Usefulness of molecular markers for detecting population bottlenecks and monitoring genetic change

Population bottlenecks are important to detect and avoid because they can reduce adaptive potential and increase risk of population extinction. I have used computer simulations and population genetic data to develop and evaluate tests for (i) detecting recent historical bottlenecks and (ii) monitoring to detect contemporary bottlenecks.; First, new tests for recent historical bottlenecks detected bottlenecks at a rate of 50-80% when using either allozyme or microsatellite data from 21 natural populations known to have been recently bottlenecked. These tests seldom mistakenly identified non-bottlenecked populations as bottlenecked populations (1-7% error rate) when applied to data from 72 natural populations. Simulations suggested that at least 10-13 polymorphic loci are required to achieve an 80% probability of detecting a recent historical bottleneck of size 20 (N{dollar}rmsb{lcub}e{rcub}{dollar} = 20) when 30 individuals are sampled. Smaller bottlenecks are likely to be detected with as few as six loci and 20 individuals. Power may be reduced slightly by extremely skewed sex ratios or post-bottleneck population growth.; Second, the following six statistical tests for monitoring to detect bottlenecks were evaluated: a test for loss of heterozygosity, two tests for loss of alleles, two tests for change in the distribution of allele frequencies, and the temporal method for estimating Ne from temporal change in allele frequencies. The temporal method was the most powerful; it provided an 85% probability of detecting a one-generation bottleneck of 10 breeding individuals when sampling 30 individuals and monitoring only five microsatellite loci (i.e., loci with 4-5 alleles). This represents nearly a ten-fold increase in power over the test for loss of heterozygosity, and it allows for detection of bottlenecks before 5% of a population's heterozygosity has been lost. The temporal method differentiated between bottlenecks of effective size 10 and size 40 when monitoring 10-15 microsatellite loci and sampling 30-60 individuals. Tests for loss of alleles were second most powerful, but had reduced power for detecting bottlenecks caused by skewed sex ratios.; This research develops and evaluates bottleneck-detection tests and population monitoring methods that are critically needed for conservation of animal and plant populations throughout the world. Guidelines are provided for applying the methods and for the number of loci and individuals required to achieve reasonable statistical power.