Population analysis of the Borzoi breed Genetic analysis of the Kennel Club pedigree records of the UK Borzoi population has been carried out with the aim of estimating the rate of loss of genetic diversity within the breed and providing information to guide a future sustainable breeding strategy. The population statistics summarised provide a picture of trends in census size, the number of animals used for breeding, the rate of inbreeding and the estimated effective population size. The rate of inbreeding and estimated effective population size indicate the rate at which genetic diversity is being lost within the breed. The analysis also calculates the average relationship (kinship) among all individuals of the breed born per year and is used to determine the level of inbreeding that might be expected if matings were made among randomly selected dogs from the population (the expected rate of inbreeding). Summary of results The analysis utilises the complete computerised pedigree records for the current UK Kennel Club registered Borzoi population, and statistics were calculated for the period 1980-2014. 1
Figure 1: a plot of number of registrations by year of birth, indicative of any changing trend in popularity of the breed, followed by the yearly trend in number of animals registered (and 95% confidence interval). Breed: Borzoi Figure 1: Number of registrations by year of birth Trend of registrations over year of birth (1980-2014) = -4.73 per year (with a 95% confidence interval of -5.59 TO -3.86). 2
Table 1: census statistics by year, including sire use statistics. Table 1: by year (1980-2014), the number of registered puppies born, by the number of unique dams and sires; maximum, median, mode, mean and standard deviation of number of puppies per sire; and the percentage of all puppies born to the most prolific 50%, 25%, 10% and 5% of sires. year #born #dams #sires puppies per sire %puppies sired by most prolific sires max median mode mean sd 50% sires 25% sires 10% sires 5% sires 1980 296 69 45 29 5 1 6.58 6.56 81.42 55.07 35.81 19.26 1981 250 59 48 18 4 1 5.21 4.47 80.8 55.2 30.8 14 1982 245 60 44 20 5 4 5.57 4.06 75.1 47.76 24.49 15.92 1983 213 39 28 21 7 1 7.61 5.14 74.65 47.89 24.88 9.86 1984 264 61 43 23 5 1 6.14 4.96 78.41 52.27 27.65 16.29 1985 189 43 34 14 5 1 5.56 3.92 79.37 50.79 21.16 14.81 1986 246 49 38 19 5.5 5 6.47 4 72.76 47.56 24.8 14.23 1987 229 45 36 25 5 3 6.36 4.74 74.24 48.03 29.26 18.34 1988 247 47 27 39 7 7 9.15 8.39 78.54 56.68 34.01 15.79 1989 197 33 22 24 7 7 8.95 6.24 73.6 51.78 24.37 12.18 1990 221 45 26 33 6 4 8.5 7.63 79.19 58.82 34.84 14.93 1991 177 41 30 20 5 2 5.9 4.59 77.4 53.11 27.68 21.47 1992 218 41 28 31 6 1 7.79 6.46 77.52 52.29 29.36 14.22 1993 220 43 33 13 7 8 6.67 3.06 70.45 37.73 15.91 10.91 1994 151 27 25 15 6 6 6.04 3.55 74.17 42.38 25.17 9.93 1995 170 36 26 21 5.5 5 6.54 4.7 74.12 51.18 29.41 12.35 1996 204 40 27 18 8 8 7.56 4.77 76.96 47.55 23.53 8.82 1997 197 38 27 22 7 9 7.3 5.16 77.66 49.75 27.41 11.17 1998 160 35 31 16 5 5 5.16 3.29 73.13 47.5 23.75 17.5 1999 213 45 30 20 7 1 7.1 4.92 75.59 49.77 24.88 18.31 2000 144 29 22 20 6 1 6.55 5.36 80.56 54.86 25 13.89 2001 184 42 36 36 3.5 1 5.11 6.39 87.5 60.33 38.59 27.72 2002 99 21 21 10 5 1 4.71 3.24 81.82 43.43 19.19 10.1 2003 159 37 30 25 2.5 1 5.3 5.64 88.05 63.52 32.7 24.53 2004 173 33 29 19 6 1 5.97 4.33 78.61 46.24 26.01 10.98 2005 133 30 27 18 5 1 4.93 3.95 82.71 51.13 27.82 13.53 2006 136 33 30 12 3.5 1 4.53 3.43 82.35 54.41 23.53 16.91 2007 126 31 28 14 3.5 1 4.5 3.64 83.33 53.97 26.98 11.11 2008 160 40 38 11 4 1 4.21 3.2 82.5 53.13 25 13.75 2009 102 25 23 13 5 1 4.43 3.36 83.33 50.98 22.55 12.75 2010 121 29 27 26 2 1 4.48 5.29 87.6 61.98 38.02 21.49 2011 84 25 21 12 2 1 4 3.74 84.52 59.52 28.57 14.29 2012 118 33 29 13 4 1 4.07 3.42 86.44 53.39 27.97 11.02 2013 127 31 28 22 3 1 4.54 4.61 85.04 57.48 33.86 17.32 2014 82 22 22 9 3.5 1 3.73 2.59 79.27 52.44 20.73 10.98 3
Generation interval: the mean average age (in years) of parents at the birth of offspring which themselves go on to reproduce. Mean generation interval (years) = 4.35 Figure 2: a plot of the annual mean observed inbreeding coefficient (showing loss of genetic diversity), and mean expected inbreeding coefficient (from random mating ) over the period 1980-2014. Expected inbreeding is staggered by the generation interval and, where >2000 animals are born in a single year, the 95% confidence interval is indicated. Figure 2: Annual mean observed and expected inbreeding coefficients 4
