Technical Comments

Comment on “A Diverse Assemblage of Late Cretaceous Dinosaur and Bird Feathers from Canadian Amber”

See allHide authors and affiliations

Science  17 Feb 2012:
Vol. 335, Issue 6070, pp. 796
DOI: 10.1126/science.1216208


McKellar et al. (Reports, 16 September 2011, p. 1619) analyzed Late Cretaceous amber specimens from Canada and identified some filaments as dinosaurian protofeathers. We argue that their analysis and data do not provide sufficient evidence to conclude that such filaments are feather-like structures. Further investigation, including destructive sampling, must be carried out for more convincing conclusions.

McKellar et al. (1) recently examined Late Cretaceous amber specimens from Canada and reported stages I through V of feather evolution according to Prum’s (2) model. They concluded that the amber specimens contained dinosaur feathers because (i) specimens UALVP 52821 [figure 1B in (1)] and UALVP 52822 [figure 1, C and D, in (1)] were comparable to nonavian dinosaur fossil compressions; and (ii) fibers in specimen TMP 96.9.334 [figure 2, A to C, in (1)] exhibited microstructure coiling comparable to modern birds. Because they could not identify the fibers as any organism of an “end-member” evolutionary-developmental spectrum, and the fibers occurred concurrently with modern feather types, they inferred that the fibers were dinosaurian.

The interpretations of figure 1, B to D, in (1); figure 2, A to C, in (1); and the supporting figures in (3) convince us that adequate analysis was not conducted on these specimens and that overstated conclusions were made on subjective observations. Other figures in (1) (figure 2, D to F, and figure 3) are comparable with the feather microstructure in modern birds and cannot be regarded as anything but the ultimate stage of feather evolution.

The filaments described as stage I [UALVP 52821, figure 1B in (1)] are an order of magnitude smaller than the filaments of compression fossils of Sinosauropteryx prima (4) with which they were compared. According to McKellar et al. (1), the largest width of the filaments in UALVP 52821 falls within the range of the S. prima’s integument structures; however, the cited work [(5), p. 1719] reported “smaller ones are considerably narrower than 0.1 mm” in diameter. This does not represent a comparable scale to the 0.027 mm [(3), p. 3] measurements from the specimen in amber. Further, other researchers who examined the S. prima impressions reported measurements no thinner than 0.05 mm and declared the fibers to be collagen and not protofeathers (6). Although the filament lengths of UALVP 52821 were not measured in this study, the authors report that these “are consistent with the finer filaments found in this specimen [S. prima], and fall within the range of observed lengths” [(3), p. 5].

The study reports diameter measurements for UALVP 52821 as being within the lower range of modern hair measurements (of Australian mammals), but the authors curiously excluded hair as a possible fiber based on diameter and hollowness [(3), pp. 3–4 and figure S1]. Our interpretation of figure S4, B to D, and figure S5B in (3) is that the fibers do show internal divisions and do not appear to be hollow the entire length of the filament (Fig. 1, A and B). Additionally, comparing the amber fibers to specimens of fossil hair found in Canada (TMP 96.9.998) and France (dated Early Cretaceous) does not exclusively rule out UALVP 52821 as including hair filaments based on surface texture (cross-hatching) and diameter alone [figure S4, B to E, in (3)]. This analysis is open to subjective interpretation based on the published images.

Fig. 1

(A to C) Images from (3) of filaments within UALVP 52821 [(A) and (B)] and UALVP 52822 (C), with white arrows added to point to our interpretation of possible internal septa, cell walls, or other features that appear as evidence to dispute hollowness. (D) Photomicrograph of cottonwood seed (Populus tremuloides) with enlarged inset showing similar overall characteristics to the filaments in the amber specimens (15 μm width and outside wall thickness is average 42% of total width). Scale bar, 0.1 mm).

