Corrigendum: Three keys to the radiation of angiosperms into freezing environments

Journal name:
Nature
Volume:
521,
Page:
380
Date published:
DOI:
doi:10.1038/nature14371
Published online

Nature 506, 8992 (2014); doi:10.1038/nature12872

corrigendum Nature 514, 394 (2014); doi:10.1038/nature13842

Three readers pointed out that in this Letter we applied the threshold of 0.044 (the size at which freezing-induced embolisms are believed to become frequent at modest tensions) to the area of the conduit (in mm2) rather than the diameter (in mm). As a consequence, our analysis assumed far too few extant taxa as having a large conduit diameter, which altered the quantitative results considerably for conduit diameter. We now show that (1) the state combination with the largest persistence time is ‘large’ conduit ‘freezing unexposed’; (2) there are fewer transitions out of ‘large’ conduit ‘freezing exposed’ than we previously reported owing to many more extant taxa exhibiting this particular state combination; and (3) climate occupancy is more labile than conduit diameter (that is, the ratio of climate to trait is 5.67). Although these quantitative results change for conduit diameter, the interpretation of the possible pathways from ‘large’ conduit ‘freezing unexposed’ to ‘small’ conduit ‘freezing exposed’ is still qualitatively the same at the 0.044-mm-diameter threshold. That is, we still find the trait is more likely to evolve prior to a shift in climate occupancy (the trait-first interpretation) at 53.5%. The trait-first pathway, however, is no longer supported for the secondary 0.030-mm-diameter threshold reported on page 11 under “Coordinated evolution of growth habit, leaf phenology, and conduit diameter with climate occupancy” of the Supplementary Information of the original Letter.

The original Letter has not been corrected online. Figure 1 of this Corrigendum shows the corrected Fig. 2b and d. The Supplementary Information of this Corrigendum shows the corrected Extended Data Tables 2, 3 and 4 of the original Letter, with updated conduit diameter results in Extended Data Tables 2 and 3 and updated –lnL for the AABCD model in Extended Data Table 4. Please refer to the corresponding author A.E.Z. for additional details. We thank E. Edwards, J. deVos and M. Donoghue for bringing this issue to our attention.

Figure 1: This is the corrected Fig. 2b and d of the original Letter.

Supplementary information

PDF files

  1. Supplementary Information (232 KB)

    This file contains three tables, replacing the original Letter’s Extended Data Tables 2, 3 and 4. Extended Data Tables 2 and 3 show updated conduit diameter results. Extended Data Table 4 shows an updated –lnL for the AABCD model.

Additional data