Black Welsh Mountain Sheep Cluster Analysis

Data was provided by the American Black Welsh Mountain Sheep Association (ABWMSA), and included pedigree records from 9,054 animals born 1974 to present.  Only animals with registered status with ABWMSA were included in the analysis, for a total of 7,773 animals.  The number of animals registered by birth year is shown in Figure 1.

Figure 1. Number of Animals Registered by Birth Year

Inbreeding

The pedigree file was traced back until all ancestors were unknown.  There were 420 unique sires and 2,092 unique dams.  Next, inbreeding coefficients were computed for the population.  The average calculated inbreeding for all animals was 14.4%, with a range from 0 to 52.0%. The inbreeding trend over time is shown in Figure 2.  Although inbreeding levels have increased over time, they have remained steady in recent years and have even decreased from the peak in 1999-2000.

Figure 2. Inbreeding Trend by Birth Year

Another statistic of interest is to identify how many animals are non-inbred (an inbreeding coefficient of 0).  Figure 3 shows how the number of animals with an inbreeding coefficient greater than 0 has changed over time.  In the past 4 years, only 1 animal has had an inbreeding coefficient of 0.

Figure 3. Percent of Animals with an Inbreeding Coefficient Greater than 0 by Birth Year

The frequency distribution of inbreeding percentage for animals born 2008-present (n=1,413) is shown in Figure 4.  Such a small pool of non-inbred and lowly inbred animals makes future mating decisions to minimize inbreeding very difficult without the addition of new animals to the population.

Figure 4.  Inbreeding Coefficient Frequency Distribution, 2008-Present

Effective Population Size

 The effective population size (N_e) is a tool used to measure the genetic health of a population in terms of the true size of the breeding population.  The effective population size is defined as the number of individuals that would generate the current level of inbreeding and is computed as

                                                                N_e = 1 / (2  ?F)

where ?F is the increase in inbreeding per generation.   The Food and Agricultural Organization of the United Nations (FAO) suggests the N_e for a breed be maintained above 50.  The N_e for BWM sheep by generation is shown in Figure 5.  As shown in Figure 5, the N_e was healthy throughout the late 1990s and 2000s; only in recent years has that number dipped below the FAO threshold.

Figure 5. Effective Population Size by Generation (including the FAO threshold of 50)

Cluster Analysis

Next, cluster analysis was performed in order to determine how well the current BWM population is represented in the NAGP repository.  Genetic relationships between rams and ewes born 2010 to present plus rams with semen in the NAGP repository were computed.  Even if animals were included that have been removed from the breeding population (for example, via castration), they still represent a theoretical genetic pool for the repository.  And, considering the high level of relationship within the breed, a relative is likely to be available if an animal of interest is no longer available.

There were 57 rams with semen in the NAGP repository, with a total of 472 rams included in the cluster analysis.  There were 584 ewes included in the analysis, for a total of 1,056 animals clustered.  Their relationships were computed using their pedigrees traced back to unknown ancestors.  This included a total of 2,330 animals.  The PROC CLUS procedure of SAS was used to generate the clusters.  T-statistics were compared to determine the number of distinct clusters of BWM sheep that exist in the population.  For this analysis, 8 clusters were determined to be the logical number of clusters.  This break is shown in Figure 6.

Figure 6. Tree Dendogram showing 8 Clusters for Black Welsh Mountain Sheep

Table 1 shows the relationship within each cluster and the number of rams represented in the repository for each cluster.  The average relationship among all animals included in the analysis was 0.211.  Considering a half-sib relationship is 0.25 and a parent-offspring or a full-sib relationship is 0.5, the clusters are doing a good job of grouping highly related animals.  The rams in the repository are only represented in 2 of the 8 clusters.

Table 1. Relationship with Clusters

Table 2 demonstrates the effectiveness of the clustering procedure.  The relationships within clusters (diagonal) should be higher than the relationships between clusters (off-diagonal).  The more red the box, the higher the relationship; conversely, the more blue the box, the lower the relationship.

Table 2. Relationship Within and Between Clusters

Although relationships are quite high within and between clusters, the diagonal relationships are higher than the off-diagonal relationships, suggesting the clustering procedure was successful.

The relationship between the repository rams was also computed.  The average relationship was 0.212 with a range of 0-0.68. 

Pseudo-matings

Pseudo-matings of females in each cluster to males in each cluster were performed in order to compute the average, minimum, and maximum relationship of their potential offspring.  The average, minimum, and maximum relationship is shown in Figures 7, 8, and 9, respectively.

Figure 7. Average Relationship of Matings Within and Between Clusters, (pink represents females and blue represents males) 

Figure 8. Minimum Relationship of Matings Within and Between Clusters, (pink represents females and blue represents males)

Figure 9. Maximum Relationship of Matings Within and Between Clusters, (pink represents females and blue represents males)

Rams for NAGP

To develop a secure collection of Black Welsh Mountain Sheep semen for the repository, we would like to have 20 to 30 rams from across all clusters.  Although the repository already has more than the required number of rams, only 2 clusters are represented.  Focus on the remaining clusters is needed, with care taken to collect rams that are as lowly related to one another as possible.  A list of animals by cluster is provided in Appendix 1.

Appendix 1. Clustered animals by gender, cluster, inbreeding, and living status (highlighted animals are in NAGP repository)