The digits of the wing of birds are 1, 2, and 3. A review.

Fossil evidence documenting the evolutionary transition from theropod dinosaurs to birds indicates unambiguously that the digits of the wing of birds are digits 1, 2, and 3. However, some embryological evidence suggests that these digits are 2, 3, and 4. This apparent lack of correspondence has been described as the greatest challenge to the widely accepted theropod-bird link (Zhou 2004. Naturwissenschaften 91:455-471). Here we review the pertinent literature regarding the debate on the origin of birds and wing digital identity and the evidence in favor of a 1, 2, 3 identity of the wing digits. Recent molecular evidence shows that the expression of Hoxd12 and Hoxd13 in the developing wing supports the theropod-bird link. In the chicken foot and in the mouse hand and foot, digit 1 is the only digit to combine the expression of Hoxd13 with the absence of expression of Hoxd12. The same is observed in the anterior digit of the wing, suggesting it is a digit 1, as expected for a theropod. Nevertheless, Galis et al. (2005. J Exp Zool (Mol Dev Evol) in press), argue that Hoxd12 and Hoxd13 expression patterns in mutant limbs do not allow distinguishing the most anterior digit in the bird wing from digit 2. They also argue that constraints to the evolution of limb development support the 2, 3, 4 identity of the wing digits. However, the case put forward by Galis et al. is biased and flawed with regard to interpretation of mutant limbs, developmental mechanisms, stages observed, and the description of the evolutionary variation of limb development. Importantly, Galis et al. do not present evidence from wild-type limbs that counters the conclusions of Vargas and Fallon (2005. J Exp Zool (Mol Dev Evol) 304B(1):85-89), and fail to provide molecular evidence to specifically support the hypothesis that the wing digits are 2, 3, and 4. The expression of Hoxd12 and Hoxd13 in the developing wing is consistent with the hypothesis that birds are living dinosaurs; this view can lead to a greater understanding of the actual limits to the evolutionary variation of limb development.