Impact of different panels of microsatellite loci, different numbers of loci, sample sizes, and gender ratios on population genetic results in red deer

Population genetic parameters from different studies might be significantly influenced by differences in sample size, fraction of males and females, marker number, and sets of markers used, reducing the comparability between studies. This hypothesis was tested on a red deer population of 205 individuals with an estimated size of 1000 animals. Four tests were performed: (1) the population was subdivided into 10 populations each with 10 to 150 individuals and genotyped with 16 markers, (2) the total population was genotyped 10 times with different panels of microsatellite loci containing 2 to 14 markers, (3) a subset of 8 microsatellite loci was used to genotype the total population; markers of this subset were replaced one by one with a different marker set and genotyping results were compared to the results of the original subset and (4) the effect of sex was estimated. Additionally, 24 references from literature, including 256 European red deer populations, were analyzed. A median of 25 individuals per population was investigated in published studies using 11 microsatellite markers (5 to 22). Sixty-eight percent of possible study comparisons matched with less than 10% of microsatellite loci. Our results show that the factors investigated, except for the factor gender, lead to significant deviations in the population genetic results, especially with sample sizes below 30, with less than 6 to 8 microsatellite markers and with the use of different panels of microsatellite loci. This is also true with respect to population genetic structure and the use of Bayesian methods. Therefore, populations from different studies should be compared with each other with caution.