DEVELOPMENTAL MORPHOLOGY OF FLIPPERS IN SEA TURTLES AND PENGUINS. Grace W. Kwong. A Thesis Submitted to the Faculty of

Transcription

1

2 DEVELOPMENTAL MORPHOLOGY OF FLIPPERS IN SEA TURTLES AND PENGUINS By Grace W. Kwong A Thesis Submitted to the Faculty of The Charles E. Schmidt College of Science in Partial Fulfillment of the Requirements for the Degree of Master of Science Florida Atlantic University Boca Raton, Florida December 2006

3 Copyright by Grace W. Kwong

4 DEVELOPMENTAL MORPHOLOGY OF FLIPPERS IN SEA TURTLES AND PENGUINS By Grace W. Kwong This thesis was prepared under the direction of the candidate's thesis advisor, Dr. Jeanette Wyneken, Department of Biology, and has been approved by the members of her supervisory committee. It was submitted to the faculty of The Charles E. Schmidt College of Science and was accepted in partial fulfillment of the requirements for the degree of Master of Science. SUPERVISORY COMMITTEE: 6?[( Dr. S. KajlUra :_)~J L. {fdt;o Dr. S. Milton a: Dean, Graduate Studies :grams CPf~l1K# # ollege of Science Ill

5 ACKNOWLEDGEMENTS I gratefully acknowledge the guidance and feedback provided by my thesis committee members Drs. J. Wyneken, S. Kajiura, and S. Milton. I thank Dr. Wyneken especially for her considerable guidance as my graduate advisor. Thanks to S. Oulette, J. Foote, C. Whelan, K. Halager, and K. Rusenko for their assistance with sea turtle embryo collection and S. Floyd and H. Urquhart at the New England Aquarium for penguin embryo collection. Dr. Kajiura and A. Williams helped immeasurably with morphometries techniques and equipment. B. Botson and N. Desjardin assisted greatly with statistical analysis and technical support. Finally, I thank my friends and family for their encouragement on this gratifying journey. Research was conducted under Florida Fish and Wildlife Conservation Commission Marine Turtle Permit #073 to J.Wyneken. IV

6 ABSTRACT Author: Title: Grace W. Kwong Developmental Morphology of Flippers in Sea Turtles and Penguins Institution: Thesis Advisor: Degree: Year: Florida Atlantic University Dr. Jeanette Wyneken Master of Science 2006 There are no modem anatomical studies of flipper development or particularly any examining limb formation across distantly related taxa converging on similar flipper morphology. This study compares and contrasts the development of flippers in sea turtle (Caretta caretta) and penguin (Spheniscus demersus, Eudyptula minor) embryos. Embryos were fixed, cleared and stained for cartilage anlagen, and prepared as whole mounts. Skeletal elements forming the flipper and changes in their growth rates were described across developmental stages. Results suggest skeletal elements contribute differently to sea turtle and penguin flipper blades and there are significant differences in bone shape and growth patterns. Greater proportional increases in lengths and areas were found in sea turtles elements compared to penguins. Sea turtles appear to depend on a pathway resulting in elongation of distal elements to build a flipper, whereas penguin limbs undergo flattening and expansion of fewer elements to meet a similar structural goal. v

7 TABLE OF CONTENTS List of Tables vii List of Figures viii Introduction Materials and Methods Results Discussion Appendix Appendix Appendix Appendix Appendix Appendix Appendix Appendix Literature Cited VI

8 LIST OF TABLES Table 1. Comparison of sea turtle and penguin flipper growth and form Vll

9 LIST OF FIGURES Fig. 1. Turtle and penguin flippers differ in skeletal form and pattern. The basic tetrapod forelimb pattern is shown for comparison on the left. All have a humerus, radius, ulna, carpals, metacarpals, and phalanges. Sea turtles (center) form the flipper by elongation and some flattening of the carpals, metacarpals and phalanges. Penguins (right) have reduced and fused carpals and metacarpals which form most of the elongated and flattened carpometacarpus. The radius and ulna are also flattened and contribute to the flipper blade. The digits (II and III) form comparatively little of the penguin flipper (Louw, 1992) Fig. 2. Photo of sea turtle flipper without the humerus showing elements measured in C. caretta ( 29). H=humerus, Ra=radius, Ul=ulna, I=intennedium, Rc=radiale=radial carpal bone, Uc=ulnare=ulnar carpal bone, C=centrale, P=pisiform, Dc=distal carpal, M=metacarpal, Ph=phalanx Fig. 3. Diagram of penguin flipper showing elements measured. Key: H=humerus, Ra=radius, Ul=ulna, Rc=radiale=radial carpal bone, Uc=ulnare=ulnar carpal bone, M=metacarpal, Ph=phalanx Fig. 4. Detailed view of the cleared, stained flipper in a stage 37 S. demersus. The pollex or digit I (labeled*) is present during penguin flipper development but later becomes fused to metacarpal II (Louw, 1992; Parsons, 191 0) Fig. 5. Proportional changes in size of humerus, radius, ulna, wrist elements, and distal carpals of C. caretta (stages 21-29) Fig. 6. Proportional changes in size of metacarpals and phalanges of C. caretta (stages 21-29) Fig. 7. Proportional changes in size ofhumerus, radius, ulna, metacarpals, and carpals of S. demersus (stages 28, 30, 37), E. minor (stages 41, 42), and unknown penguin species (stages 25, 42). Lengths were not measured for the carpal elements Fig. 8. Proportional changes in size of the pollex and phalanges of S. demersus (stages 28, 30, 37), E. minor (stages 41, 42), and unknown penguin species (stages 25, 42) Fig. 9. Growth trends of humerus, radius, and ulna in C. caretta as a function of stage (stages 21-29). Error bars are shown Fig. 10. Growth trends of wrist elements in C. caretta as a function of stage (stages 21-29). Error bars are shown Fig. 11. Growth trends of distal carpal elements in C. caretta as a function of stage (stages 21-29). Error bars are shown Fig. 12. Growth trends of metacarpal elements in C. caretta as a function of stage (stages 21-29). Error bars are shown Vlll

10 Fig. 13. Growth trends of phalanx 1 in C. caretta as a function of stage (stages 21-29). Error bars are shown Fig. 14. Growth trends of phalanx 2 in C. caretta as a function of stage (stages 21-29). Error bars are shown Fig. 15. Growth trends of phalanx 3 in C. caretta as a function of stage (stages 21-29). Error bars are shown Fig. 16. Growth trends ofhumerus, radius, and ulna ins. demersus (stages 28, 30, 37), E. minor (stages 41, 42), and unknown penguin species (stages 25, 42) as a function of stage. Error bars are shown Fig. 17. Growth trends of carpal and metacarpal elements ins. demersus (stages 28, 30, 37), E. minor (stages 41, 42), and unknown penguin species (stages 25, 42) as a function of stage. Lengths were not measured for carpal elements. Error bars are shown Fig. 18. Growth trends ofthe pollex and phalanges ins. demersus (stages 28, 30, 37), E. minor (stages 41, 42), and unknown penguin species (stages 25, 42) as a function of stage. Error bars are shown IX

