Figure legends
Figure 1. A. Distribution map of the black-faced (red: breeding range; gold: wintering range) and royal spoonbills (blue; modified from Hancock et al. 1992). The arrows indicate the two sampling sites, Tainan, Taiwan (blue) and Hong Kong (orange), of the black-faced spoonbill used in current study. B. The cross-validation (CV) error of theADMIXTURE analysis (K : 1- 5) of the black-faced spoonbill from Taiwan and Hong Kong was smallest when K was equal to 1. The origin of samples is labeled at the bottom: T: Tainan, Taiwan; HK: Hong Kong. C. Samples of the black-faced spoonbill from Tainan, Taiwan and Hong Kong did not form distinct clusters in the plot of scores of PC1 (1.33% of genetic variance) versus PC2 (1.28% of genetic variance). The origin of the samples is indicated by colors: blue: Tainan, Taiwan; orange: Hong Kong.
Figure 2. A. The historical effective population size (N e) trajectories estimated by SMC ++ for both the black-faced and the royal spoonbill since the last glacial maximum (c.a. 22,000 years before the present); the fullN e trajectories of the two spoonbill species estimated by SMC ++ are shown within the inset. The black-faced spoonbill in magenta and the royal spoonbill in apricot. B. Demographic parameters of the recent bottleneck model for the black-faced spoonbill estimated from simulations of Fastsimcoal2 based on the folded site frequency spectrum of the 215,722 unlinked autosomal SNPs. The solid line is the maximum likelihood estimate of variousN e; The dashed line is the 95% CI ofN anc and N cur.T bot: initiation time of the bottleneck;T endbot: termination time of the bottleneck;N anc: long term N e before the bottleneck; N bot: N eduring the bottleneck; and N cur: the post-bottleneck N e.
Figure 3. A. The black-faced spoonbills sequenced had more runs of homozygosity (> 100 Kb; NROH) and longer total runs of homozygosity (SROH) than the royal spoonbills sequenced. The black-faced spoonbills are shown in circles, and the royal spoonbills in triangles. B. The linkage (r 2) between SNPs on the same scaffold decayed more slowly in the black-faced spoonbill genome than in the royal spoonbill genome. C. The proportion of singletons in the folded site frequency spectrum for unlinked (r 2< 0.2) autosomal SNPs was significantly lower in the black-faced spoonbill genomes (21.9%) than in the royal spoonbill genome (31.7%; χ 2= 24.82, p = 0.000). D. The Tajima’s D value for all autosomal SNPs of the black-faced spoonbill (0.24± 0.98, n = 4,522,032) is higher than that of the royal spoonbill (0.05± 0.99,n = 4,437,044; t = 296.51, p < 0.001). The black-faced spoonbill is shown in magenta and the royal spoonbill in apricot.
Figure 4. A. Significantly more genes in the black-faced spoonbill (2,094, 37.0%) are under relaxed selection (relaxation parameter k <1,p < 0.05) than in the royal spoonbill (1,542, 27.2%;χ 2= 123.04, p = 2.2e-16; Figure 4A); but not for genes under intensified selection χ 2= 3.17, p = 0.075). B. The mean number of nonsynonymous substitutions found in each individual of the black-faced spoonbill genome (8294.00± 332.66 substitutions) is higher than that found in the royal spoonbill genome (8025.78± 211.77 substitutions; t = -1.91,p < 0.035). C. The average Grantham’s score for each deleterious nonsynonymous substitution (Grantham’s score > 50) is significantly higher for the black-faced spoonbill (56.98± 0.69) than for the royal spoonbill (53.25± 1.83; two-sample t test= 6.5154, p = 3.6e-6). D. The ratio of homozygous to heterozygous nonsynonymous substitutions in the genome of the black-faced spoonbill (0.13± 0.04) is higher than that in the royal spoonbill genome (0.11± 0.007; t = -1.86, p < 0.04). The black-faced spoonbill is shown in magenta and the royal spoonbill in apricot.