Population structure and genomic inbreeding in nine Swiss dairy cattle populations

Abstract

BackgroundDomestication, breed formation and intensive selection have resulted in divergent cattle breeds that likely exhibit their own genomic signatures. In this study, we used genotypes from 27,612 autosomal single nucleotide polymorphisms to characterize population structure based on 9214 sires representing nine Swiss dairy cattle populations: Brown Swiss (BS), Braunvieh (BV), Original Braunvieh (OB), Holstein (HO), Red Holstein (RH), Swiss Fleckvieh (SF), Simmental (SI), Eringer (ER) and Evolèner (EV). Genomic inbreeding (F ROH) and signatures of selection were determined by calculating runs of homozygosity (ROH). The results build the basis for a better understanding of the genetic development of Swiss dairy cattle populations and highlight differences between the original populations (i.e. OB, SI, ER and EV) and those that have become more popular in Switzerland as currently reflected by their larger populations (i.e. BS, BV, HO, RH and SF).ResultsThe levels of genetic diversity were highest and lowest in the SF and BS breeds, respectively. Based on F ST values, we conclude that, among all pairwise comparisons, BS and HO (0.156) differ more than the other pairs of populations. The original Swiss cattle populations OB, SI, ER, and EV are clearly genetically separated from the Swiss cattle populations that are now more common and represented by larger numbers of cows. Mean levels of F ROH ranged from 0.027 (ER) to 0.091 (BS). Three of the original Swiss cattle populations, ER (F ROH: 0.027), OB (F ROH: 0.029), and SI (F ROH: 0.039), showed low levels of genomic inbreeding, whereas it was much higher in EV (F ROH: 0.074). Private signatures of selection for the original Swiss cattle populations are reported for BTA4, 5, 11 and 26.ConclusionsThe low levels of genomic inbreeding observed in the original Swiss cattle populations ER, OB and SI compared to the other breeds are explained by a lesser use of artificial insemination and greater use of natural service. Natural service results in more sires having progeny at each generation and thus this breeding practice is likely the major reason for the remarkable levels of genetic diversity retained within these populations. The fact that the EV population is regionally restricted and its small census size of herd-book cows explain its high level of genomic inbreeding.Electronic supplementary materialThe online version of this article (10.1186/s12711-017-0358-6) contains supplementary material, which is available to authorized users.

Assessing genomic diversity and signatures of selection in Original Braunvieh cattle using whole-genome sequencing data

“…The minor contribution of long ROH to the genomic inbreeding indicates that recent inbreeding is relatively low in OB possibly due to use of many sires in natural matings. Our results based on ROH inferred from whole-genome sequencing variants corroborate that genomic inbreeding is lower in OB than most mainstream breeds [25]. However, comparing the number and size distribution of ROH across studies is difficult and subject to bias because misplaced genomic segments might break ROH into multiple small-and medium-sized ROHs and different ROH-detection approaches yield results that are not readily comparable [29,34,56,86,87].…”

“…The relatively high genomic diversity in OB cattle is assumed to be the result of many different sires contributing to the gene pool due to frequent use of natural mating [25]. Our whole-genome sequencing based estimate of genomic inbreeding (0.14) is substantially higher than previous estimates obtained using either 50K SNP microarray data (F ROH = 0.029, [25]) or pedigree records (F PED =0.02, [24]) for the same population. Genotype data obtained using SNP microarrays with medium density (e.g., BovineSNP50) can only capture long ROH (>1 Mb) due to lower SNP density (~1 SNP per 50 kb) and underestimates the genomic inbreeding, particularly when short and medium-sized ROH contribute substantially to genomic inbreeding as it is the case in our study.…”

“…However, genomic inbreeding is slightly higher in OB than New Red Danish cattle (0.11), an admixed breed that contains genes from old Danish and other red breeds [84]. The relatively high genomic diversity in OB cattle is assumed to be the result of many different sires contributing to the gene pool due to frequent use of natural mating [25]. Our whole-genome sequencing based estimate of genomic inbreeding (0.14) is substantially higher than previous estimates obtained using either 50K SNP microarray data (F ROH = 0.029, [25]) or pedigree records (F PED =0.02, [24]) for the same population.…”

“…In our data, short and medium-sized ROH accounted for 80.48% of total inbreeding. Most short and medium-sized ROH are not reliably detectable with SNP microarrays used in Signer-Hasler et al [25], resulting in a massive underestimation of genomic inbreeding. Our estimate of the genomic inbreeding using WGS variants also includes short and medium-sized ROH that were previously missed using SNP array data, thus representing a realistic estimate of total genomic inbreeding.…”

“…Since then, the Swiss OB population is slowly but steadily growing due to governmental subsidies. In spite of the small population size, pedigree-and microarray-derived estimates of population genetic parameters indicate that genetic diversity is higher in OB than many mainstream breeds likely due to the use of many sires in natural mating [24,25].…”