11 INTRODUCTION Evolutionary and developmental events are intimately connected processes in the natural history of vertebrates and invertebrates. Groups of organisms which are not closely related and are otherwise distinct may share similar features (common structure, function, or development). Sea turtles and penguins represent two distantly-related groups that have independently invaded the marine ecosystem. Although both groups are secondarily aquatic, they have highly specialized morphologies differing from their respective ancestors and accompanying their ecological divergence in niche utilization. Interestingly, sea turtles and penguins share similar morphological attributes, such as streamlining of the body and modified forelimbs as flattened, blade-like flippers. Flippers are supported by flattened, elongated bones and contrasting levels of skeletal element reduction in sea turtles and penguins. Sea turtle flippers are semi-rigid, hypertrophied, and wing-like structures with elongate third to fifth digits and flattened carpal elements with movement-restricting joint surfaces (Hirayama, 1994). The flipper is modeled as well by the flattening of bones of the autopodium (Wyneken, 2001 ). Elements in the wing skeleton of penguins are dorsoventrally flattened, and the joints allow for little movement between bones (Raikow, 1985). Flippers serve both sea turtles and penguins in locomotion, allowing for lift-based endurance swimming (Clark and Bemis, 1979; Hui, 1988; Wyneken 1997). Members ofboth taxa use their flippers for propulsion and steering (Clark and Bemis, 1979; Webb, 1984; Lohmann and Lohmann, 1992; Wyneken, 1997). Swimming is a fundamental component of the lifestyles ofboth

12 groups of animals, enabling them to make extensive migrations between feeding and breeding grounds. Developmentally, the shape of the forelimbs as flippers in sea turtles and penguins is conserved or converges in spite of their independent origins. Given that analogous structures evolved independently in these two tetrapod groups, both similarities and differences are expected in the development and form of the forelimb elements. Parallels may stem from similarities in (i) basic developmental programs, (ii) selection pressures acting on the forelimb shape, and (iii) the flipper components of the Bauplan, which defines a fundamental structural plan. Differences may result from the distinct evolutionary and developmental histories and may take the form of losses of elements, heterochronic changes in developmental events, and/or completely novel aspects of the developmental programs. It is possible that sea turtles and penguins evolved quite unique skeletal arrangements within the limb, but these alternative solutions meet the same functional requirements. Examination of a range of embryonic stages from limb-bud to near hatching can give insight into the patterns of limb formation in two distinct lineages that share similar end-morphology. Such a comparative approach is useful in discerning how two distantly-related groups converge (Wagner, 1994 ). There have been no modem anatomical studies examining limb formation across distantly related taxa that converge on similar flipper morphology. Although sea turtles and penguins historically shared a distant common reptile ancestor and the forelimbs are flippers in both, the gross similarities in flippers evolved 2

13 independently. Characteristic of avian species is the reduction in number and size of the carpus, metacarpus, and phalanges in the forelimb during development. Although present in early avian development, the first and fifth digits as well as several carpal bones disappear (Rol'nik, 1968). In contrast, sea turtles retain all five phalanges and most of the carpal and metacarpal bones, although they may become flattened and fused. I compared and contrasted the development of flippers in the cheloniid sea turtle (Caretta caretta) and the penguins (Spheniscus demersus and Eudyptula minor). Here I describe the skeletal elements and the changes in their growth rates as they form the flipper blade. Examination of forelimb development from early limb to near hatching stages allowed us to track the patterns of flipper formation in these two distinct lineages. Given that sea turtles and penguins are adapted to pelagic lifestyles and both evolved flippers, I tested the hypothesis that development of the flipper in both taxa is comparable, diverging only skeletally and only in late embryonic stages. I found that skeletal elements contribute differently to sea turtle and penguin flipper blades and there are significant differences in both shape and growth patterns of the bones. The results of this study did not support my hypothesis. MATERIALS AND METHODS I collected representatives of all stages of sea turtles, but obtained an incomplete series of penguins. Miller' s (1985) staging criteria were used for sea turtles (limbs formed during stages 16-31) and Herbert's (1967) stages were used for penguins (limbs 3

14 formed during stages 17-45). Herbert's (1967) staging criteria were based on Adelie penguin (Pygoscelis adeliae) specimens; however, this species is morphologically similar to the two species considered here. Appendices 1 and 2 describe limb development by stages in sea turtles and penguins, respectively. Loggerhead sea turtle embryos (Caretta caretta, Lineaeus 1758) were obtained from eggs during nest evaluations along the coast of Florida during two consecutive nesting seasons (2003 and 2004). All embryos came from unhatched eggs and did not complete development due to tidal inundation or other natural phenomena; abnormal embryos were not included in the series. Embryos of two species of penguins, the African penguin (Spheniscus demersus, Lineaeus 1758) and Little Blue penguin (Eudyptula minor, Forster 1791), were obtained from the New England Aquarium, along with two penguin embryos whose species were unidentified. These specimens were from eggs that were not incubated to term. Only embryos in good condition were used for analyses of flipper morphology. Sea turtle and penguin embryos were fixed in 10% buffered formalin and subsequently cleared and stained with Alcian blue (Wassersug, 1976) for cartilage anlage (Dingerkus and Uhler, 1977), but not counterstained for bone. Specimens were prepared as flat whole mounts. In some cases, it was necessary to disarticulate overlapping elements in order to take accurate measurements. Individual skeletal elements were digitally photographed through a dissecting microscope (Olympus SZX12); 4

15 measurements were collected using a slide micrometer and ImageJ software (Abramoff et al., 2004). Magnification used for photographs ranged from 3-8X. Comparisons covered the stages when at least the cartilage condensations of the stylopodium and zeugopodium were present. The earliest stages when these elements were present and distinct (sea turtle stage 21, penguin stage 25) were used as the baselines for growth comparisons. Because of the incomplete penguin series, I could not start comparisons with the first limb bud stages and I could not compare all stages in both taxa. s and species of penguin specimens included in this study are presented in Appendix 3. The proportional increase in size of each element in subsequent stages as the flipper developed was determined by using measurements from the earliest stages of sea turtle and penguin specimens as baselines for comparison. I used either the left or right limb depending on which was stained more thoroughly. Proportional growth was determined for both length and area of the developing skeletal elements. Using a micrometer in digital photographs of each flipper, I used ImageJ (Abramoff et al., 2004) to set scale and establish pixel:length relationships to determine length and area of each element. Proportional growth measures were graphed to identify which components diverged and which contributed most to the taxon-specific flipper form. Least squares linear regressions (Legendre, 1805) were used to analyze the relationship between embryonic stage and proportional growth in length and area of each skeletal element. 5

16 RESULTS General Description of Flipper Morphology of Sea Turtles and Penguins The flippers of sea turtles and penguins share similarities in their development. As with other vertebrates, the limb buds extend laterally from the flank region. Proximal elements form before intermediate elements and distal structures (e.g. wrist and hand elements) form last (Fig. 1). The digital plates appear early in limb formation (sea turtle stage 21 and penguin stage 25). Photo atlases of sea turtle and penguin flippers at the developmental stages studied are presented in Appendices 4 and 5, respectively. Variation in Flipper Morphology of Sea Turtles and Penguins The shapes of the flippers differ; the sea turtle flipper is broader at the distal end while the penguin flipper is tapered distally. The underlying skeletal pattern also diverges in numbers and shapes of elements present in turtle and penguin flippers (Table 1). There are fewer autopodial (wrist and hand) elements in the penguin flipper than in the sea turtle flipper. Because of these differences, I measured 31 skeletal elements in sea turtles and just 11 in penguins (Figs. 2-3, respectively). The penguin wrist elements include the ulnare (ulnar carpal bone), radiale (radial carpal bone), and the carpometacarpus (fused carpals and metacarpals) whereas sea turtles have a greater complement, consisting of the intermedium, ulnare, centrale, pisiform, radiale, and five 6

17 7 distal carpals and metacarpals. Penguins have fewer metacarpals (two) and digits than sea turtles; remnants of metacarpal I and the pollex (digit I) are reportedly fused to metacarpal II in adults (Louw, 1992; Parsons, 191 0). However, in embryonic penguins the pollex is clearly visible in Alcian blue stained specimens (Fig. 4). Digital plate development differed in the two taxa; the sea turtle digital plate was rounded and that of the penguin was oval and proximodistally elongated. Elongation of specific elements also varied between sea turtle and penguin flippers. Sea turtle phalanges showed greater elongation relative to those of penguins. In penguins, there was flattening of zeugopod and autopod elements through the pronounced addition of cortical bone to preaxial and postaxial shaft regions; this was not seen in sea turtles. Variation in Growth Trends of Skeletal Elements in Sea Turtles and Penguins There are clear differences between the sea turtle and penguin flipper in the proportional growth of individual elements. Skeletal elements in the sea turtle flipper grew proportionately longer throughout all stages of development with the greatest rate of outgrowth in the final stages before hatching (between stages 27-29, Figs. 5-6). Proportional changes in penguin flipper elements tend to be relatively small. Growth increases with embryonic age in penguins, but rate increase is greater in sea turtles (Figs. 7-8).

