|
Model | Process driving divergence | Migration and gene flow | Discrete patches (populations) | Method | Notes |
|
[16] | Divergent selection | Yes | Clinal divergence rather than discrete patches | Analytical | Divergent selection is on major locus, modifier locus affects the fitness of alleles at the major locus, but is not under divergent selection itself, barrier to neutral gene flow not examined. Also, focus of the model is how the modifier, once fixed, alters the shape of the cline, rather than on how the modifier fixes in the first place. |
[18] | Divergent selection | Yes | Clinal divergence rather than discrete patches | Analytical | Divergent selection is on major locus, modifier locus causes DMI but is not under divergent selection itself, barrier to neutral gene flow not examined. |
[27] | Stochastic (drift) | No | Yes | Simulation | Drift fixed new mutations in small populations; two loci considered. |
[28] | Unspecified | No | Yes | Analytical | The classic paper on the accumulation of DMIs. No assumptions are made about evolutionary causes of substitutions driving DMIs. |
[29, 30] | Uniform selection, not explicitly divergent | Yes | Yes | Analytical | Migration can allow the spread of a gene combination whose component alleles are not individually favored, but the conditions for this are restricted (low migration, strong advantage for particular genotypic combinations). |
[31] | Uniform selection or drift | No | Yes | Analytical and simulation | Population subdivision has no effect on time to speciation by drift. Under selection, speciation occurs more rapidly between two large populations than between many small ones. |
[32] | Unspecified; populations initially fixed for different incompatible alleles | Yes | Yes, stepping stone | Analytical | Focused on the effect of intrinsic incompatibilities on the barrier to neutral gene flow. |
[33] | Unspecified | No | Yes | Analytical | Adds stochasticity of molecular evolution to the results of Orr [28]. |
[34] | Uniform selection | Yes | Yes, between neighboring patches | Simulation | Speciation in subdivided populations occurs most rapidly when there is some migration (which spreads incompatible alleles); many (250) loci are considered. |
[35] | Uniform selection | Yes | Yes | Simulation | Strong selection counters the inhibitory effects of gene flow, permitting the evolution of intrinsic postmating isolation. |
[36] | Uniform selection | Yes | Yes, | Analytical | Migration occurs through a single, spatially structured population and affects the accumulation of DMIs. |
[37] | Uniform selection | Yes | Yes | Analytical | Loci causing DMIs themselves are under selection (i.e., new alleles at these loci can be favored), but selection is uniform. Barrier to gene flow is considered, including the effect of chromosomal inversions. |
(Gavrilets [14]) and previous models by same author (e.g., [38, 39]) | Divergent selection, uniform selection, or drift | Yes | Yes | Analytical | Incompatible alleles advantageous in one environment but neutral in other (thus wider range of scenarios examined in current study), barrier to gene flow examined. |
[40] | Uniform selection | Yes | No, clinal with IBD | Simulation | IBD and outbreeding depression were sufficient to drive speciation, in the absence of ecological differences. |
[41] | Uniform selection | Yes | Yes | Simulation | Focused on the origin of DMIs due to different, incompatible mutations diverging between populations adapting to identical selection regimes. Gene flow strongly impeded strong divergence via such a “mutation-order” process (c.f., Schluter [26]) |
Current study | Divergent selection | Yes | Yes | Analytical and simulation | Focus on the origination of incompatibilities in the face of gene flow. Examined scenarios with and without direct, divergent selection on modifier loci which also contribute to intrinsic isolation. Examined barrier to neutral gene flow caused by evolution of intrinsic isolation. |
|