Limits...
Spatially Heterogeneous Environmental Selection Strengthens Evolution of Reproductively Isolated Populations in a Dobzhansky – Muller System of Hybrid Incompatibility

View Article: PubMed Central - PubMed

ABSTRACT

Within-species hybrid incompatibility can arise when combinations of alleles at more than one locus have low fitness but where possession of one of those alleles has little or no fitness consequence for the carriers. Limited dispersal with small numbers of mate potentials alone can lead to the evolution of clusters of reproductively isolated genotypes despite the absence of any geographical barriers or heterogeneous selection. In this paper, we explore how adding heterogeneous natural selection on the genotypes (e.g., gene environment associations) that are involved in reproductive incompatibility affects the frequency, size and duration of evolution of reproductively isolated clusters. We conducted a simulation experiment that varied landscape heterogeneity, dispersal ability, and strength of selection in a continuously distributed population. In our simulations involving spatially heterogeneous selection, strong patterns of adjacency of mutually incompatible genotypes emerged such that these clusters were truly reproductively isolated from each other, with no reproductively compatible “bridge” individuals in the intervening landscape to allow gene flow between the clusters. This pattern was strong across levels of gene flow and strength of selection, suggesting that even relatively weak selection acting in the context of strong gene flow may produce reproductively isolated clusters that are large and persistent, enabling incipient speciation in a continuous population without geographic isolation.

No MeSH data available.


Three-dimensional histograms of changes in the mean duration of reproductively isolated clusters of individuals (in generations; row 1, A–C), mean number of reproductively isolated clusters (row 2, D–F), and mean size of reproductively isolated clusters (individuals; row 3, G–I). Columns in the figure represent different levels of landscape aggregation of the two habitat types involved in environmental section of the genotypes contributing to reproductive isolation (column 1, A,D,G is H1, highly heterogeneous; column 2, B,E,H is H5, intermediate heterogeneity; column 3, C,F,I is H9, high aggregated patterns of the two habitat types). The 6 × 6 parameter space in each subfigure shows the combinations of six levels of dispersal (D3—3% of landscape extent, D5—5% of landscape extent, D10—10% of landscape extent, D15—15% of landscape extent, D25—25% of landscape extent, D50—50% of landscape extent) across six levels of selection (S2—2% difference in relative fitness of genotypes aabb and AABB in each of the two habitats, S4—4% difference in relative fitness, S8—8% difference in relative fitness, S16—16% difference in relative fitness, S32—32% difference in relative fitness, S64—64% difference in relative fitness). See Supplementary Videos for these histograms as they change through time.
© Copyright Policy
Related In: Results  -  Collection

License
getmorefigures.php?uid=PMC5121238&req=5

Figure 2: Three-dimensional histograms of changes in the mean duration of reproductively isolated clusters of individuals (in generations; row 1, A–C), mean number of reproductively isolated clusters (row 2, D–F), and mean size of reproductively isolated clusters (individuals; row 3, G–I). Columns in the figure represent different levels of landscape aggregation of the two habitat types involved in environmental section of the genotypes contributing to reproductive isolation (column 1, A,D,G is H1, highly heterogeneous; column 2, B,E,H is H5, intermediate heterogeneity; column 3, C,F,I is H9, high aggregated patterns of the two habitat types). The 6 × 6 parameter space in each subfigure shows the combinations of six levels of dispersal (D3—3% of landscape extent, D5—5% of landscape extent, D10—10% of landscape extent, D15—15% of landscape extent, D25—25% of landscape extent, D50—50% of landscape extent) across six levels of selection (S2—2% difference in relative fitness of genotypes aabb and AABB in each of the two habitats, S4—4% difference in relative fitness, S8—8% difference in relative fitness, S16—16% difference in relative fitness, S32—32% difference in relative fitness, S64—64% difference in relative fitness). See Supplementary Videos for these histograms as they change through time.

Mentions: To explore the main effects and the predominant interaction between landscape heterogeneity and dispersal we produced histograms in a dispersal × selection space, across the three levels of landscape heterogeneity (Figures 2A–C; Supplementary Video S1 duration.avi). These charts illustrate two main patterns. First, reproductively isolated clusters persist for the entire simulation time when dispersal is low and environmental selection is high. Second, the duration of reproductively isolated clusters increases across levels of dispersal and selection as landscapes become less heterogeneous. For example, at H1, the most heterogeneous configuration, reproductively isolated clusters persist for the full simulation time at combinations of dispersal between 3 and 5% and selection levels of 32 or 64 (Figure 2A; Supplementary Video S1 duration.avi). At the H5 level of heterogeneity, reproductively isolated clusters persist for the full simulation time for dispersal 3% when selection is 8 or above, at dispersal 5% when selection is 16 or above, at 10% dispersal when selection is 32 or above, and at dispersal 25% when selection is 64. The pattern continues at the highest level of aggregation, H9, when clusters have duration across the full extent of the simulation time or nearly the full extent for all combinations of dispersal and selection producing clusters (diagonal across dispersal-selection space from D3 to S64).


