First approaches for a combined use of microsatellites and pedigree data to estimate relationships.

The main objective of conservation is the preservation of genetic diversity. Among several tools recommended to measure genetic variability, relationships coefficients are often used. They can be estimated from pedigree but also from direct knowledge of the genotype. The Skyros pony is an example of a small breed concerned by the management of genetic variability. However, the analysis of the pedigree showed its very poor quality. A part of the population was therefore genotyped for 16 microstellites and two estimators of relationships were tested: the Lynch & Ritland (L&R) (anly based on genotypes and described as having the best performances for all population comositionsof the mammal species) and a new estimator (developed for this study and combining pedigree and microsatellite information). In order to compare the results, a Principal Component Analysis (PCA) was performed. The new estimator explained a higher percentage of information within the 3 principal factors of the PCA (41.99%) than the L&R estimator (20.61%). The graphic PCA representation showed a better separation between the reference group and the rest of the populationin the case of the new estimator. The new estimator showed interesting preliminary results and results support interest in combining information in case of incomplete pedigrees and/or use of a limited number of markers.