18 In all elements measured, greater proportional increases in both lengths and areas were found in sea turtles compared to penguins (Figs. 5-8). Length and area of the sea turtle humerus increased 13- and 100-fold, respectively from stages 21 to 29, while the penguin humerus showed 7.5- and 29-fold increases, respectively. The radius and ulna of sea turtles increased 11- to 12-fold in length while these elements increased 6.5- to 7.5- fold in penguins. The sea turtle radius increased 40-fold in area and the ulna increased by 60-fold. In contrast, the penguin radius and ulna showed a 23- to 24-fold change in areas. Wrist elements of the sea turtle flipper increased 5- to 11-fold in length and 6- to 28-fold in area. Lengths were not recorded for wrist elements in penguin specimens due to their irregular margins; however, their area increased 7- to 11-fold. The distal carpals (grouped together for this category) in sea turtles increased 2.5- to 5-fold in lengths and 5- to 25-fold in areas. Distal carpal bones are not present in penguins. The metacarpals (considered together) in sea turtles increased 8- to 12-fold in lengths and 17- to 33-fold in areas. In contrast, metacarpals in penguins underwent just a 2-fold change in length and 2- to 9.5-fold changes in area. In penguins, the pollex (digit I) increased~ 2-fold in length and area. Phalangeal measures were grouped together by the proximodistal identity of the elements (e.g. the measures of phalanx 1 in all 5 digits were grouped). Phalanx 1 in sea turtles increased in length 14- to 22-fold and 10- to 22-fold in area, while in penguins, phalanx 1 increased just 1.5- to 2-fold in lengths and 2.5- to 3-fold in area. Phalanx 2 in sea turtles increased 5- to 17-fold in length and 3- to 32-fold in area. Phalanx 2 of digit III increased by the greatest proportional length and that of digit V changed the least. There was no clear trend in the growth of phalanx 2 in penguins as it varied through stages; however, length measured at the largest size found was just ~ 1.5X 8

19 the starting size and area changed~ 1.6-fold. Phalanx 3 in sea turtles increased 6- to 13- fold in lengths and 7- to 38-fold in areas. Digit III showed the greatest proportional change while digit IV showed the least. Phalanx 3 is absent in the penguin digit. Generally, the lengths of flipper elements in penguins showed proportional increase from stages 25-37; at stage 41, length increases proportionally slowed. At stage 42, lengths increased again slightly. Areas of penguin flipper elements showed similar trends as were seen in lengths, with two exceptions: (1) metacarpal II showed~ 9.5-fold increase in area whereas metacarpal III showed~ 2-fold increase, while maintaining similar lengths and (2) the radial carpal bone increased in area from stage 30 to 37; then growth appeared to be isometric to stage 41 (inferred since I lack intermediate stages). Growth decreased from stage The ulnar carpal bone increased in area from stage 30-37, like the radial carpal. However, its growth is large (11-fold increase). By stage 41, the ulnar carpal bone decreased in size, increasing again by stage 42 ( ~ 11-fold increase since it first appeared). Least squares regression analyses indicated that embryonic stage significantly predicted proportional changes in lengths and areas of most sea turtle and penguin skeletal elements (Appendices 6-8, respectively). Regressions were significant for all sea turtle elements, except for digit IV, phalanx 3 area. For penguins, regressions significantly explained most changes in flipper element size with the exceptions of lengths and areas of digit II and III phalanges as well as areas of the ulnar carpal bone and pollex. In sea turtles, general growth trends are similar for 26 elements (for both length and area); the exceptions are the radiale, centrale, distal carpals 2 and 3, and digit IV 9

20 phalanx 3, which grew irregularly (Figs. 9-15). All other trends are nearly linear throughout stages observed and growth rates increased between stages 27 and 29 (latter developmental stages), especially in the metacarpals and phalanges. In penguins, length and area generally increased for all 11 elements from stages (Figs ). However, at stage 41, both length and area decreased in most elements exclusive of the radius, metacarpal III, and the pollex in which lengths increased slightly. By stage 42, length and area both increased again in all elements. DISCUSSION The basic tetrapod forelimb pattern consists of a proximal stylopod (humerus), intermediate zeugopod (radius and ulna), and distal autopod (wrist bones and phalanges) (Andrews and Westoll, 1968; Ahlberg and Milner, 1994; Wagner and Chiu, 2001; Coates et al., 2002). This arrangement may be a fundamental characteristic of the archetypal tetrapod limb (Wagner and Chiu, 2001), which serves as the basis for extensive functional adaptation (Gardiner et al., 1998). Sea turtles and penguins follow the basic tetrapod pattern generally but differ in the proportional contributions of the skeletal elements that each makes to the flipper. The wrist and phalangeal elements in both sea turtles and penguins are elongated and somewhat flattened to form the flipper blade. However, they diverge from the basic tetrapod design in the patterns ofbone losses, fusions, and reductions in flipper elements characteristic to each group. 10

21 Early flipper development is similar in sea turtles and penguins. As with other vertebrates, limb bud outgrowth occurs proximodistally and differentiates in a similar manner. Establishment of the dorsoventral (knuckle/palm) and anteroposterior (thumb/pinky) axes are also conserved. These similarities in patterning illustrate the conservative nature of early developmental events, suggesting that there are basic rules for limb formation among tetrapods. General morphological characters give rise to those having increasing levels of specialization until finally resulting in the most specialized forms (Gilbert, 2003). Thus, although limbs are similar during early stages of formation, development diverges in latter stages until the characteristic forms of the sea turtle and penguin flippers emerge. Among the mechanisms by which differences may arise, heterochrony, the change in timing or rate of developmental events, is among them. This may especially be the case for sea turtle phalanges which are greatly elongated compared to other secondarily aquatic groups (penguins, whales, ichthyosaurs). Gould (1977) proposed four models of heterochrony. Of those types ofheterochrony, the two likely to be responsible for the elongation of phalangeal elements are acceleration (of development) and hypermorphosis (extension of growth period), both of which lead to exaggerated adult features. I hypothesize that, like in other vertebrate character divergences, shifts in timing of gene expression during limb formation could lead to this morphology through extended time spent in a particular phase of development, resulting in elongated elements (Smith, 2003). Heterochrony involving regulatory factors, downstream signals, or receptors can also produce changes in alternative developmental pathways resulting in various phenotypes. 11

