Chicken Fingers: A case study of the relationship between development and homology

Lecture © Jonathan M. G. Perry
BIOL 606 Session, University of Alberta, February 9, 2000

Different definitions of homology exist and many different methods are used to ascertain homology. Some use morphological resemblance as evidence for homology, while others dismiss this method due to the probability of morphological convergence arising in response to similar functional demands. Some rely on developmental patterns as proof of homology; however, many instances of clearly homologous structures that arise through different developmental pathways (Wagner and Gauthier, 1999) argue for caution when equating development to homology.

Initial use of the term homology is usually attributed to Richard Owen although he was probably inspired by Renaissance anatomists who used this concept (=same organ in different animals) to relate all living animals to an animal 'archetype' or 'ideal' (Panchen, 1994; Hall, 1995). For Owen, a homologous part in two different organisms occupied a similar position and was connected to the same structures, but it could have a different form and function in each organism. Darwin introduced the use of genealogy to determine homology, whereas Haeckel urged that similar embryological origin was a sign of homology in adult structures. Several different methods are used to determine homologies today.

The question of the homologies of the avian forelimb digits is an old one (Hinchliffe and Hecht, 1984) but remains unanswered. Based on similar finger morphology in birds and theropod dinosaurs, many paleontologists maintain that the three avian fingers are digits I, II, and III (in reference to the pentadactyl 'primitive' tetrapod condition). In contrast, studies of embryological sequences in birds suggest that the fingers are actually II, III, and IV.

In their 1997 Science paper, Burke and Feduccia asserted that the avian fingers are digits II, III, and IV. They traced the development of chicken fingers through several embryonic stages and identified pre-cartilage condensations for digits II, III, and IV (V is present but fails to ossify in most birds). They saw no evidence of a condensation for digit I. Digits were identified based on the observation that digit IV is developed first (and seems necessary for the emergence of the other digits) in all tetrapods known from embryological sequences. Burke and Feduccia pointed out that early theropod dinosaurs show reduction of the digits identified as IV and V (e.g., Herrarasaurus) and progressive loss of these through theropod evolution toward a more bird-like, tridactyl morphology (e.g., Deinonychus). The incongruity between the evolutionary reduction of digits IV and V in theropods and the embryological evidence of digit IV in birds caused the authors to suggest that the presumed close relationship between these two groups is fallacious. However, many other characters support this relationship.

Wagner and Gauthier (1999) proposed a solution intended to restore the bird-theropod relationship. They argued that a 'frame shift' took place in the evolution of derived (bird-like) theropods. This frame shift caused the condensation for digit II to give rise to an adult structure morphologically identical to digit I. Equally, condensation III 'becomes' digit II and condensation IV 'becomes' digit III. Thus, Wagner and Gauthier contended that morphology, rather than developmental precursor, determines homology. Derived theropods (such as maniraptorans) were under strong selection for a grasping predatory hand, a feature already achieved in digits I, II, and III. Therefore, such a frame shift would have been advantageous. It was also necessary due to the initial development of condensation IV. Unfortunately, there is no direct evidence that this frame shift took place.

This debate continues as we lack sufficient theropod and early bird embryos to trace the 'true' homologies of their fingers. Even if such data turns up, we will always disagree on the 'true' homologies because workers continue to use different criteria for determining homology (e.g., Burke and Feduccia, 1997 vs. Wagner and Gauthier, 1999). Modern biologists must not rely on a one-to-one relationship between developmental precursors and adult structures, and thereby developmental methods for determining homology are cast in the same doubt that has always plagued morphological methods (similar morphologies may arise convergently). This case study provides no consensus on the best criteria for determining homology. 


Burke, A.C. and Feduccia, A. 1997. Developmental patterns and the identification of homologies in the avian hand. Science 278: 666-668.

Hall, B.K. 1995. Homology and embryonic development. Evolutionary Biology 28: 1-37.

Hinchliffe, J.R. and Hecht, M.K. 1984. Homology of the bird wing skeleton: embryological versus paleontological evidence. Evolutionary Biology 18: 21-39.

Panchen, A.L. 1994. Richard Owen and the concept of homology. In Homology: The Hierarchical Basis of Comparative Biology. Edited by B.K. Hall. Academic Press Inc. San Diego, pp. 21-62.

Wagner, G.P. and Gauthier, J.A. 1999. 1,2,3 = 2,3,4: a solution to the problem of the homology of the digits in the avian hand. Proc. Natl. Acad. Sci. USA 96: 5111-5116.


Rapporteur: Sean Graham

Did the paper really pin down the issue of homology of the 3 bird digits with those in other living and extinct amniotes?

It was first pointed out that the staining technique used in the paper may sometimes be inconsistent &endash; it was not clear why more advanced techniques (X-ray, NMR??) had not been used. An unspoken assumption of the paper was that digits can not be fused or lost "internally" in a sequence. This seemed reasonable to some, but not all, participants in the discussion.

A general unease surrounded discussion of this paper and the central question: "What is a homologous character?" Although there was some disagreement about how this should be defined and whether we can talk about "levels of homology," it was generally agreed that the question of homology must be carefully framed at an appropriate structural and phylogenetic level: Bat wings and bird wings are homologous limbs, but bat flight and bird flight are not homologous.

Botanists and zoologists in the audience differed in their level of comfort over the use of "serial homology" to think about homology within an organ. The possibility of developmental "frameshifts" led to much discussion. Can a frameshifted digit really be homologous to the original digit it replaces? The consensus in the audience was "NO". A useful distinction was drawn between developmental and phylogenetic definitions of homology. The latter seems to be the more powerful framework for thinking about homology, and was an issue which the discussion paper largely failed to address.

It was pointed out that the paper did not have much to say about bird relationships to theropods (and that even if birds ARE very closely related to theropods, that does not mean that digit sequence is homologous in the two groups &endash; this would require careful reconstruction on a phylogenetic tree). A possible simple solution to assessing the homology of digit sequencing in birds may be to find living or extinct birds with less reduced hands. Some suggestions were made about candidate taxa for study.

An ultimate aim of phylogenetic reconstruction may be to remove all spurious homology assessments in a data set by re-coding characters, after they are reconstructed on a tree. This goal does not seem to apply to trivial characters (like nucleotides). Does this highlight a deep split in the goals of molecular and non-molecular systematics?

The clearest definition of homology may be "presence in a common ancestor." Most participants left feeling a greater appreciation of the complexity of the issue. To paraphrase one participant "It is sometimes good to know that you don't know something very well."