Figure 1

Measurements of shells (rib width) of Terebratulina specimens examined in this study. Boxes in (B–D) depict data between the 25th and the 75th percentiles, vertical lines illustrate the full range of data. (A) Mean width of shell ribs of all investigated specimens (n = 39; 23 IceAGE samples supplemented by 16 shells from the brachiopod collection of the Museum für Naturkunde, Berlin). Note the discontinuity between two clusters at about 0.35 mm (red arrow). (B) The same data set depicted as box-and-whisker plots showing a significant difference between two groups (A: mean rib width 0.416 mm ± 0.078 SD, and B: 0.289 mm ± 0.037 SD; Mann-Whitney U-test: p ≤ 0.0001), which according to Curry & Endo16 should represent the two Atlantic species T. retusa (group A) and T. septentrionalis (group B). (C) Rib widths of T. retusa (0.345 mm ± 0.05 SD, n = 15) and T. septentrionalis (0,282 mm ± 0.038 SD, n = 10) are still significantly different (Mann-Whitney U-test: p = 0.003), when species are identified based on mitochondrial sequence data (12 S and 16 S rRNA), but the difference is less obvious than in (B ). Same as in (C), but species identification based on nuclear sequence data (28 S rRNA) leading to non-significant differences (Mann-Whitney U-test: p = 0.154) between rib-widths of T. retusa (0.335 mm ± 0.049 SD, n = 12) and T. septentrionalis (0.298 mm ± 0.059 SD, n = 12).