INFORMATION TO USERS

This was produced from a copy of a document sent to us for microfilming. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the material submitted.

The following explanation of techniques is provided to help you understand markings or notations which may appear on this reproduction.

1. The sign or “target” for pages apparently lacking from the document photographed is “Missing Page(s)”. If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting through an image and duplicating adjacent pages to assure you of complete continuity.

2. When an image on the film is obliterated with a round black mark it is an indication that the film inspector noticed either blurred copy because of movement during exposure, or duplicate copy. Unless we meant to delete copyrighted materials that should not have been filmed, you will find a good image of the page in the adjacent frame.

3. When a map, drawing or chart, etc., is part of the material being photo­ graphed the photographer has followed a definite method in “sectioning” the material. It is customary to begin filming at the upper left hand corner of a large sheet and to continue from left to right in equal sections with small overlaps. If necessary, sectioning is continued again—beginning below the first row and continuing on until complete.

4. For any illustrations that cannot be reproduced satisfactorily by xerography, photographic prints can be purchased at additional cost and tipped into your xerographic copy. Requests can be made to our Dissertations Customer Services Department.

5. Some pages in any document may have indistinct print. In all cases we have filmed the best available copy.

University Microfilms International 300 N. ZEEB ROAD. ANN ARBOR. Ml 48106 18 BEDFORD ROW. LONDON WC1R 4EJ, ENGLAND 8001783

M e s s in g e r, D a v id S t e v e n

MECHANICS OF WALKING AND SWIMMING OF THE ANAS PLATYRHYNCHOS

The Ohio State University Ph.D. 1979

University Microfilms International300 N. Zeeb Road, Ann Arbor, MI 48106 18 Bedford Row, London WC1R 4EJ, England

Copyright 1979 by Messinger, David Steven All Rights Reserved PLEASE NOTE: In all cases this material has been filmed in the best possible way from the available copy. Problems encountered with this document have been identified here with a check mark .

1. Glossy photographs______2. Colored illustrations______3. Photographs with dark background______4. Illustrations are poor copy______5. P-int shows through as there is text on both sides ______of page 6. Indistinct, broken or small print on several pages throughout

7. Tightly bound copy with print lost in______spine 8. Computer printout pages with indistinct______print 9. Page(s) _____ lacking when material received, and not available from school or author______10. Page(s) ______seem to be missing in numbering only as text follows______11. Poor carbon copy______12. Not original copy, several pages with blurred __ type 13. Appendix pages are poor ______copy 14. Original copy with light ______type 15. Curling and wrinkled pages______16. Other ______

University Micnjfilms international

300 N. 2 E E 3 RD.. ANN ARBOR. Ml 48106 1313) 761-4700 MECHANICS OF

WALKING AND SWIMMING OF THE DUCK

ANAS PLATYRHYNCHOS

DISSERTATION

Presented in Partial Fulfillment of the Requirement for

the Degree Doctor of Philosophy in the Graduate

School of The Ohio State University

by

David Steven Messinger, B.S., M.S., M.S.

Vc ->V - k 4c -k

The Ohio State University

1979

Reading Committee: Approved By

Dr. Abbot S. Gaunt

Dr. George H. Dalrymple

Dr. Roy A. Tassava

Dr. Walter G. Venzke Advisor ACKNOWLEDGMENTS

I am grateful to Dr. Abbot S. Gaunt for the oppor­

tunity to pursue this study in his laboratory. Dr. G.

Dalrymple and Dr. R. Tassava offered encouragment and

helpful suggestions. I am indebted to Dr. J. Munford,

Mr. and Mrs. Clapp, M. Dautorus and M. Anderson, for

their able assistance during my experiments.

I thank Mr. A. Macias for the use of his photographic

expertise. Many technical problems were resolved through

the assistance of Dr. F. Heft. Dr. S. Weber provided

valuable assistance in the preparation of the manuscript.

Without the encouragement of Mr. G. Cronin, Dr. J.

Hostettler, Dr. W. Butts, Dr. L. Sohaki and Dr. W. Wilson,

Dr. G. Goslow and Mrs. Vedder this study would never have

been attempted.

