Table 1.
Description and source of publicly available outgroup data on 274 individuals across 11 breeds.
Fig 1.
Average linkage disequilibrium (LD) decay in Angler (RVA), Red-and-White dual-purpose (RDN) and Red Holstein (RH).
Fig 2.
Trends in LD based effective population size for Angler (RVA), Red-and-White dual-purpose (RDN) and Red Holstein (RH).
A) Across 1000 generation and B) across 52 generations ago.
Table 2.
Number of animals (N), average genomic relationship (Ø gRel), average observed heterozygosity (Ho) and most recent effective population size (Ne5Gen) for Angler (RVA), Red-and-White dual-purpose (RDN) and Red Holstein (RH).
Fig 3.
Frequency distribution of the number of ROH in different length classes for Angler (RVA), Red-and-White dual-purpose (RDN) and Red Holstein (RH).
Table 3.
Number of animals with and without identified ROH, total and average number of ROH per breed and average sum of ROH segment lengths for Angler (RVA), Red-and-White dual-purpose (RDN) and Red Holstein (RH).
Table 4.
Mean inbreeding coefficients calculated from ROH with minimum length of 4 (FROH>4), 8 (FROH>8), and 16 (FROH>16) Mb, from the excess of homozygosity (FHOM) and from pedigree (FPED) for Angler (RVA), Red-and-White dual-purpose (RDN) and Red Holstein (RH).
Fig 4.
Regression plot of FROH>4Mb (A), FROH>8Mb (B) and FROH>16Mb (C) on FPED (left) and on FHOM (right) for Angler (green) and Red-and-White dual-purpose (blue). The broken line (black) is a regression line for both breeds and corresponds to the regression equation presented.
Fig 5.
Principal component analysis of 14 cattle breeds distinguished by shape and colour.
Based on 19,971 SNPs, the first four components together accounted for 61.7% of variation in the data.
Fig 6.
Admixture graphs for k = 2, 5, 9, 13 and 19 with the optimal cluster level marked with an arrow.
The amount of a colour in a cluster reflects a breed’s proportion of genetic variation originating from that colour. Breed names are coloured according to geographical origin: Germany (red), Northern Europe (grey), Norway (green), Channel Islands (dark red), Switzerland (blue), France (purple), Scotland (brown) and Burkina Faso (orange).
Fig 7.
Visualization of high-resolution population networks for 14 cattle breeds.
K-NN = 10 (A) investigated small-scale structures, K-NN = 100 (B) targeted large-scale structures and breeds are distinguished by node colour.
Fig 8.
Network visualisation of 14 cattle breeds effectively separated along geographical lines with K-NN = 35.
Admixture results at k = 19 were included in the network construction (in Box) such that node colouration reflects individual level of admixture.