In the spinning disk confocal microscopy (SDCM) and laser scanning confocal microscopy (LSCM) analysis, McKellar et al. (1) state that the β-keratin autofluorescence results were influenced by background interference of the tested amber specimens, so we are confused as to why they conclude that UALVP 52821 [(3), p. 7)] contains β-keratin. Although the graphs have similar profiles, the intensity peaks and wavelength excitation values for β-keratin are different [figure S10, B and E in (3)], and we question the assumption that the materials are similar in nature.

McKellar et al. (1) explicitly state that reflective emissions of UALVP 52821 and TMP 96.9.997 are inconclusive [(3), pp. S7–S8] because values could be from a reflective surface when the specimen pulled away, or from fractures in the amber and not from β-keratin matrix. We feel that this represents an insufficient conclusion based on circumstantial evidence.

Another flaw in this particular analysis is that comparisons were made only to a feather in TMP 96.9.997 and not to the specimen that they reported as being hair (TMP 96.9.998). Confocal microscopy comparing the unknown fiber to the amber specimen containing hair would be more appropriate and could possibly rule out α-keratin as a component.

Because the reported stage II morphotypes [figure 1, C and D, in (1) and figure S5 in (3)] did not undergo the same analytical tests as UALVP 52821, we cannot assume that those fibers are the same [(1), p. 1620] and believe that the authors should have conducted similar analyses on that specimen instead of basing conclusions on visual observations.

The interpretations of fibers in TMP 96.9.334 [figure 2, A to C in (1)] are not analogous to modern feather microstructures of birds. Feather barbules (pennula) in modern birds are cylindrical and do not coil in both right and left directions on the same axis (Fig. 2A), nor do the twists extend distally. Twisting on modern birds only occurs on the flattened, straplike cells of bases [as shown in figure S12E in (3)]. McKellar et al.’s conclusions are invalid because (i) the amber specimen shows coiling at mid and distal positions of the fiber [figure 2A in (1)]; (ii) no base cell is observable; (iii) internodes are not flattened in modern birds; and (iv) figures S6A, S6B, S7A (lower arrow), and S7B (3) do not clearly show a rachis structure or a barb ramus, but rather seem to be a concentration of filaments focused on a central structure (observed on figure S7B). We have observed similar coiling on fibers of seed hairs (Fig. 2B), that is, Populus trichocarpa.

Fig. 2

Similarities of morphological characteristics of amber specimen and seed hair filaments. (A) Image from McKellar et al. (1) of TMP 96.9.334, with white arrows added to show the coiled distal filaments. Scale bar, 0.2 mm. (B) Cottonwood seed hairs (Populus trichocarpa) with thick arrows showing similar coiling, and arrowheads showing semiflattened areas similar to filament observed in figure 2C in (1) (filament width of 9 to 11 μm). Scale bar, 0.1 mm.

Although exploring amber specimens for clues to feather evolution may seem novel, this study lacks evidence and vigor to conclude that the fibers in UALVP 52821, UALVP 52822, and TMP 96.9.334 are dinosaurian. The analysis was not complete for each specimen, did not conclusively rule out hair or specialized plant parts as possible fibers, makes incorrect comparisons to modern feather microstructure, and cannot be cited as early stages of feather evolution. Because the topic of dinosaur feathers has been disputed, we feel that better analysis of the material in question, including destructive sampling of the amber specimens, is paramount.

Without concise identification of the various filaments depicted, there is no basis for assigning any of them to a particular group of organisms, to say nothing of dinosaurs.

References and Notes

  1. Supporting online material for (1).
  2. Acknowledgments: The Feather Identification Lab at the Smithsonian is funded through interagency agreements with the U.S. Air Force, the U.S. Navy, and the U.S. Federal Aviation Administration. L. C. Straker is a predoctoral fellow funded through the Coordenação de Aperfeiçoamento de Pessoal de Nível Superior/Fulbright Commission, Brazil.

Stay Connected to Science


Navigate This Article