22 Where penguins form phalanges, they show some limited degree of elongation; however, not as much as is seen in sea turtles. Both species share more in common with one another's flipper structure than with the alternative flipper skeletal form, hyperphalangy, found in cetaceans, ichthyosaurs, and plesiosaurs. Hyperphalangy is a mechanism that forms flippers through an increase in the number of phalangeal elements, which contribute to the flipper blade, rather than phalangeal elongation (Sues, 1987; Richardson and Oelschlager, 2002). Elongation of the digits is not an uncommon mechanism that allows expansion of ecological niches through the establishment of novel locomotor modes. Among mammals, bats (order Chiroptera) modified forelimbs into wings capable of powered flight. Greatly elongated third, fourth, and fifth digits support a membrane of skin that comprises the wing (Sears et al., 2006). Developmentally, digital elongation is accomplished through increased rates of proliferation and differentiation of cartilage cells within the hypertrophic zone of the wing growth plate, which seems to be at least in part controlled by members of the bone morphogenetic protein (Bmp) family of secreted growth factors (Sears et al., 2006). In particular, Bmp2 was found to have a significant role in elongation of bat wing digits. Interestingly, Bmps are suggested to play roles in establishing digit identities (Randall and Fallon, 2000) and are involved in interdigital cell death (Zou and Niswander, 1996). Future studies looking for expression of Bmps may show similar roles in the elongation of digits in the sea turtle flipper. 12

23 Differences between sea turtle and penguin flippers are obvious when one considers the wrist and hand elements. Sea turtles retain most wrist and hand elements, while penguins retain relatively few. Sea turtle elements are greatly elongated and somewhat flattened, whereas penguin elements are greatly flattened. Such divergences could be due to heterochrony, phylogenetic inertia (fundamental differences associated with ancestral patterning), or other responses to selection acting on the limb in the two groups. Penguin flipper elements receive contributions from the addition of cortical bone to both preaxial and postaxial shaft regions of bones. This is not likely to be a general bird characteristic since most fliers have evolved hollow or pneumatic bones for aerodynamic efficiency. However, as penguins do not employ this locomotor mode, the added bone may serve to flatten the flipper for increased hydrodynamic efficiency. There must also be differences in apoptotic patterns where sea turtles retain large areas of interdigital soft tissue "webbing" in the flipper blade while penguins show a lesser degree of webbing. This is not surprising given the lack of digits in the penguin flipper. In addition, penguin digits are more closely spaced together within the flipper than sea turtle digits which are separated within the flipper webbing. There are differences also in the shape of the digital plate. These could be the result of which digits contribute to the flipper distally. The digital plate of the penguin may be more tapered because fewer digits contribute to the blade, primarily digits II and III. Distinct differences in growth rates of skeletal elements were found between sea turtle and penguin flippers. In sea turtles, outgrowth occurs throughout developmental stages, peaking at late stages. Outgrowth is proportionally less in embryonic penguins. 13

24 This may be related to differences in the ecology and behavior of the two groups. Hatchling sea turtles use their flippers within a few days of hatching to exit the nest and swim offshore to nursery grounds (Carr, 1986; Wyneken and Salmon, 1992) whereas penguin chicks remain on land during an extended growth period under parental care, using their wings (flippers) little (Darby and Seddon, 1990). A study on post-hatching ontogeny oflimb skeletons in semiprecocial birds (California gulls, Larus californicus) showed that differences in growth between wings and hindlimbs are responses to different locomotor demands (Carrier and Leon, 1990). Wings are not used in locomotion until gulls are fully grown; however, the hindlimbs become functional shortly after hatching. Pectoral size and strength were found to increase very little until the juveniles reached adult size, whereupon surface area of the wing doubled. It might be expected for penguins to follow a similar developmental program. Sea turtle and penguin flipper development growth trends were compared; however, the resolution in the penguin proportional growth plots was not as great as in sea turtles as I lacked a complete representation of earlier stages of development. Ideally, additional early and intermediate stages in the penguin series would fill those gaps. Measurements of skeletal elements were taken from specimens of two penguin species, S. demersus and E. minor; thus, it is possible that using pooled data from two species could obscure differences in size and growth rates. E. minor is smaller at hatching than S. demersus; however, the flippers in both species are small compared to body size (H. Urquhart, pers. comm.). Additionally, two unidentified penguin specimens (most likely S. demersus or rockhopper penguin Eudyptes chrysocome, both intermediate in size) 14

25 representing stages 25 and 42 were included in this study. The abrupt decrease in growth for most elements from penguin stage 37 to stage 41 and increase at stage 42 may reflect such a sampling artifact. However, a phylogeny of all extant penguin species based on morphological characters recovered a strongly supported suprageneric clade composed of the Spheniscus and Eudyptula genera (Bertelli and Giannini, 2005). The Eudyptes genus fell into a clade consisting of all genera excluding Spheniscus and Eudyptula; the genera Eudyptes and Megadyptes were found to be sister taxa. If the details of limb development are conserved in this clade, then their inclusion may realistically reflect flipper growth patterns. The flippers of sea turtles and penguins are a striking example of convergent evolution. In the distinct lineages leading to sea turtles and penguins, the forelimbs in both groups have been modified in uniquely similar ways resulting in wing-shaped flippers which serve to propel the animals through their aquatic environments. Flippers may be similar in gross morphology due to the constraints and limitations that accompany the achievement of hydrodynamic efficiency through a dense medium, such as the need to increase lift while reducing drag. The shape of a flipper, or of any foil, is directly related to the functions it must perform. Flippers must efficiently serve the locomotor needs of the organisms which possess them. In both sea turtles and penguins, flippers are used to propel the animals through the water as well as for steering; however, the shapes of the two kinds of flippers differ in detail. Sea turtle flippers are broader distally, while in penguins, the flipper is tapered at its distal end. Thus, the two groups of flippers have differing aspect ratios, the ratio of length to width of the flipper. Penguins 15

26 appear to have higher aspect ratio flippers than sea turtles. This translates into the penguin flipper being more suitable for long distance endurance swimming while the sea turtle flipper is more efficient at maneuvering (Westneat and Walker, 1999). Aspect ratio affects the stability and maneuverability of a foil. Foils with high aspect ratios are welldesigned for high performance endurance behaviors, whereas foils with low aspect ratios are more suitable for maneuvering behaviors. Functional morphology of the limb is certainly influenced and constrained by developmental patterning within the limb. Component parts forming the limb appear to face differential constraints during evolution and development. The appearance of differential constraints may reflect developmental modularity in developing flippers (Schlosser and Wagner, 2004). It has been suggested that modularity offers a certain degree of "evolvability" to a system (e.g. limb skeleton) by allowing certain features to be altered without significantly changing the integrity of the system. Limbs can be regarded as anatomical modules, both in their entirety and also in their components. Limbs are subject to evolution, as can be observed by the astounding variability in appendage forms and types. Of the three major limb segments, the autopod shows the highest degree of evolvability and morphological variability. This is true of vertebrate autopods in general, including the flippers of sea turtles and penguins. There is great variability in both numbers and types of elements comprising the autopod in the two groups examined here. The autopod is generally less constrained than the stylopod and zeugopod and more likely to undergo losses, duplications, or fusions (Shubin et al., 1997; Richardson et al., 2004). It appears that sea turtles rely on the elongation of distal 16

27 elements to build a flipper, while penguins rely on flattening and expansion of fewer elements to meet a similar structural and functional goal. Variability in limb form is common in tetrapods and may be quite extensive in certain cases, as is evident in the autopodia of bat wings and flippers of marine vertebrates. 17

28 APPENDIX 1. Limb development based on sea turtle staging series by Miller (1985) Changes in Limb Development Small limb buds present on lateral body wall folds Limb buds project latera-posteriorly from lateral body folds Limb buds extended as projections from body wall; digital plates forming distally Limb buds resemble paddles and extend laterally from body wall Limbs have thickened digital plates; delineated by distinct apical ridge Paddle-like digital plates lack grooves and ridges; digital plates not defined by ridge at distal end Distal ridge bounds digital plate; proximal ridge delimits limb; no digital ridges visible Forelimb elongated; digital ridges evident on digital plates Digital ridges well developed; 1 51 digit protrudes beyond webbing but lack claw Forelimbs unpigmented and lacking scales; claws present on 1 51 digit and phalanges are well-defined Scales may be present on leading and trailing edges of flippers Flipper has scales and some pigmentation Accent colors on limbs distinct Basic hatchling morphology and pigmentation present 18

