A Spatially Explicit Individual-Based Model of Reinforcement in Hybrid Zones

Abstract

An individual-based model consisting of two dioecious populations in a two-dimensional environmental grid was constructed. Each population began with, and never exceeded, 1000 individuals; extinction was allowed. Genomes consisting of 30 biallelic loci for male sexual advertisement call, female mate preference, and population origin were constructed, and lineages of each individual in the starting populations were followed for 2000 generations. Type and level of hybrid disadvantage, initial population distribution, patchiness of environmental resources, and level of mate choice were varied. Persistence of bimodal hybrid zones was nonexistent at low levels of hybrid disadvantage and universal at high levels of hybrid disadvantage, with a narrow threshold in which persistence was unpredictable. Persistence occurred at lower levels of hybrid disadvantage when populations were initially parapatric rather than sympatric, and environments were patchy rather than homogeneous. Increased divergence in mating systems occurred when hybrid disadvantage was high, hybrids were infertile, populations were initially parapatric, and increased female choice was allowed. Mating system divergence was much higher in interacting populations compared with noninteracting populations, indicating that reinforcement caused most of the observed divergence. When hybrids were infertile, reinforcement contributed to speciation, because under hybrid infertility the probability of persistence at low levels of hybrid disadvantage was positively related to mate choice. The results agree with previous one-dimensional spatial models in finding that population persistence is more likely in parapatric and patchy population distributions. In addition, the results show that hybrid infertility may facilitate the process of reinforcement and speciation.

Genetic monitoring of two decades of hybridization between allis shad (Alosa alosa) and twaite shad (Alosa fallax)

“…It is possible hybrids are at a mating disadvantage through sexual selection against hybrid phenotypes (Sadenin and Littlejohn 2003), a concept shown in three-spined sticklebacks where hybrid males were outcompeted in mesocosm experiments by purebred limnetic males for territorial space preferred by the limnetic females (Vamosi and Schluter 1999). If shad species use different breeding habitats within the Solway Firth system, it is plausible that hybrid males could be similarly disadvantaged through competition for mates.…”

Optimal Annual Routines: New Tools for Conservation Biology

“…Beyond this, it is difficult to generalize about IBMs, because the behaviors of interest will vary according to the application of the model. For example, if the model is aimed at assessing the consequences of assortative mating (e.g., Sadedin and Littlejohn 2003), the key behaviors might include matechoice and mate-search rules. By contrast, if the purpose of the model is to assess habitat use (e.g., Sutherland and Allport 1994), the key behavior might involve movement to maximize intake rate.…”

An agent-based computational model of the evolution of reproductive isolation in populations of digital plants

“…These models may also be used to test hypotheses about the underlying mechanisms of a biological system (e.g. Orr 1991, Sadedin andLittlejohn 2003). This chapter explores various aspects and methods of computational models.…”

“…These are easily analyzed using traditional equations of population genetics. More recent models involved numerous loci per trait (Sadedin and Littlejohn, 2003;Kaneko, 2002;Dieckinann and Doebeli, 1999;Gavrilets, 1999;Kondrashov and Kondrashov, 1999). Forming and solving a system of equations grows increasingly complex as more loci are added.…”

“…With rare exception, individual based models of speciation implement the genetics as sets of biallelic (typically "zero" and "one") loci. The traits are frequently quantitative (Rice, 1984;Doebeli, 1996;Dieckmann and Doebeli, 1999;Doebeli and Dieckmann, 2000;Gavrilets et al, , 2000Sadedin and Littlejohn, 2003) although there are models implementing the Hamming distance (number of allelic mismatches across loci of two binary strings) (Kondrashov and Shpak, 1998;Gavrilets, 1999), dominance (Orr, 1991), epistasis (Buerkle et al, 2000), or combining methods of phenotypic computation (Kondrashov and Mina, 1986;Kondrashov and Kondrashov, 1999). There are other, more abstract implementations as well, such as Kaneko (2002) and Wilke et al (2001) where the phenotype is determined by a set of equations or operations that make up the genotype of the individual.…”