Correction

Table Three (see Table 1 of this erratum) of our recently published paper [1] contains several errors. Here we present the corrected version of Table Three (see Table 2 of this erratum) and explain the new data. The authors regret the errors.

Table 1 Original and incorrect Table Three presented in Leroy et al. (2009)
Full size table
Table 2 Corrected Table Three
Full size table

Results

Partition of diversity

Errors concern the computation of the CW component developed by Ollivier and Foulley [2]. In the new version, CW ranged from -1 to 0.78. As aggregate diversity D is defined as a linear combination of CW and contribution to between-breed diversity, column D had also to be corrected, and ranged from -0.30 to 1.18. Consequently, the Pearson correlation between CW and ΔGD WS was found to be -1 (instead of -0.72 in the previous version), and the Pearson correlation between D and ΔGD T was found to be -0.59 (P = 0.008).

Discussion

Conservation priorities

In spite of the above modifications, the populations that contributed most to the total diversity, according to the approaches of Ollivier and Foulley [2] and Caballero and Toro [3], still remain mostly the non-endangered breeds (AR, PFS, TF) [instead of AR, PS, SF, TF in the previous version].

On the contrary, when considering the eight breeds classified as endangered or endangered/maintained by the FAO (ARD, AUX, BOUL, LAND, MER, POIT, POT, TDN) and the approach of Ollivier and Foulley [2], a change is noted for the breeds exhibiting the highest contributions to aggregate diversity D, which are now MER, LAND and POT, instead of BOUL, MER and POIT.

Finally, since the discussion on breed conservation is based on the use of several other methods and parameters, the above new results do not change our recommendations on which breeds specifically need support.