29 APPENDIX 2. Limb development based on penguin staging series by Herbert (1967) (Hours) 17 ( days) 18 (10-11 days) 19 (11-12 days) (12-13 days) 22 (12-14 days) 23 (13-15 days) 24 (15-16 days) 25 (15-16 days) 26 (16-17 days) 27 (17-18 days) 28 (18-20 days) 29 (20-21 days) 30 (~ 21 days) 31 (21-22 days) 32 (21-22 days) Changes in Limb Development Forelimb primordia are condensations of tissue elongated antero-posteriorly on lateral body folds Limb rudiments elevated from lateral body folds (No observable changes reported) Forelimb is more clearly demarcated from the body Forelimb is adducted (inclined posteriorly) Forelimb is flattened Forelimb shows first signs of elbow flexion Olecranon joint of flattened forelimb is bent at angle of and is slightly grooved Forelimb has protuberance on leading edge, and a ridge extends for one-third of the distance from the tip Elbow flexion is almost 90 and protuberance is very pronounced Forelimb has two well-defined ridges (incipient radius and ulna) Forelimbs are larger than previous stage; protuberance on leading edge is elongated Forelimb is more pointed distally than in previous stage Pointed forelimb is slightly curved; incipient carpometacarpal articulation is recognized half-way along its length Forelimb is more wedge-shaped and its protuberance has almost disappeared Forelimb is more flipper-like 19

30 33 (22-23 days) 34(~23 days) (~ days) 40 (28-29 days) 41 (29-31 days) 42 (31-33 days) 43 (32-34 days) 44 (33-36 days) 45 (Hatching) Faint line of feather germs is present along trailing edge of flipper Long smooth flipper is wedge shaped (No observable changes reported) Hind edge of flipper covered with long black down feathers still in feather sheaths; rest of flipper covered with small triangular feathers Feathers on leading edge of flipper remain unpigmented while all other body feathers and claws are pigmented Feathers on leading edge of the outennost part of flipper are still unpigmented Forelimbs are completely pigmented (No observable changes reported) (No observable changes reported) 20

31 APPENDIX 3. Penguin specimens by stage and species. Unidentified specimens were possibly S. demersus or E. chrysocome. Species 25 Unidentified 28 S. demersus 30 S. demersus 37 S. demersus 41 E. minor 42 E. minor, Unidentified 21

32 APPENDIX 4. Photo atlas of sea turtle (C. caretta) flipper skeleton forming. The following stages are represented in this study: A: stage 21, B: stage 22, C: stage 23, D: stage 24, E: stage 25, F: stage 26, G: stage 27, H: stage 28, and I: stage 29. Scale = 0.5cm. The left flipper is shown in dorsal view in all photos, except D (right flipper). A )

33 APPENDIX 5. Photo atlas of penguin (S. demerus (Sd), E. minor (Em), unidentified (spp)) left.flipper skeletal development. The following stages are represented in this study: A: stage 25 (spp), B: stage 28 (Sd), C: stage 30 (Sd), 0: stage 37 (Sd), E: stage 41 (Em), F: stage 42 (Em), and G: stage 42 (spp). Scale= 0.5cm. The flipper is shown in dorsal view in all photos, except A (ventral view). ~ v 20,n 23

34 APPENDIX 6. Least squares regression analyses of proportional changes in lengths of sea turtle skeletal elements. Degrees of freedom (DF); Alpha= Analyses could not be performed on elements with incomplete data (n/a). Skeletal element Regression Equation R 2 Value DF P-value Humerus y = 1.49x << Radius y = 1.21x << Ulna y = 1.27x << Intermedium y = 1.26x < Ulnare y = 1.17x < Radiale n/a n/a -- n/a Pisiform y = 1.27x < Centrale y = 0.58x Distal Carpal 1 y = 0.57x << Distal Carpal 2 y = 0.45x Distal Carpal 3 y = 0.31 X Distal Carpal 4 y = 0.63x Distal Carpal 5 y = 0.60x < Metacarpal I y = 1.42x < Metacarpal II y=1.17x < Metacarpal III y = 1.21 X << Metacarpal IV y = 1.56x << Metacarpal V y = 1.78x < Digit I Phalanx 1 y = 1.99x < Digit II Phalanx 1 y = 2.15x < Digit III Phalanx 1 y = 2.36x << Digit IV Phalanx 1 y = 3.43x << Digit V Phalanx 1 n/a n/a -- n/a Digit I Phalanx 2 y = 0.91x < Digit II Phalanx 2 y= 1.40x Digit III Phalanx 2 y = 2.23x < Digit IV Phalanx 2 y = 3.09x < Digit V Phalanx 2 n/a n/a -- n/a Digit II Phalanx 3 y = 1.38x < Digit III Phalanx 3 y = 2.92x Digit IV Phalanx 3 y = 2.52x

35 APPENDIX 7. Least squares regression analyses of proportional changes in areas of sea turtle skeletal elements. Degrees of freedom (DF);Alpha = Analyses were not performed on elements with incomplete data (n/a). Skeletal element Regression Equation R 2 Value DF P-value Humerus y = 11.1 Ox Radius y=4.14x Ulna y = 6.1 Ox Intermedium y = 2.08x Ulnare y = 3.20x Radiale n/a n/a -- n/a Pisiform y = 3.02x Centrale y = 0.83x Distal Carpal 1 y = 2.05x Distal Carpal 2 y = 1.50x Distal Carpal 3 y = 0.68x Distal Carpal 4 y = 3.32x Distal Carpal 5 y = 0.91x Metacarpal I y = 4.76x Metacarpal II y = 3.15x Metacarpal III y = 2.92x Metacarpal IV y = 3.22x Metacarpal V y = 2.50x Digit I Phalanx 1 y = 2.21x Digit II Phalanx 1 y=3.1lx Digit III Phalanx 1 y = 3.38x Digit IV Phalanx 1 y=2.91x Digit V Phalanx 1 n/a n/a -- n/a Digit I Phalanx 2 y = 1.98x Digit II Phalanx 2 y = 3.26x Digit III Phalanx 2 y = 5.70x Digit IV Phalanx 2 y = 1.40x Digit V Phalanx 2 n/a n/a -- n/a Digit II Phalanx 3 y = 2.95x Digit III Phalanx 3 y = 8.65x Digit IV Phalanx 3 y = 1.40x

36 APPENDIX 8. Least squares regression analyses of proportional changes in lengths and areas of penguin skeletal elements. Degrees of freedom (OF) and P-values (Alpha= 0.05) are given. Analyses could not be performed on elements with incomplete data (n/a). L eng1 th s Skeletal element Regression Equation R.l Value DF P-value Humerus y = 1.21x Radius y = 1.06x Ulna y = 1.19x Radial carpal bone n/a n/a -- n/a Ulnar carpal bone n/a n/a -- n/a Metacarpal II y = 0.33x Metacarpal III y = 0.35x Pollex (Digit I) y = 0.27x ".) Digit II Phalanx 1 y = 0.18x ) " Digit III Phalanx 1 y = 0.21x Digit II Phalanx 2 y = 0.04x ) " Areas Skeletal element Regression Equation R.l Value DF P-value Humerus y = 5.40x Radius y = 5.14x Ulna y=4.81x Radial carpal bone y = 1.80x Ulnar carpal bone y = 2.23x ) " Metacarpal II y = 2.10x Metacarpal III y = 0.3 7x Pollex (Digit I) y = 0.31x ) '"' Digit II Phalanx 1 y = 0.48x Digit III Phalanx 1 y = 0.50x Digit II Phalanx 2 y = 0.13x

