Estimating the optimal number of samples to determine the effective population size in livestock

Arianna Manunza^1*Paolo Cozzi¹Paul Boettcher²Ino Curik^3,4Christian Looft⁵Licia Colli^6,7Johann Sölkner⁸Gábor Mészáros⁸Alessandra Stella¹

• ¹Institute of Agricultural Biology and Biotechnology, National Research Council (CNR), Milan, Italy
• ²Food and Agriculture Organization of the United Nations (Italy), Rome, Lazio, Italy
• ³Department of Animal Science, University of Zagreb, Faculty of Agriculture, Zagreb, Croatia, Zagreb, Croatia
• ⁴Institute of Animal Sciences, Hungarian University of Agriculture and Life Sciences, Kaposvár, Hungary
• ⁵Neubrandenburg University of Applied Sciences, Brodaer Str. 2, Neubrandenburg, Germany, Neubrandenburg, Germany
• ⁶Dipartimento di Scienze Animali, della Nutrizione e degli Alimenti Università Cattolica del S. Cuore, via Emilia Parmense 84, 29122 Piacenza (PC), Italy, Piacenza, Italy
• ⁷BioDNA Centro di Ricerca sulla Biodiversità e sul DNA Antico, Università Cattolica del S. Cuore, via Emilia Parmense 84, 29122 Piacenza (PC), Italy, Piacenza, Italy
• ⁸Institute of Livestock Sciences, BOKU University, Gregor-Mendel-Straße 33, 1180 Vienna, Austria, Vienna, Austria

Effective population size (Ne) is a key parameter in various biological disciplines, including evolutionary biology, conservation genetics and livestock breeding programs. When applying genomic approaches to estimate Ne or other indicators of genetic variation, sample size is among the critical factors that directly affect the balance between cost and precision. In this study, we investigated the impact of sample size on Ne estimates by analysing data from previous genotyping studies and simulations. Our results suggest that a sample size of 50 animals is a reasonable approximation of the "true" ("unbiased") Ne value within the populations analysed. While estimating the Ne value is an important starting point in population genetics, additional factors -such as degree of inbreeding, population structure and admixture -must be taken into account to obtain a comprehensive genetic evaluation and avoid misinterpretation. We conclude that linkage disequilibrium (LD)-based approaches are well suited for the estimation of Ne in livestock populations. However, careful interpretation of results is essential as current bioinformatics tools may introduce potential biases due to methodological assumptions, marker density or population-specific factors.