|
Isolating
Mechanisms
|
|
|
|
Roger K Butlin The University of Leeds, Leeds, UK
|
|
|
|
|
Advanced
|
|
|
doi:10.1038/npg.els.0001747
|
|
|
|
|
|
|
|
|
Isolating
mechanisms are intrinsic characteristics of species that reduce or prevent
successful reproduction with members of other species. Viewed genetically,
they are characters that act as barriers to the exchange of genes between
populations.
|
|
|
|
|
|
|
|
Introduction and
Classification
|
|
|
|
Successful sexual reproduction requires
many features of the male and female partners to be compatible. They must
be sexually mature in the same place and at the same time, they must
recognize each other as potential mates and coordinate their mating
activity (e.g. courtship behaviour in animals, flowering in plants or hyphal fusion in fungi). Following mating, the sperm
must reach and recognize the egg (in animals), the pollen tube must grow
through the style to the ovule (in plants), or nuclear migration and dikaryon formation must occur (in fungi). Nuclear
fusion must follow and the resulting zygote must be both viable and
fertile. Finally, the second generation offspring must also be viable and
fertile. In prokaryotes, successful conjugation or transformation is
required for gene exchange. See also: Sex: advantage; Bacterial reproduction and growth;
Plant reproduction; Reproduction in vertebrates: overview;
Genetics and variation in survival and
reproduction
|
|
|
|
Within species, natural selection acts to
maintain all of these interactions and, indeed, to increase their
efficiency. However, between species the sequence may be broken or impaired
at any point and the characters involved in this breakdown are known as
reproductive isolating mechanisms, or simply isolating mechanisms. This
terminology is unfortunate: the word ‘mechanism’ implies something that has
been elaborated by evolution for the function of isolation whereas
reproductive isolation is probably much more commonly an incidental effect
of divergence between populations for other reasons. For example, natural
selection may favour divergence in flowering time
in two allopatric populations of plants in
response to climatic differences. If the ranges of the populations
subsequently change so that they overlap, the flowering time difference may
result in some reproductive isolation but it certainly did not evolve as a
‘mechanism’ to prevent successful reproduction. See also: Natural selection: introduction;
Speciation: introduction
|
|
|
|
Reinforcement is the only process by which
natural selection directly favours reproductive
isolation, and it is a controversial process with limited empirical
support. For these reasons, many biologists now prefer to use the term
‘barrier to gene exchange’ because it avoids the implication of function
and focuses attention on the critical genetic consequences. However, use of
‘isolating mechanism’ is so widespread that it will be a long time before
it is displaced completely. See also: Reinforcement
|
|
|
|
It is helpful to classify isolating
mechanisms on the basis of the possible break points in the sequence
leading to successful reproduction. The classification in Table 1 is based on sexually
reproducing animals and flowering plants. Within these groups, some of the
suggested mechanisms are not universal: for example, animals with external
fertilization cannot show mechanical isolation. The classification can
easily be modified for sexual reproduction in fungi or protists.
An extension to prokaryotes is more difficult although mechanisms 1 and 2
apply whatever the mode of gene exchange, as does the requirement for
offspring to be viable and capable of further reproduction. Conjugation and
transformation require recognition and compatibility and so it is possible
to envisage mechanisms equivalent to 3–6 in Table 1. See also: Homologous genetic recombination during
bacterial conjugation
|
|
|
|
Note that geographical barriers between
populations can prevent gene exchange but they are not included in Table 1. Isolating mechanisms
are defined as intrinsic characteristics of organisms that prevent gene
exchange whereas geographical barriers are imposed externally. Both
external and intrinsic barriers to gene exchange can permit evolutionary
divergence, but external barriers are often temporary. See also: Species and speciation: overview
|
|
|
|
|
|
Premating Isolation
|
|
|
|
Premating isolation may be
the only barrier separating recently evolved species pairs. For example,
female spiders in the genus Schizocosa
will normally allow matings only by males of
their own species, which they identify by substrate-transmitted vibrations
(Stratton and Uetz,
1986). However, if the females are anaesthetized interspecific matings are
possible and they generate viable and fertile offspring. This suggests that
premating isolation can evolve more rapidly than postmating isolation and there is evidence from Drosophila
that this is a common pattern (Coyne and Orr, 1997).