37 LITERATURE CITED AbramoffMD, Magelhaes PJ, Ram SJ Image Processing with ImageJ. Biophotonics International 11 (7): Ahlberg PE, Milner AR The origin and early diversification of tetrapods. Nature 368: Andrews SM, Westoll TS The postcranial skeleton of Eustenopteron foordi Whiteaves. Trans R Soc Edinb 68: Bertelli S, Giannini P A phylogeny of extant penguins (Aves: Sphenisciformes) combining morphology and mitochondrial sequences. Cladistics 21 : Carr A New perspectives on the pelagic stage of sea turtle development. NOAA Tech Memo NMFS-SEFC-190 Carrier D, Leon LR Skeletal growth and function in the California gull (Larus californicus). J Zool London 222: Clark BD, Bemis W Kinematics of swimming of penguins at the Detroit Zoo. J Zool 188:

38 Coates MI, Jeffery JE, Ruta M Fins to limbs: what the fossils say. Evol Dev 4(5): Darby JT, Seddon PJ Breeding biology of yellow-eyed penguins (Megadyptes antipodes). In: Davis LS, Darby JT, editors. Penguin Biology. San Diego: Academic Press, Inc. p Dingerkus G, Uhler LD Enzyme clearing of alcian blue stained whole small vertebrates. Stain Technol 52: Futuyma, DJ Evolutionary biology. 2nd ed. Sunderland, Massachusetts: Sinauer Associates, Inc. Gardiner DM, Torok MA, Mullen LM, Bryant SV Evolution of vertebrate limbs: Robust morphology and flexible development. Am Zoo! 38: Gilbert SF Developmental Biology Edition. Sunderland, Massachusetts: Sinauer Associates Inc. Gould SJ Ontogeny and Phylogeny. Cambridge, MA, London, England: The Belknap Press of Harvard University Press. 28

39 Herbert C A timed series of embryonic developmental stages of the Adelie penguin (Pygoscelis adeliae) from Signy Island, South Orkney Islands. Brit Antarct Surv B No 14 p Hirayama R Phylogenetic systematics of chelonioid sea turtles. Isl Arc 3: Hui CA Penguin swimming. I. Hydrodynamics. Physiol Zool61(4): Legendre AM Nouvelles methodes pour Ia determination des orbites des cometes. Courcier, Paris. Lohmann KJ, Lohmann CMF Orientation to oceanic waves by green turtle hatchlings. J Exp Biol171:1-13. Louw GJ Functional anatomy of the penguin flipper. J S Afr Vet Ass 63(3): Miller JD Embryology of marine turtles. In: Gans C, editor. Biology of the Reptilia. New York: John Wiley & Sons. p Parsons CW Penguin Embryos. Brit Antarct (Terra Nova) Exped Zoo! Vol IV No 7 p

40 Raikow RJ Locomotor system. In: King AS, McLelland J, editors. Form and function in birds Volume 3. London: Academic Press. p Randall DD, Fallon JF Interdigital Regulation of Digit Identity and Homeotic Transformation by Modulated BMP Signaling. Science 289: Richardson MK, Oelschlager HHA Time, pattern, and heterochrony: a study of hyperphalangy in the dolphin embryo flipper. Evol Dev 4(6): Richardson MK, Jeffery JE, Tabin CJ Proximodistal patterning of the limb: insights from evolutionary morphology. Evol Dev 6(1): 1-5. Rol'nik VV Development of the muscles and skeleton. In: Mills H, editor. Bird Embryology. Leningrad: Keter Press. p Schlosser G, Wagner GP Modularity in development and evolution. Chicago: The University of Chicago Press. Sears KE, Behringer RR, Rasweiler IV JJ, Niswander LA Development ofbat flight: Morphologic and molecular evolution of bat wing digits. PNAS 1 03(17):

41 Shubin N, Tabin C, Carroll S Fossils, genes and the evolution of animal limbs. Nature 388: Smith KK Time's arrow: heterochrony and the evolution of development. Int J Dev Biol47: Sues HD Postcranial skeleton of Pistosaurus and interrelationships of the Sauropterygia (Diapsida). Zoo! J Linn Soc-Lond 90: Wagner GP Homology and the mechanisms of development. In: Hall BK, editor. Homology: The hierarchical basis of comparative biology. San Diego: Academic Press. p Wagner GP, Chiu C-H The tetrapod limb: A hypothesis on i"ts origin. Mol Dev Evol 291: Wassersug RJ A procedure for differential staining of cartilage and bone in whole formalin-fixed vertebrates. Stain Techno! 51: Webb PW Body form, locomotion, and foraging in aquatic vertebrates. Am Zoo! 24:

42 Westneat MW, Walker JA Mechanical design of fin propulsion: kinematics, hydrodynamics, morphology and motor control of pectoral fin swimming in fishes. Proceedings of the Conference on Unmanned, Untethered Submersible Technology. Autonomous Underwater Systems Institute, Durham, N.H. Wyneken J, Salmon M Frenzy and postfrenzy swimming activity in loggerhead, leatherback, and green sea turtles. Copeia 1992(2): Wyneken J Sea turtle locomotion: Mechanisms, behavior, and energetics. In: Lutz PL, Musick JA, editors. The Biology of Sea Turtles. Boca Raton: CRC Press. p Wyneken J The anatomy of sea turtles. NOAA Tech Memo NMFS SEFSC-470 Zou H, Niswander L Requirement for BMP signaling in interdigital apoptosis and scale formation. Science 272:

43 TABLE 1. Comparison of sea turtle and penguin flipper growth and form Sea Turtle Penguin Shape Broader distally; minimally tapered; 5 phalanges present Tapered distally; 2 phalanges present Stylopodium Humerus size increases throughout all observed stages; growth accelerates in late stages (27-29) Humerus size increases slowly; proportional growth similar in all stages observed Zeugopodium & Radius and ulna short with some Autopodium midshaft overlap; retain most carpal, metacarpal, and phalangeal elements; phalanges elongated distally Radius and ulna long, flattened and remain separate; retain just two each of carpals, metacarpals, and digits; phalanges do not elongate as much as in sea turtle Digital Plate Rounded at stage 22; elongation starts at stage 24 Oval and elongated, not rounded as in sea turtle; elongation is proportionately uniform 33

44 A B c Fig. 1. Turtle and penguin flippers differ in skeletal form and pattern. The basic tetrapod forelimb pattern is shown for comparison in A (left forelimb, ventral view). All have a humerus, radius, ulna, carpals, metacarpals, and phalanges. Sea turtles (B ; right flipper, dorsal view) form the flipper by elongation and some flattening of the carpals, metacarpals and phalanges. Penguins (C; left flipper, ventral view) have reduced and fused carpals and metacarpals which form most of the elongated and flattened carpometacarpus. The radius and ulna are also flattened and contribute to the flipper blade. The digits (II and III) form comparatively little of the penguin flipper (Louw, 1992). 34

45 Fig. 2. Photo ofleft sea turtle flipper (dorsal view) without the humerus showing elements measured in C. caretta ( 29). H=humerus, Ra=radius, Ul=ulna, I=intermedium, Rc=radiale=radial carpal bone, Uc=ulnare=ulnar carpal bone, C=centrale, P=pisiform, Dc=distal carpal, M=metacarpal, Ph=phalanx. The pollex is the top-most digit. 35

46 20 ~ ( I Fig. 3. Photo of a stage 42 (Em) penguin flipper (left, dorsal view) showing elements measured. H=humerus, Ra=radius, Ul=ulna, Rc=radiale=radial carpal bone, Uc=ulnare=ulnar carpal bone, M=metacarpal, Ph=phalanx. 36

47 10 20 Fig. 4. Detailed view of the cleared, stained left flipper (dorsal view) in a stage 37 S. demersus. The pollex or digit I (labeled *)is present during penguin flipper development but later becomes fused to metacarpal II (Louw, 1992; Parsons, 191 0). 37

