM> first proved by Ronald Fisher. As population genetic diversity
AM> decreases, so does the ability of the species to track changes in
AM> the environment (e.g. evolving diseases, climate shifts, changes
AM> in abundance of different prey species or plants, etc.) This
AM> leads to a gradual loss of adaptedness to the environment,
AM> placing the population at a subtle competitive disadvantage, or
AM> increasing predation pressure on it, and generally reducing its
AM> absolute fitness (its numbers). This of course leads to still
AM> stronger genetic drift, more loss of population genetic
AM> diversity, slower rates of adaptation, and still more loss of
AM> adaptedness to the environment.
AM>
AM> The adaptation loss process is the slowest -- the most insidious
AM> -- of the four. The only way to protect against adaptation loss
AM> is for the population to be large enough that loss of alleles to
AM> genetic drift is balanced by mutation and natural selection. This
AM> requires a quite healthy population size.
AM>
AM> * * * * * * *
AM>
AM> When there is no extinction crisis in progress, the extinction
AM> rate is more or less balanced by the speciation rate, or even
AM> slightly exceeded by the speciation rate, since species diversity
AM> has increased throughout the Phanerozoic. The conditions for
AM> speciation to occur are much like those for adaptation to be
AM> maintained, except more so: i.e. a species which has many locally
AM> adapted populations and much geographical variation is most
AM> likely to produce new species. This generally requires an even
AM> larger total population size than protection against loss of
AM> adaptation does.
AM>
AM> Of course speciation is also a stochastic process, and there is
AM> some possibility that a species with small numbers may
AM> nevertheless speciate: it seems quite clear that new species
AM> almost always arise from tiny *founder populations*. However,
AM> most tiny populations succumb to the four horsemen of extinction.
AM> In order for a few to speciate, many must become extinct. Thus
AM> the probability of speciation is increased when a species
AM> constantly produces many such candidate founder populations. This
AM> is most likely when the species has a very large total K
AM> throughout a large range, and much local variation of selection
AM> pressures acting upon it, i.e. a diverse environment throughout a
AM> large range.
AM>
AM> Thus we can think of a "fifth horseman" operating at an even
AM> slower rate than the adaptation loss process: a *speciation loss
AM> process* in which contraction of ranges and reductions of
AM> environmental diversity lead to decreased opportunities for
AM> speciation. When the speciation rate is exceeded by the
AM> extinction rate, as local populations become extinct the
AM> diversity of local selection pressures decreases and the ranges
AM> of species shrink, leading to still lower speciation rates. To
AM> reverse this process would require significant expansions of
AM> ranges and increases in local environmental diversity (especially
AM> species diversity) to restore local variation of selection
AM> pressures.
AM>
AM> Refs
AM>
AM> Gilpin, Michael E and Michael E. Soule, 1986. Minimum viable
AM> populations: processes of species extinction. In _Conservation
AM> Biology: The Science of Scarcity and Diversity_, Sinauer,
AM> Sunderland, Massachusetts.
AM>
AM> Leigh, E.G. 1981. The average lifetime of a population in a
AM> varying environment. J. Theoretical Biology 90:213-39
AM>
AM>
AM> -!---------
AM> Alan McGowen
AM> -- Transfer complete, hit <RETURN> to continue --

AM> -!-
AM> ! Origin: ONE WORLD Usenet<->Fidonet (1:102/129.1)

- If you scratch a cynic,
- you'll find a defeated idealist.

0 new messages