4.4.3 The Controversy Between Neutralism and Selectionism
143
one of these niches; the population as a whole is
highly polymorphic. In contrast, each individual
of a "fine-grained species" can survive in each
niche. In the latter case, there is little polymorphism as the same universal alleles are found in all
individuals. Similar terms, but with variation of
environmental parameters in time rather than in
space, are used in the "environment amplitude
variation" hypothesis to explain protein polymorphism. As these hypotheses were later
contradicted by the discovery of high polymorphism amongst deep-sea creatures and tropical
invertebrates living in constant, homogeneous
biotopes, Valentine and Ayala developed, in 1971,
the "resource predictability variation" hypothesis: With a continuous, high food supply, specialization is advantageous because it reduces intraspecific competition; this results in higher polymorphism. During periods of food shortage, all
individuals must be able to make use of any form
of nutrient and they are, therefore, genetically
similar [17, 294, 439]. Because it is never known
exactly what selective value the time- or spacevariable parameters have at anyone moment,
clear confirmation or rejection of the niche theories is hardly possible.
4.4.3 The Controversy Between Neutralism
and Selectionism
Neutralism and selectionism are not really alternatives; the question is no longer whether
(quasi-) neutral genetic alterations exist, but
rather what proportion they represent of the total
genetic variability. An irrefutable argument for
the predominance of neutral alterations in molecular evolution is the fact that the lower their
importance, the faster molecular characters
evolve, and the more they vary. If one argues
that characters of minor importance even react to
weaker selection pressure [439], then with the
assumption of very weak selection effects the
selectionist and neutral theories are reconciled.
The exact proportion of neutral genetic variation
will in fact never be determined. Neutrality, i.e.
the absence of selection, as a "negative property
cannot be conclusively proven. The particular
advantage of the neutral theories is that they
allow relatively simple, quantitative models of
molecular evolution and molecular polymorphism to be established.
The following arguments are mainly brought in
favour of selection as the cause of protein polymorphism:
1. The relative frequency of the alleles in natural
populations does not agree with the predictions of the neutral theories.
2. Protein polymorphism (in the terms of the
niche theory) is correlated with variation in
time and space of the environment.
3. Selection can be shown directly for single gene
loci:
a) Alleloenzymes have different properties in
vitro.
b) Organisms with different alleloenzymes
have different physiological properties and
are subject to selection in the laboratory or
the field.
c) Allele frequencies are correlated with the
geographical location or a site-dependent
environmental factor.
d) Reproductively isolated populations (e.g.
different species) in the same biotope have
matching allele frequencies.
Many studies have used statistical methods to
examine whether the allele frequencies in natural
populations agree with the predictions of the
neutralist theories [183, 254,294], and a negative
result has often been interpreted as evidence for
the correctness of the selectionistic explanation of
protein polymorphism. There are, however, many
basic problems with such studies. Most of the
equations are only valid for equilibrium; however, this can be disturbed by brief changes in
population size and may not be immediately reestablished. Phenomena such as coupling and
epistatic interaction between genes are not taken
into account. Finally, the models are always concerned with on the one hand, the relationship
between deterministic forces, like the mutation
rate v, the selection coefficient s or the migration
rate m, and on the other hand, probability, which
is proportional to lINe. Terms, such as Nev, Nes or
Nem, are derived in which neither the very high
value of Ne nor the very low values of v, s or m can
be sufficiently accurately determined; almost any
model can be "proven" with appropriate plausible assumptions for these parameters [237].
According to the niche theories, protein polymorphism should be correlated with the variability of living conditions [126]. Generalists, i.e. animal species with large distribution areas, broad
ecological niches and large populations, should
have a higher mean heterozygosity than specialists; this trend is confirmed by the analysis of severallarge collections of data [21]. The low heterozygosity of the mammals (Table 4.8) is also indicated by the niche theories; mobile mammals
143
one of these niches; the population as a whole is
highly polymorphic. In contrast, each individual
of a "fine-grained species" can survive in each
niche. In the latter case, there is little polymorphism as the same universal alleles are found in all
individuals. Similar terms, but with variation of
environmental parameters in time rather than in
space, are used in the "environment amplitude
variation" hypothesis to explain protein polymorphism. As these hypotheses were later
contradicted by the discovery of high polymorphism amongst deep-sea creatures and tropical
invertebrates living in constant, homogeneous
biotopes, Valentine and Ayala developed, in 1971,
the "resource predictability variation" hypothesis: With a continuous, high food supply, specialization is advantageous because it reduces intraspecific competition; this results in higher polymorphism. During periods of food shortage, all
individuals must be able to make use of any form
of nutrient and they are, therefore, genetically
similar [17, 294, 439]. Because it is never known
exactly what selective value the time- or spacevariable parameters have at anyone moment,
clear confirmation or rejection of the niche theories is hardly possible.
4.4.3 The Controversy Between Neutralism
and Selectionism
Neutralism and selectionism are not really alternatives; the question is no longer whether
(quasi-) neutral genetic alterations exist, but
rather what proportion they represent of the total
genetic variability. An irrefutable argument for
the predominance of neutral alterations in molecular evolution is the fact that the lower their
importance, the faster molecular characters
evolve, and the more they vary. If one argues
that characters of minor importance even react to
weaker selection pressure [439], then with the
assumption of very weak selection effects the
selectionist and neutral theories are reconciled.
The exact proportion of neutral genetic variation
will in fact never be determined. Neutrality, i.e.
the absence of selection, as a "negative property
cannot be conclusively proven. The particular
advantage of the neutral theories is that they
allow relatively simple, quantitative models of
molecular evolution and molecular polymorphism to be established.
The following arguments are mainly brought in
favour of selection as the cause of protein polymorphism:
1. The relative frequency of the alleles in natural
populations does not agree with the predictions of the neutral theories.
2. Protein polymorphism (in the terms of the
niche theory) is correlated with variation in
time and space of the environment.
3. Selection can be shown directly for single gene
loci:
a) Alleloenzymes have different properties in
vitro.
b) Organisms with different alleloenzymes
have different physiological properties and
are subject to selection in the laboratory or
the field.
c) Allele frequencies are correlated with the
geographical location or a site-dependent
environmental factor.
d) Reproductively isolated populations (e.g.
different species) in the same biotope have
matching allele frequencies.
Many studies have used statistical methods to
examine whether the allele frequencies in natural
populations agree with the predictions of the
neutralist theories [183, 254,294], and a negative
result has often been interpreted as evidence for
the correctness of the selectionistic explanation of
protein polymorphism. There are, however, many
basic problems with such studies. Most of the
equations are only valid for equilibrium; however, this can be disturbed by brief changes in
population size and may not be immediately reestablished. Phenomena such as coupling and
epistatic interaction between genes are not taken
into account. Finally, the models are always concerned with on the one hand, the relationship
between deterministic forces, like the mutation
rate v, the selection coefficient s or the migration
rate m, and on the other hand, probability, which
is proportional to lINe. Terms, such as Nev, Nes or
Nem, are derived in which neither the very high
value of Ne nor the very low values of v, s or m can
be sufficiently accurately determined; almost any
model can be "proven" with appropriate plausible assumptions for these parameters [237].
According to the niche theories, protein polymorphism should be correlated with the variability of living conditions [126]. Generalists, i.e. animal species with large distribution areas, broad
ecological niches and large populations, should
have a higher mean heterozygosity than specialists; this trend is confirmed by the analysis of severallarge collections of data [21]. The low heterozygosity of the mammals (Table 4.8) is also indicated by the niche theories; mobile mammals