|
|
|
|
Habitat and temporal isolation (mechanisms
1 and 2 in Table 1) may be particularly
important where speciation is sympatric since isolation may arise as an
incidental consequence of adaptation to alternative resources in the
environment, such as host plants for phytophagous
insects. The fruitfly Rhagoletis
pomonella appears to be in the process of
divergence into apple- and hawthorn-associated species. Currently, there is
probably still some gene exchange between host-associated populations but
it is impeded by the tendency of flies to mate on the fruits of the host
tree from which they emerged, and for the peaks of emergence of the two
populations to be timed to fit the different fruiting seasons of their
hosts (Feder et al., 1997). See also: Speciation: sympatric and parapatric; Adaptation and natural selection:
overview
|
|
|
|
Both flowering time and association with
habitats are likely to be important in reproductive isolation of plant
species. A good example involving both barriers is the partial isolation
between populations of the grass Agrostis
tenuis growing on pastures contaminated with
heavy metals from mining activities and those in surrounding grassland. The
populations are spatially separated as a result of their ecological
tolerances and also isolated by divergence in flowering time which has both
environmental and genetic components (McNeilly and Antonovics,
1968).
|
|
|
|
Ethological isolation has attracted a
great deal of research attention because the complex sequence of behavioural acts leading to mating in most animals is
potentially easily disrupted, and because the characters involved may be
expected to diverge rapidly under sexual selection (Butlin and Ritchie, 1994). This type
of isolation can be quantified relatively easily under laboratory
conditions using mate choice experiments to detect positive assortative mating: a tendency to mate more readily
with partners of the same species, population or phenotype. Although they
vary in design, these experiments generate an ‘isolation index’, the
frequency of interspecific matings
relative to intraspecific matings,
which can be readily compared across studies. Coyne and Orr (1997), for
example, compared many pairs of Drosophila species and showed that
the time taken to reach an isolation index of I = 1.0, indicating
complete premating isolation, is in the region of
3 million years for allopatric species pairs but
considerably less for sympatric pairs.
|
|
|
|
Closely related species frequently differ
strongly in sexual signal characters, even when they are difficult to
separate on the basis of other phenotypic or genetic traits. There are many
examples: closely related grasshopper, frog and bird species typically have
distinctive ‘songs’ used to attract and stimulate mates; male cichlid fish
in the spectacular radiations in African lakes differ strikingly in their
breeding coloration; fireflies flash with species-specific patterns; and
female moths have precisely defined pheromone blends that attract only
males of their own species. Where signals can be generated or modified
artificially, it has been possible to dissect the features that contribute
to reproductive isolation. Thus, both contact pheromone blend (Coyne et al., 1994) and
the temporal characteristics of the sound produced by wing vibration (Ritchie et al., 1999)
contribute to premating reproductive isolation
between Drosophila melanogaster and D. simulans. On the other hand, the unusually wide
head of Drosophila heteroneura males is a
species-specific character that is under sexual selection but does not
contribute to isolation between D. heteroneura
and its sibling species D. silvestris (Boake et al., 1997). See also: Bird song: steroid hormones and plasticity;
Pheromones in vertebrates;
Precopulatory reproductive strategies
|
|
|
|
Pheromones involved in recognition of
compatible cells or mycelia of opposite mating type have been described in
fungi (Casselton and Olesnicky,
1998). This type of signalling
probably underlies ‘interfertile’ versus ‘intersterile’ interactions between mycelia which can be
used to define ‘breeding groups’ in some fungi. For example, the
morphologically defined basidiomycete fungus Armillaria mellea
can be divided into many such groups (at least five in Europe and 10 in
North America) which are likely to represent biological species (see Brasier, 1997).
