Fig. 2. PSMC plot of one Xerces Blue (L05) and one Silvery Blue (L13) specimen. The two samples are those with higher average coverage. Individual PSMC plots were bootstrapped 100 times each (lighter lines). One year of generation time and a mutation rate of ยต=1.9x10-9 were used. The peak of the MIS 7 interglacial is marked in yellow.
We generated PSMC curves from the remaining lower-coverage individuals and down-sampled data from specimen L05 to 50% and 75% of the total coverage to explore the effects of coverage on estimation of heterozygous sites. Although there was a reduction in the effective population size estimates, as expected, the temporal trajectories in lower-coverage individuals were similar to their respective, higher-coverage Xerces Blue and Silvery Blue references (Fig. S6).
We subsequently explored the heterozygosity of each individual and found that Xerces Blue had 22% less heterozygosity on average than the Silvery Blue, a difference that is statistically significant (T-test; p=0.0072) (Fig. S7, Table S1). We searched for runs of homozygosity (RoH) that can indicate the existence of inbreeding in a dwindling population. The total fraction of the genome presenting RoH, although limited, is much higher in Xerces Blue (up to 6% of the genome) than in Silvery Blue, especially in short RoH of size between 100 and 500 kb (Fig. 3 and Fig. S7), consistent with background inbreeding. The limited presence of long RoH discards consanguinity as a common scenario in Xerces Blue.