Estimated effective population size: the rate of inbreeding (slope or steepness of the observed inbreeding in Figure 2) is used to estimate the effective population size of the breed. The effective population size is the number of breeding animals in an idealised, hypothetical population that would be expected to show the same rate of loss of genetic diversity (rate of inbreeding) as the breed in question. It may be thought of as the size of the gene pool of the breed. Below an effective population size of 100 (inbreeding rate of 0.50% per generation) the rate of loss of genetic diversity in a breed/population increases dramatically (Food & Agriculture Organisation of the United Nations, Monitoring animal genetic resources and criteria for prioritization of breeds, 1992). An effective population size of below 50 (inbreeding rate of 1.0% per generation) indicates the future of the breed many be considered to be at risk (Food & Agriculture Organisation of the United Nations, Breeding strategies for sustainable management of animal genetic resources, 2010). Where the rate of inbreeding is negative (implying increasing genetic diversity in the breed), effective population size is denoted n/a. Estimated effective population size = 918.8 NB - this estimate is made using the rate of inbreeding over the whole period 1980-2014 5
Table 2: a breakdown of census statistics, sire and dam usage and indicators of the rate of loss of genetic diversity over 5 year periods (1980-4, 1985-9, 1990-4, 1995-9, 2000-4, 2005-9, 2010-14). Rate of inbreeding and estimated effective population size for each 5-year block can be compared with the trend in observed inbreeding in Figure 2. Table 2: by 5-year blocks, the mean number of registrations; for sires the total number used, maximum, mean, median, mode, standard deviation and skewness (indicative of the size of the tail on the distribution) of number of progeny per sire; for dams the total number used, maximum, mean, median, mode, standard deviation and skewness of number of progeny per dam; rate of inbreeding per generation (as a decimal, multiply by 100 to obtain as a percentage); mean generation interval; and estimated effective population size. years 1980-1984 1985-1989 1990-1994 1995-1999 2000-2004 2005-2009 2010-2014 Mean #registrations 253.6 221.6 197.4 188.8 151.8 131.4 106.4 Total #sires 140 102 99 98 112 112 104 Max #progeny 85 79 44 51 60 36 29 Mean #progeny 9.0429 10.853 9.9596 9.6122 6.7679 5.8571 5.0192 Median #progeny 6 7 8 6 4 3 2.5 Mode #progeny 1 3 8 2 1 1 1 SD #progeny 11.576 12.135 8.6543 10.243 9.0045 6.5249 5.4202 Skew #progeny 3.4753 2.8319 1.947 2.0246 3.217 2.1104 2.0182 Total #dams 221 163 163 150 146 142 119 Max #progeny 34 22 17 27 26 23 16 Mean #progeny 5.7285 6.7914 6.0491 6.2867 5.1918 4.6197 4.3866 Median #progeny 5 6 6 5.5 5 4 3 Mode #progeny 1 5 6 1 1 1 1 SD #progeny 4.5596 4.5044 3.379 4.259 4.3284 3.8311 3.7328 Skew #progeny 1.9567 1.0876 0.55559 1.2553 1.7941 1.3851 1.1375 Rate of inbreeding 0.040749 0.020417 0.020729-0.00082 0.001408-0.00665-0.03638 Generation interval 4.375 4.476 4.0814 4.5848 4.1932 4.404 4.3483 Effective pop size 12.27 24.489 24.121 n/a 355.15 n/a n/a 6
Figure 3: a histogram ( tally distribution) of number of progeny per sire and dam over each of the seven 5-year blocks above. A longer tail on the distribution of progeny per sire is indicative of popular sires (few sires with a very large number of offspring, known to be a major contributor to a high rate of inbreeding). Figure 3: Distribution of progeny per sire (blue) and per dam (red) over 5-year blocks (1980-4 top, 2010-14 bottom). Vertical axis is a logarithmic scale. 7
Comments As with most breeds, the rate of inbreeding was at its highest in this breed in the 1980s and 1990s. This represents a genetic bottleneck, with genetic variation lost from the population. However, since the mid-1990s the rate of inbreeding has slowed and even declined slightly, implying maintenance and even some replenishment of genetic diversity (possibly through the use of imported animals). It appears that the extensive use of popular dogs as sires has eased a little (the tail of the blue distribution shortening in figure 3). It should be noted that, while animals imported from overseas may appear completely unrelated, this is not always the case. Often the pedigree available to the Kennel Club is limited in the number of generations, hampering the ability to detect true, albeit distant, relationships. 8