The value is comparable to, or perhaps slightly higher, than heterozygosity estimates of ostrich subspecies, which appears to be the only other ratite that has been studied with microsatellite analyses. The level of genetic diversity we recorded for Dinornis, combined with the discrimination power documented in the ��Probability of Identity�� analysis, suggest that these six markers a highly informative and suitable for population genetic studies of moa. Further analyses and interpretations of the genetic diversity are, however, not the scope of this paper and data will be presented elsewhere. A critical evaluation of the final data confirmed its integrity. One minor deviation from Hardy-Weinberg proportions was observed in the Moa_MA21 locus, but as these data refer to birds living over a span of.4000 years, they do not reflect randomly mating individuals at one point in time. Hardy-Weinberg proportions are not necessarily, GDC-0941 therefore, expected by default. Still, the general accordance with expected HW-proportions was an indication that allelic dropout had a minimal effect on the compiled data. Use of additional software, designed specifically to identify scoring problems owing to dropout and stuttering, did not reveal any issues. Lastly, we documented a link between the probability of dropout and CT value, but clearly rejected a negative correlation between observed individual heterozygosity and CT in the compiled data. This also provided strong support for the integrity of the data, indicating that DNA preservation had not affected the final results. We suggest, therefore, that on the basis of our proposed criteria, the moa microsatellite dataset is of high fidelity despite representing template molecules of c. 600 to 5000 years of age. We argue that variants of the applied methodology will be valid for most scenarios involving aDNA and microsatellites. We emphasise, however, the importance of Torin 1 assessing each case on its merits by including pilot studies and the generation of preliminary data to develop a specific strategy for the material at hand. The proposed Criterion 2 for example, could prove insufficient for types of data where false peaks are difficult to discriminate from true alleles or in situations where contamination is of greater concern. The achievement of generating a high qualitymicrosatellite dataset for an extinct species does not mean that there is no room for methodological or procedural improvement. A major drawback to the procedure presented here, is the considerable workload associated with single-plexed PCRs, repeated many times with DNA added from one tube at a time. A few experiments with two-plexes were, however, unsuccessful. Methods have been developed for improving microsatellite PCR results from degraded DNA e.g., and further research might elucidate whether our stringent setup can be relaxed somewhat, and whether these novel approaches can be used on aDNA templates without increasing the risk of cross-contamination.