48 Proportional Change in Length ofturtle Humerus Proportional Change in Area of Turtle Humerus " c: ~ 10 0 ci. 5 e H Ill c: ~ 100 (.) ci. 50 e Embryonic Embryonic Proportional Change in Lengths of Turtle Radius and Ulna Proportional Change in Areas ofturtie Radius and Ulna " c: ~ 10 0 ci. 5 e R -+-U -.-R -+-U ~ 22 n ~ ~ ~ v u 29 Embryonic Embryonic Proportional Change in Lengths of Turtle Wrist Elements Proportional Change in Areas ofturtle Wrist Elements " c: ~ 10 0 ci. 5 e U...-r c: ~ 20 0 ci 10 e n. 0 --I p c Embryonic Embryonic Proportional Change in Lengths of Turtle Distal Carpals 1-5 ~ ~ n ~ ~ ~ v ~ ~, &34 c: ~ 20 0 ci. 10 e o 11. Proportional Change in Areas ofturtle Distal Carpals 1-5 ~ 22 n ~ ~ ~ v u 29 Embryonic Embryonic Fig. 5. Proportional changes in size of humerus, radius, ulna, wrist elements, and distal carpals of C. caretta (stages 21-29). 38

49 ~ 15 c: ~ 10 () ci. s e o 0.. Proportional Change in Lengths of Turtle Metacarpals I V Embryonic, -+- U -.- II _.._ rv v &40 i 30.1:: 0 20 ci 10 0 ~ 0 Proportional Change in Areas of Turtle Metacarpals 1-V Embryonic..._ --- rv..,_ v & 25 ; 20 l: 15 ~ Q: 0 Proportional Change in Lengths of Turtle Phalanx 1 ~ 22 n ~ ~ ~ D ~ 29 Embryonic -+- I & 25 i 20.1:: 15 ~ 10 g. 5 ~ 0 Proportional Change in Areas of Turtle Phalanx Embryonic Proportional Change in Lengths of Turtle Phalanx 2 Proportional Change in Areas of Turtle Phalanx 2 ~ 20 ; 15 'fi e a. 0 ~ 22 n ~ ~ u D u 29., rv v & 40 ; 30.1:: t,) 20 ci 10 0 ct I -+- II -+- II rv v Embryonic Embryonic Proportional Change in Lengths of Turtle Phalanx 3 Proportional Change in Areas of Turtle Phalanx 3 ~ 15 c: ~ 10 u ci. s a. e Embryonic ,_ rv &~ ; 30.1:: () 20 ci. 10 e o Embryonic Fig. 6. Proportional changes in size of metacarpals and phalanges of C. caretta (stages 21-29). 39

50 Proportional Change in Length of Penguin Humerus Proportional Change in Area of Penguin Humerus &a ~ 6 ~ 4 ci.2 0 0: a H ~40 ; 30 r ci. 10 ~ 0 Q H Embyronic Embryonic Proportional Change in Lengths of Penguin Radius and Ulna Proportional Change in Areas of Penguin Radius and Ulna &a ; 6 ~ 4 Q.2 0 0:. 0 ~ u 30 n ~ ~ &30 c ~20 0 ci. 10 e o Q Embryonic Embryonic Proportional Change in Lengths of Penguin Metacarpals Proportional Chango in Areas of Penguin Metacarpals II HI II 3 Cll 2.5 ; 2 '5 1.5 ri 1 e o.s IL II Ill 12 & 10 ; II '5 6 ri 4 e z IL u u ~ ~ " ~ ~ Embryonic Embryonic Proportional Change in Areas of Penguin Carpals Gl 15 Cll c ~ 10 (J Q. 5 e Q t _...u Embryonic Fig. 7. Proportional changes in size ofhumerus, radius, ulna, metacarpals, and carpals of S. demersus (stages 28, 30, 37), E. minor (stages 41, 42), and unknown penguin species (stages 25, 42). Lengths were not measured for the carpal elements. 40

51 Proportional Change in Length of Penguin Pollex!Digit I) Proportional Change in Area of Penguin Pollex (Digit I) 1:1 = 2 ; 1.5 ~ u.. 1 ~ 0.5 a p ~ 2.5 ; 2 ~ 1.5 u 1 ~ 0.5 a. 0 ~ ~ ~ u ~ ~...- P Embryonic Embryonic Proportional Change in Lengths of Penguin Phalanx 1 Proportional Change in Areas of Penguin Phalanx 1 = 2 01 ; 1.5 ~ u 1 ~.. ~ 0.5 a Embryonic -+- II ~ Q c: 3 II ~ 2 ~ 1 a Embryonic ---Ill Proportional Change in Lengths of Penguin Phalanx 2 Proportional Change in Area of Penguin Phalanx 2 Gl 2 Q ; 1.5 ~ u 1 ~ a. 0 ~ II Q 2 Q ; 1.5 ~ 1 ~ a Embryonic Embryonic Fig. 8. Proportional changes in size of the pollex and phalanges of S. demersus (stages 28, 30, 37), E. minor (stages 41, 42), and unknown penguin species (stages 25, 42). 41

52 10 00 Humerus 0.8 E ~ 0.6 i 03 r; "' (lj I "' I I OA Humerus I 1 <! 1 1 c: 0.4 e 02 _J ;s 71 :1Jl D 00 Radius OB 0 4 Radius ~ 06 '1 03 ",!ii OA!'! 0.2 _J <! l "' "' f OS Ulna OS 0.4 ", I!' "' j c: OA 1 N~ 0.3 <! 0.2 Ulna I 02 I 0 1! I OD "' Fig. 9. Growth trends of humerus, radius, and ulna in C. caretta as a function of stage (stages 21-29). Error bars are shown. 42

53 025 OO>J Centrale Centrale : ; 0.01!5 "' c:.:t '" _J '" 6 I A 20,. 27 "' ,. 26 Z7 I "' 29 "' lntermedium ~ 0.1, N '";" O.D1S "' c: 0.10 I i!! _J '" I <( S O.Q ~ lntermedium ", MD 020 Pisiform 0020 ",. 2G Pisiform ~ 0~ "' "E M:!l.;,;_ '" l 001e5 C) c: 0., < _J '" 001() ", , "' Q20 Radiale o.mo Radiale E o.1e.;,;_ ~ c:..j '" QOO QOO Ni "' <( "' t,. " A 25 Z7 "' 29 "' Ulnare l 0025 Ulnare : 0.15 Ni 0020 _J "'.<:: 0015 "' i!! c: 010 <( '" ", Fig. 10. Growth trends ofwrist elements in C. caretta as a function of stage (stages 21-29). Error bars are shown. 43

54 0.14 Distal Carpal ]' CJ) ~ loooe ~ "' coos Distal Carpal 1 I 0.00 '---~------~ ' ~ ~ ~ ~ u ~ ~ n ~ ~ ~ A 25 2S X ) 0.14 Distal Carpal Distal Carpal E o.1o ~ 0.06 ~ 006 _.J 004 I I ~E ~ ooce "' ~ E o.1o -"'-.c > ODS ~ ~ D D H B ~ V ~ ~ ~ o,. Distal Carpal ~ 0010 ~ OOCE "'.?_ '" ooo; 0.01) ~ Distal Carpal 3 W ~ D D H ~ V ~ ~ D '---~ ~-~-~~-----< zo ~ 2S 27 2S Distal Carpal Distal Carpal E o.1o ~ ~ ;, :E 0.06 _.J ~ " D D H B ~ n ~ ~ ~ ~E ~ I!! "'..: oooa ~ I 20 ~ Distal Carpal Distal Carpal 'E o.10 u O.CE CJ) ~ O.C6 _.J j "'!"..: < '---~------~ ' ~ ~---~~-----< ~ Fig. 11. Growth trends of distal carpal elements in C. caretta as a function of stage (stages 21-29). Error bars are shown. 44

