Figure 2 Bayesian and maximum likelihood (ML) trees for dunlin mitochondrial DNA haplotypes from Jiangsu in different years and from different breeding areas. (a) 2015 and (b) 2016. Numbers near nodes are posterior probabilities from Bayesian analysis and bootstrap values from ML analysis (BI/ML). Colored circles indicate subspecies of birds sampled in breeding areas or at Jiangsu (see key), which are grouped into Alaskan (ALA, pink shading) and Beringian (BER, yellow shading) lineages. The number of individuals from different lineages in different years is in parentheses.
Sixty-seven individuals were identified as males and 99 individuals were identified as females using a molecular method. The proportion of sexed birds decreased from 85.90% in 2015 to 57.23% in 2016. Owing to incomplete morphological data, the number of individuals was different in different morphological statistics. Scatter plots indicated little overlap in bill and wing lengths between the two years (Fig. 3). For males, wing length and bill length data conformed to a normal distribution, and both were significantly different between years (wing length, t = 3.004, df = 44, two-tailed P = 0.004; bill length, t = 7.612, df = 44, two-tailed P< 0.001). Wing and bill lengths of males were shorter in 2015 than in 2016 (Fig. 3b). For females, wing length data conformed to a normal distribution, whereas bill length data did not fit a normal distribution. Only bill length was significantly different between the two years for females (wing length, t = 1.657, df = 84, two-tailed P = 0.105; Mann–Whitney U-test: bill length, Z = –2.126, P = 0.033). Female dunlins in 2015 tended to have longer bills and wings than those in 2016 (Fig. 3a).