Speciation (cont.)

Peripheral isolates

Drift can move a population across an adaptive valley, if one population is small enough.

Peripheral isolates are small populations that become isolated on the periphery of the parent population's range.
Because of small effective size, alleles that are selected against as heterozygotes may be fixed by drift.
If the populations are brought back together, there is selection to avoid hybridization because hybrid offspring have low fitness.

Recall the case of selection with multiple adaptive peaks.
 
 

Drift could move a population from one peak to another.
If the populations If the populations come back into contact, there would then be selection to not interbreed because hybrids would have reduced fitness.
             Recall that this is referred to as "reinforcement".

Example: Chromosomal inversions.
If an organism is heterozygous for an inversion on a chromosome, then crossing over in the inverted region during Metaphase 1 of Meiosis will cause half of the gametes to have abnormal chromosomes.
 
 

Since heterozygotes are much less fit than homozygotes, an inversion could only go to fixation in a very small population.

Example:  Plants of the genus Clarkia   (Evening Primrose Family)
    Localized species on the periphery of more widespread species show different chromosomal arrangements.

Molecular comparisons show that the localized species is often very genetically similar to the supposed "parent".

Individuals of Clarkia species can self fertilize, which would facilitate the spread of an inversion or translocation.

Once hybrids have low fitness, speciation in furthered by reinforcement, manifest here as changes in flowering time.

In general, plants can evolve to prevent interbreeding by:
Changing flowering time.
Changing flower morphology, to attract different polinators.
Preventing the pollen tube from developing.

Parapatric and Sympatric Speciation
An early idea was that disruptive selection alone could drive sympatric speciation.

Thoday & Gibson (1962) reported getting reproductive isolation in the lab as a result of simply imposing disruptive selection on fruit fly populations.
Subsequently, many attempts to replicate this result have failed; implying that it does not work.

Theory predicts that disruptive selection alone should not be enough.
Recall the case of selection on one locus with heterozygote inferiority (the simplest case of disruptive selection).  Here, disruptive selection causes one or the other allele to be lost, depending on where the population starts out.

As long as there is random mating, hybrids with intermediate (and therefore less fit) phenotypes keep being produced. This tends to drive rare phenotypes extinct, since most of their offspring will be hybrids. This is sometimes called being "mated to extinction".

Thus, we must combine disruptive selection with assortative mating.

If individuals with extreme phenotypes are both more fit and prefer mates with the same phenotype, then sympatric speciation is possible.

One way to achieve this is with gametic disequilibrium between a locus on which disruptive selection is occurring and another locus for which there is assortative mating (organisms like to mate with others that have the same genotype at that locus).
This is probably what was happening in Thoday & Gibson's original experiment.

The problem is, recombination is always breaking up the adaptive combinations.
Thus, the loci would have to be tightly linked in order for there to be enough time for selection to produce substantial reproductive isolation.

Another way to achieve such "linkage" is if organisms choose slightly different habitats to forage in and then tend to mate with others in the same habitat. This produces assortative mating by distance.

Example: In lab experiments, flies were allowed to distribute themselves in a "maze" with different conditions (light level, humidity, etc) in different places. Only those that chose extreme conditions were allowed to mate - doing so with others that also chose those environments. Offspring were then put back in the middle of the maze and allowed to distribute themselves.
After ~30 generations, reproductive isolation achieved.

We conclude that sympatric speciation is most likely when there is disruptive selection on a number of traits combined with differences in habitat use.

Traits that experience disruptive selection and also induce assortative mating are sometimes referred to as "magic" traits.

Example: Sticklebacks (Fish)
Many species of sticklebacks have entered fresh water from marine (which is ancestral). In British Columbia, many lakes have two distinct species of sticklebacks:

In each lake, these are reproductively isolated from one another.

However, benthic species from one lake can interbreed (in the lab) with benthics from other lakes. The same holds for limnetic species.

Since there are many such lakes, this indicates that reproductive isolation between the two species in any one lake is a result of adaptation to their different environments.

There appears to be sexual selection on body size, which would explain the pattern (large like to mate with large, small with small).

When hybrids are produced in the lab, they have intermediate morphology that is not well adapted to any habitat within these lakes, they are thus strongly selected against.

Speciation by Hybridization and Polyploidy
Common in plants.

Hybrids of separate species are generally infertile since their chromosomes do not come in homologous pairs, and thus can not pair properly in Meiosis.

Occasionally, a plant cell will double its DNA without dividing. In such a case, the cell becomes effectively tetraploid and each chromosome has a homologue.

If the plant can self fertilize (which many can), then it can produce an entire population of tetraploids in the next generation.
These are reproductively isolated from the parent species (different numbers of chromosomes) and are usually bigger.

Example: Galeopsis (in the Mint family) studied by Muntzing (1930).

Started with two species, G. speciosa and G. pubescens, each with 16 chromosomes (2n=16).
Through a series of hundreds of crosses, produced a tetraploid hybrid (2n=32) that closely resembled another species, G. tetrahit.
(Specifically, Muntzing crossed the two initial species and, from many offspring, got a single triploid. Backcrossing this to G. pubescens yielded a single tetraploid.)
The lab produced hybrid was not interfertile with the original parent species, but was interfertile with the naturally occurring tetraploid species. Jul 8, 2021