|
|
|
|
In many plants, features of floral
morphology, flower colour and scent determine the
range of pollinators that visit flowers of a particular species. If related
species attract different pollinator species, or even if individual
pollinators tend to move between similar flowers, interspecific
fertilization will tend to be restricted. Floral differences between
closely related species can be dramatic, just as sexual signal differences
are in animals. Two species of the monkey flower genus Mimulus
differ in flower colour (pink with yellow nectar
guides versus red without nectar guides), corolla form (wide with
forward-thrust petals versus narrow with recurved
petals), nectar volume, nectar concentration, and anther and stigma
positions (Bradshaw et al., 1995).
The first is adapted for bumblebee pollination and the second for
hummingbird pollination. Despite these extreme differences, which almost
completely prevent hybridization in nature, the species produce vigorous
and fertile hybrids when artificially cross-pollinated. See also: Flowers; Stamen and pollen development
|
|
|
|
|
|
Assortative Fertilization
|
|
|
|
Isolation that occurs between mating and
formation of the zygote is often referred to as ‘assortative
fertilization’. It is likely to be of great importance in organisms with
external fertilization, especially in sessile aquatic organisms that
broadcast their gametes although allochronic
isolation is also important in these cases. However, assortative
fertilization can also have a role in plant species through differential
pollen tube growth or in animal species with internal fertilization if
females mate repeatedly and store sperm.
|
|
|
|
Howard and Gregory (1993)
described a striking example in the field crickets Allonemobius
fasciatus and A. socius.
These two species form a mosaic hybrid zone in eastern North America and patterns in
the field suggest that they are strongly reproductively isolated. However, interspecific pairs mate readily and produce viable and
fertile offspring. The cause of isolation was a mystery until the offspring
of females that had been mated once to each of two males was examined using
genetic markers. The results are given in Table 2. Clearly there is
strong assortative fertilization, regardless of
mating order, which will act as a substantial barrier to gene exchange in
the field, provided multiple mating is common. See also: Interspecific interaction
|
|
|
|
In Allonemobius,
it is uncertain whether the cause of assortative
fertilization is differential sperm survival or fertilization success but
in sea urchins with external fertilization there is good evidence that
disruption of sperm–egg recognition is a major cause of reproductive
isolation. Studies on closely related species of the genus Echinometra show that eggs are very rarely
fertilized by heterospecific sperm even when they
are applied in excess under laboratory conditions; see Palumbi (1998) for a review. This
isolation appears to be due to a failure of attachment by a sperm protein, bindin, to the egg surface. The gene coding for bindin has been identified and sequenced. It shows an
unusual pattern of evolution: while part of the molecule is strongly
conserved, the remainder has many amino acid substitutions between species
– many more than expected from the number of synonymous base substitutions
in the DNA sequence. This suggests divergence driven by natural selection
but does not identify the source of the selection pressure. At the same
time, the bindin sequence is highly polymorphic
within species and evidence is starting to accumulate for a relationship
between bindin variation and variation in
fertilization success between pairs of urchins. See also: Sperm–egg interactions: sperm–egg binding
in invertebrates
|
|
|
|
|
|
Postzygotic Isolation
|
|
|
|
Successful mating and fertilization are
not enough for gene exchange: the resulting zygote must also be able to
complete development, survive to sexual maturity and be fertile. In many
pairs of closely related species, it is possible to overcome premating barriers to gene exchange in the laboratory
but the offspring produced are inviable or
sterile: there is postzygotic isolation. Frequent
production of unfit hybrids is rarely seen in nature (i.e. there is rarely postzygotic isolation without prezygotic
isolation). This may be because prezygotic
isolation evolves more quickly or because of the inherent instability of
pairs of populations that produce unfit hybrids: if they mate at random,
the rarer population produces more hybrid offspring than the commoner one
and so tends to be driven to extinction. The exception to this
generalization is the widespread occurrence of hybrid zones where the
ranges of divergent taxa meet. Here, the constant
removal of hybrids by selection is counterbalanced by movement of
individuals into the area from the parental populations (Barton and Hewitt, 1985). See also: Speciation: chromosomal mechanisms;
Hybrid zones; Speciation: genetics
|
|
|
|
Early research quickly identified a
pattern in the fitness of these hybrid offspring, now known as ‘Haldane’s rule’ (Haldane, 1922): in many cases only
one sex is inviable or sterile and this sex is
almost always the heterogametic sex (i.e. the sex with two different sex
chromosomes: the XY male in mammals and Drosophila, the ZW female in
butterflies and birds). See also: Fitness
|
|
|
|
Haldane’s rule has been
amply confirmed by subsequent research (Orr, 1997), and has been shown
to be an early stage in the speciation process, by a compilation of data on
reproductive isolation in relation to genetic distance between pairs of Drosophila
species (Coyne and Orr, 1997).