55 ~ 0.3.<: 0, c: 0.2 _j "' 0.1 Metacarpal I 0.04, ,;-;---;-----;-;-----,-----, Metacarpal I 003 i -002 "' ~ !----~ ';.----~~-~----< >J Z v E ~ 0.3 "' ~ 0.2 _j Metacarpal II 0~.----~~~--~~------, Metacarpal II ~ Z3 24 2!l I "' 27 2S I I "' 05 Metacarpal Ill 004 Metacarpal Ill 0.4 E.B. 0.3.<: ""' ~ 02 _j ~E' B "' 1! <>: ~ >J Z3 I "' I ! I "' 27 2S "' 3) 05 Metacarpal IV 004 Metacarpal IV 04 ~ 03 t a; 02 1 _j j "' Q) ~ oo+-~-t-~-t---~----~ n v & 2~ E o.3 B O~r , Metacarpal V C1 c: Q) I _J 0.1 I I "' 27 2S I "' 30 0~,-----~~~--~~~-----, Metacarpal V oro 0( I OIXl S! "' 30 Fig. 12. Growth trends of metacarpal elements in C. caretta as a function of stage (stages 21-29). Error bars are shown.

56 =.,----:-,---;--=:---:-----: , Digit I Phalanx owo,----=~~~~~-~~---. Digit I Phalanx 1 002l5 ~ 0.3 g, ~ 0.2 _j ! I! ~ ~ ~ ~ ~ ~ v ~ ~ ~ 0.5,------=-,----,----,-c-=-,-----,------= , Digit II Phalanx w 21 ~ n ~ ~ ~ n ~ ~ ~ OUD,----=~~~~~--~-- Digit II Phalanx ~ ~5 0.2 CJ) c: (1) _j 0.1 I ~ (1) "' ~ ~ 0.000! ~ n n u ~ ~ v ~ ~ ~ 0.5, =-,----,-,--,-:-=c---,---, , Digit Ill Phalanx ~ 03 g, 0.2 c: (1) _j 0 1 I i "' (1) ~ ~-~--- ;-~~-,----~-~-1 w ~ n n u ~ ~ v ~ ~ ~ 00~,----=~~:-:-:--=-:----:---, Digit Ill Phalanx ~ l W ~ D D ~ ~ ~ V ~ ~ ~ 0.5,-----=~~----=--=-:---:----:----. Digit IV Phalanx ~ CJ) c:...j "' 0.1 w ~ n n u ~ v ~ ~ ~ 0~ =~~~~-:-~---,---. Digit IV Phalanx 1 002~ i - 001~ I I ~ ~ D D H ~ ~ V ~ ~ ~ o =-,---~~~----,----, Digit V Phalanx 1 DA w ~ ~ ~ ~ 25 ~ v ~ ~ ~ 0030,.----=~-::-c:-;-=:----;---;----, Digit V Phalanx 1 002~ ~ 03, 02 a;...j ! ~i Ill ~ DOW 001~ W ~ D D H ~ V ~ ~ ~ ~-~-----~~-,----~~----=-1 W ~ D D H ~ ~ V ~ ~ ~ Fig. 13. Growth trends ofphalanx 1 in C. caretta as a function of stage (stages 21-29). Error bars are shown. 46

57 08,-----~~~~~--~~---- Digit I Phalanx 2 OS 0~,-----~~~~--~--~ Digit I Phalanx ~~ :;------~--~-1 ~ ~ n n u ~ ~ v ~ ~ ~ ~~------;..- ;;c.._~ ~ ~ ~ D M ~ ~ V ~ ~ ~ 'E -"- Iii "' _J Digit II Phalanx 2 004,-----~:-:-:--:-c:--=-:---:-----:: Digit II Phalanx I ~ ~ ~ D 24 ~ ~ ~ ;..-~ W ~ n D M ~ 26 V ~ ~ ~ I 08 O.S 'E -"- ~ 0 4 c Digit Ill Phalanx 2 I Ni ro I!' <( 003 I I Digit Ill Phalanx 2 I ~ "' c ~ ~ ~ D 24 ~ ~ V ~ ~ ~ os,-----~~~~~~--~----, Digit IV Phalanx 2 OS OA _J "' 02 1 ooo ~--~--~---; ~ ~ 26: Z ) 004,------:::-:--:-:-:-:-:--::::-:-~----::---. Digit IV Phalanx w ~ n n " ~ ~ v ~ ~ ~ 000 ~ 21 n 23 " ~ ~ ~ ~ 0.8,-----~~:-:--c:-=-:---:----:;:-----, Dig it V Phalanx Digit V Phalanx 2 'E -"- "' c ~ 02 i ro!!! <( oo +-~~----~ :;I I ~~ W ~ ~ D ~ ~ ~ V ~ ~ ~ 000 ~ 21 n 23 I " ~ 2S ~ "' I Fig. 14. Growth trends of phalanx 2 in C. caretta as a function of stage (stages 21-29). Error bars are shown. 47

58 :::-:---:-:--:-:-c=-:----:------::c Digit II Phalanx ~ 0.2 c: _] "' 0.1 OD20,.-----=-:-:-,.---,-,--=:-:---:----=-----, Digit II Phalanx i.._, OD10 ro "'.;;: 0005 ~ ~-~-~-~-~-~---~ ,---~~~~~-,------,-----, Digit Ill Phalanx 3 OJ 02! ~-~-~-~-~ < , ,=-:---:-:--::--c:-=:---:----=---, Digit IV Phalanx I j ro 2! <! D ~ ~ ~ u ~ ~ n ~ ~ ~ =c----,-----,-cc---=-,-----, , , Digit Ill Phalanx 3 001: I I OCDO +--~~~-~+- ' ~I-~~~ S "' =-::---:-:--:-:--:-::,..,----:------:::-----, Digit IV Phalanx ~----~-~-~-~-~-~--4 D ~ ~ ~ ~ ~ n ~ ~ ~-~---~ ;:.._~-~---~ 'Z J I I Fig. 15. Growth trends of phalanx 3 in C. caretta as a function of stage (stages 21-29). Error bars are shown. 48

59 Humerus e ~.. ~ c '" Humerus Radius Radius ~...c ;;,. I Ulna " Ulna " Fig. 16. Growth trends ofhumerus, radius, and ulna ins. demersus (stages 28, 30, 37), E. minor (stages 41, 42), and unknown penguin species (stages 25, 42) as a function of stage. Error bars are shown. 49

60 Radial Carpal Ulnar Carp~l "' stage " Metacarpal II Metacarpal II.. oo+--~~~~~-~~~~~ ~ :l ~ n '' n,,. ",. 40,, u u,-----:-:----:-----:-:-:-:-----, Metacarpal Ill Metacarpal Ill e ~. i stage Fig. 17. Growth trends of carpal and metacarpal elements ins. demersus (stages 28, 30, 3 7), E. minor (stages 41, 42), and unknown penguin species (stages 25, 42) as a function of stage. Lengths were not measured for carpal elements. Error bars are shown. 50

61 Pollex (Digit I) Pollex (Digit I) ""'., =:----:-:--:-:--=:---:--: Digit II Phalanx I.I Digit II Phalanx I! Dig~ Ill Phalanx I.l stage Digit Ill Phalanx I Digit II Phalanx 2 Digit II Phalanx 2 stage Fig. 18. Growth trends of the pollex and phalanges ins. demersus (stages 28, 30, 3 7), E. minor (stages 41, 42), and unknown penguin species (stages 25, 42) as a function of stage. Error bars are shown. 51

62