Genomic Inbreeding and Relationships in Wild Panda Populations

Captive breeding of giant panda uses kinship coefficient for controlling inbreeding under the assumption of unrelated wild founders. Increased inbreeding is expected to be an inevitable consequence of habitat loss and fragmentation but estimates of the actual inbreeding levels in wild pandas were unavailable. Applying joint variant calling by Samtools mpileup and BCFtools to published sequence data of 49 pandas including 34 wild pandas, we called 6,993,254 SNPs with MAF>0.1 after removing indels. Given the abundance of SNPs, we further required each SNP to have no missing genotype for all 49 pandas and to pass HWE test with p>0.01, and 150,025 SNPs satisfied these requirements. Using these SNPs, we estimated genomic inbreeding coefficient, additive relationship, dominance relationship, IBS probability, IBD probability, frequency of identity by genotype (IBG), and frequency of non-shared genotypes. Parentage testing and observed IBG excluded the possibility of parent-offspring relationship, monozygotic twins or duplicate samples. Assuming correct DNA samples and absence of major sequencing errors, a Xiaoxiangling panda and a Qionglai panda would be fullsibs with the strongest genomic additive and dominance relationships among all 49 pandas. The four largest habitats, Qionglai, Minshan, Qinling and Liangshan, were genetically unrelated, indicating that crossing between habitats could avoid inbreeding. Pandas between these four habitats on average shared 66.0-69.1% common alleles and 45.6-48.6% common genotypes, and pandas within each habitat shared 71.8-77.0% common alleles and 51.7-60.4% common genotypes. Qinling and Liangshan pandas had the highest levels of inbreeding and relatedness, whereas the inbreeding levels in Qionglai and Minshan were 28-45% of those in Qinling and Liangshan. These results lead to the recommendation of genome-guided and habitat-controlled captive breeding to avoid hidden inbreeding, and provide genomic evidence for the urgency to address the problem of habitat loss and fragmentation.