Sterility and inviability of hybrid males evolves
rapidly and is followed by much slower accumulation of female sterility or inviability. There has been much debate about the
genetic basis of Haldane’s rule,
and about the origin of postzygotic isolation in
general. It now seems likely that both result from the accumulation of new
alleles in diverging populations that work well in the genetic environment
of their own population but are incompatible with alleles present in other
populations. See also: Speciation: allopatric
|
|
|
|
In the yellow monkey flower Mimulus guttatus
various simple genetic systems have been identified that give rise to postzygotic isolation via either early death of hybrid
seedlings or male sterility in mature hybrid plants (Christie and Macnair,
1987). All involve interactions among loci, as expected. In some
cases it has been possible to show that the interaction causing sterility
is a pleiotropic effect of an allele that confers
tolerance to copper contamination in the soil. Thus, in these cases,
isolation is an incidental side effect of rapid adaptation rather than a
result of long-term divergence by genetic drift. See also: Adaptation: genetics; Heavy metal adaptation
|
|
|
|
Postzygotic isolation need not
necessarily be due to genetic incompatibility, it may also result from
reduced competitive ability of hybrids. In some Canadian lakes, the
three-spine stickleback Gasterosteus
aculeatus occurs in two morphs: a small
slender form (‘limnetic’) that feeds on plankton
in open water, and a larger, deeper bodied form (‘benthic’) that feeds on
invertebrates on the bottom and on vegetation. Hybrids have intermediate
body form but are viable and fertile, as expected from the very recent
divergence of the parental types. However, competition experiments show
that the hybrid phenotype performs worse than the limnetic
form in open water, and worse than the benthic form in its preferred
habitat (Hatfield and Schluter,
1999). This leads to selection against intermediates and reduces
gene exchange between the forms.
|
|
|
|
The final class of isolating mechanism,
hybrid breakdown, has been documented relatively rarely because it is only
manifest in second generation hybrids. Two races of the grasshopper Caledia captiva
in Australia provide a
well-documented example (Shaw et al., 1986). The
Moreton and Torresian
races show extensive chromosomal differentiation due to pericentric
inversions. The F1 is viable and fertile but the F2
(F1 × F1) and backcross (F1 × parental)
offspring show viability reductions of up to 50%. This is apparently due to
break-up of coadapted gene complexes by recombination
in the F1 hybrid. It represents a substantial barrier to gene
exchange, as evidenced by the narrow hybrid zone where the ranges of the
two taxa meet.
|
|
|
|
|
|
Originally published: January
2000
|
|
|
|
|
|
References
|
|
|
|
Barton NH and Hewitt GM (1985) Analysis of hybrid zones. Annual
Review of Ecology and Systematics 16:
113–148. Links
|
|
|
|
Boake CRB, De Angelis MP and Andreadis
DK (1997) Is sexual selection and species recognition a continuum? Mating
behaviour of the stalk-eyed fly Drosophila heteroneura.
Proceedings of the National Academy of Sciences of the USA 94: 12442–12445. Links
|
