Empirical analysis of fish mating systems and the effect of microsatellite null alleles
Dakin, Elizabeth Erin
MetadataShow full item record
The use of highly polymorphic microsatellite markers has rapidly expanded the ability of scientists to evaluate the genetic mating systems of almost every sort of living organism. Fishes provide an excellent group in which to study mating systems due to the incredibly diverse array of tactics which they employ. Here I have used micrsatellite markers to empirically assess the mating systems of two species of fish, the straight-nosed pipefish, Nerophis ophidion, and the pirate perch, Aphredoderus sayanus. The pipefish is of interest to behavioral ecologists because it is said to be sex role reversed, meaning that sexual selection acts more strongly on the females and that males are the sex that is more choosy about who to mate with. Genetic testing of eggs confirmed that each male carries the eggs of a single female at one time, and suggests that females might mate with multiple males during the same time period. In the pirate perch, very little was previously known about their spawning behavior in the wild. By genotyping pirate perch offspring, we were able to determine that the large numbers of adult pirate perch seen swarming around nesting sites were most likely all there to find mates, and not to consume the young. The final section of this dissertation involves microsatellite null alleles, a potential drawback to parentage studies. Null alleles are caused by a mutation in the priming region of the primers used to amplify a microsatellite locus, causing no product to be formed. I have reviewed 234 articles in which null alleles are either found or are suspected, and I report on the different methods used to deal with them. I also carried out a series of computer simulations in order to determine whether null alleles can impact the calculation of the exclusion probability, an important statistic in parentage studies. Fortunately, under realistic null allele frequencies, there seems to be little change in the exclusion probability.