Spatially Heterogeneous Environmental Selection Strengthens Evolution of Reproductively Isolated Populations in a Dobzhansky – Muller System of Hybrid Incompatibility
Three-dimensional histograms of changes in the mean duration of reproductively isolated clusters of individuals (in generations; row 1, A–C), mean number of reproductively isolated clusters (row 2, D–F), and mean size of reproductively isolated clusters (individuals; row 3, G–I). Columns in the figure represent different levels of landscape aggregation of the two habitat types involved in environmental section of the genotypes contributing to reproductive isolation (column 1, A,D,G is H1, highly heterogeneous; column 2, B,E,H is H5, intermediate heterogeneity; column 3, C,F,I is H9, high aggregated patterns of the two habitat types). The 6 × 6 parameter space in each subfigure shows the combinations of six levels of dispersal (D3—3% of landscape extent, D5—5% of landscape extent, D10—10% of landscape extent, D15—15% of landscape extent, D25—25% of landscape extent, D50—50% of landscape extent) across six levels of selection (S2—2% difference in relative fitness of genotypes aabb and AABB in each of the two habitats, S4—4% difference in relative fitness, S8—8% difference in relative fitness, S16—16% difference in relative fitness, S32—32% difference in relative fitness, S64—64% difference in relative fitness). See Supplementary Videos for these histograms as they change through time.
© Copyright Policy
Related In: Results  -  Collection

License
Show All Figures
getmorefigures.php?uid=PMC5121238&req=5

Figure 2: Three-dimensional histograms of changes in the mean duration of reproductively isolated clusters of individuals (in generations; row 1, A–C), mean number of reproductively isolated clusters (row 2, D–F), and mean size of reproductively isolated clusters (individuals; row 3, G–I). Columns in the figure represent different levels of landscape aggregation of the two habitat types involved in environmental section of the genotypes contributing to reproductive isolation (column 1, A,D,G is H1, highly heterogeneous; column 2, B,E,H is H5, intermediate heterogeneity; column 3, C,F,I is H9, high aggregated patterns of the two habitat types). The 6 × 6 parameter space in each subfigure shows the combinations of six levels of dispersal (D3—3% of landscape extent, D5—5% of landscape extent, D10—10% of landscape extent, D15—15% of landscape extent, D25—25% of landscape extent, D50—50% of landscape extent) across six levels of selection (S2—2% difference in relative fitness of genotypes aabb and AABB in each of the two habitats, S4—4% difference in relative fitness, S8—8% difference in relative fitness, S16—16% difference in relative fitness, S32—32% difference in relative fitness, S64—64% difference in relative fitness). See Supplementary Videos for these histograms as they change through time.
Mentions: To explore the main effects and the predominant interaction between landscape heterogeneity and dispersal we produced histograms in a dispersal × selection space, across the three levels of landscape heterogeneity (Figures 2A–C; Supplementary Video S1 duration.avi). These charts illustrate two main patterns. First, reproductively isolated clusters persist for the entire simulation time when dispersal is low and environmental selection is high. Second, the duration of reproductively isolated clusters increases across levels of dispersal and selection as landscapes become less heterogeneous. For example, at H1, the most heterogeneous configuration, reproductively isolated clusters persist for the full simulation time at combinations of dispersal between 3 and 5% and selection levels of 32 or 64 (Figure 2A; Supplementary Video S1 duration.avi). At the H5 level of heterogeneity, reproductively isolated clusters persist for the full simulation time for dispersal 3% when selection is 8 or above, at dispersal 5% when selection is 16 or above, at 10% dispersal when selection is 32 or above, and at dispersal 25% when selection is 64. The pattern continues at the highest level of aggregation, H9, when clusters have duration across the full extent of the simulation time or nearly the full extent for all combinations of dispersal and selection producing clusters (diagonal across dispersal-selection space from D3 to S64).

View Article: PubMed Central - PubMed

ABSTRACT

Within-species hybrid incompatibility can arise when combinations of alleles at more than one locus have low fitness but where possession of one of those alleles has little or no fitness consequence for the carriers. Limited dispersal with small numbers of mate potentials alone can lead to the evolution of clusters of reproductively isolated genotypes despite the absence of any geographical barriers or heterogeneous selection. In this paper, we explore how adding heterogeneous natural selection on the genotypes (e.g., gene environment associations) that are involved in reproductive incompatibility affects the frequency, size and duration of evolution of reproductively isolated clusters. We conducted a simulation experiment that varied landscape heterogeneity, dispersal ability, and strength of selection in a continuously distributed population. In our simulations involving spatially heterogeneous selection, strong patterns of adjacency of mutually incompatible genotypes emerged such that these clusters were truly reproductively isolated from each other, with no reproductively compatible “bridge” individuals in the intervening landscape to allow gene flow between the clusters. This pattern was strong across levels of gene flow and strength of selection, suggesting that even relatively weak selection acting in the context of strong gene flow may produce reproductively isolated clusters that are large and persistent, enabling incipient speciation in a continuous population without geographic isolation.

No MeSH data available.