Support came from the Ohio State Department of Zoology

the Josyln Van Tyne Memorial Fund, and the Frank M. Chapman

Award from the American Museum of Natural History. Addi-

' tional support was provided through National Science Founda

tion grants GB40069 under Dr. A.S. Gaunt and RF4074A1 under

Dr. R. McGhee. VITA

February 13, 1950...... Born - Amityville, New York

1972 ...... B.S. State University of New York, Oneonta

1974 ...... M.S. Northern Arizona University, Flagstaff, Arizona

1977 ...... M.S. The Ohio State University, Columbus, Ohio

1978-1979...... Teacher, Columbus Public Schools

PUBLICATIONS

"Ankle Extensor Activity in the Walking Cat." Amer. Zool. 14:1267 Abstract, 1974

FIELD OF STUDY

Major Field: Zoology, Anatomy, Physiology

Studies in . Professor George E. Goslow, Jr.

Studies in Animal Locomotion. Professor Abbot S. Gaunt TABLE OF CONTENTS

Page

ACKNOWLEDGMENTS...... ii

VITA...... iii

LIST OF FIGURES...... vi

LIST OF PLATES...... vn

INTRODUCTION...... 3

Chapter

I. MATERIAL AND METHODS...... 14

Animals...... 14

Recording Techniques...... 15

Observation Methods: Walking...... 19

Observation Methods: Swimming...... 21

Analytical Methods...... 23

Stereo Viewing...... 23

Stick Figures...... 27

II. RESULTS...... 29

Walking: Knee, Ankle and Foot Angles 29

Walking: Front and Rear Views...... 32 iv Page

Swimming: Knee, Ankle and Foot Angles.... 33

III. DISCUSSION...... 35

Comparison of the Joint Angles During Walking and Swimming...... 35

Walking and Swimming Compared...... 39

Comparison of the Intertarsal (Ankle) Angles of the Walking Duck, Pigeon and ...... 43

IV. CONCLUSIONS...... 47

V. FURTHER INVESTIGATIONS...... 49

VI. BIBLIOGRAPHY...... 50

VII. APPENDIX...... 53

v LIST OF FIGURES

Figure Page

1. Walkway for Observing Walking...... 53

2. Tank for Observing Ducks Swimming...... 54

3. Diagram of Duck's Leg with Joint References... 55

4. Stick Diagrams of the Walking Duck...... 56

5. Stick Diagrams of the Swimming Duck...... 58

6 . Graph of Knee, Ankle and Foot Angles 60 During the Walk Cycle (Graph #1)......

7. Graph of Knee, Ankle and Foot Angles During the Swim Cycle (Graph #2)...... 61

8 . Graph Comparing Knee Angles of the Walking and Swimming Duck (Graph #3) ...... 62

9. Graph Comparing Ankle Angles of the Walking and Swimming Duck (Graph # 4 ) ...... 63

10. Graph Comparing Foot Angles of the Walking and Swimming Duck (Graph #5)...... 64

11. Graph Comparing Ankle Angles of the Walking Duck, Pigeon, and Gull (Graph # 6 )...... 65

vi LIST OF PLATES

Plates Page

I. Rear and Lateral Views of Walking with Lateral Views of Swimming...... 66

II. Rear View of Walk with Lateral View of Swim...... 68

III. Lateral View of the Walk...... 70

IV. Moderately Fast Swim...... 72

V. Closeups and Multiple Exposures of the Walking Duck...... 74

VI. Frontal and Lateral Views of the Walk...... 76

VII. Fast and Slow Swimming...... 78 ABSTRACT

The mallard duck, Anas platyrhynchos was studied to compare its walking and swimming movements. Special equip­ ment was constructed so that the subjects could be observed under conditions which would insure that the movements recorded would represent the animal in a near natural state

Video recordings were found to be the most appropriate means of studying the fast locomotor actions of walking and swimming ducks. Conventional techniques were modified through the use of lighting by a xenon flash as described within this report.

The data collected during this investigation were analyzed using a new method which allows the viewer to see the photographic images in such a way that they appear to b three dimensional. This is a new technique that presently provides only qualitative rather than quantitative data.,

For this reason, more conventional stick figures were made from tracings made from photographs of the original video tape recordings. Measurements were made of the knee, ankle and foot angles from which graphs were made.

The movements of each joint are described, and all the angular displacements of the joints compared with each other. Relationships are described and explained. 2

During the walk cycle it was found that the knee and ankle move synchronously while the foot (intertarsal) angle behaves completely out of phase. The same was found true for the swimming duck.

It was found that the swimming duck has a knee that remains at a relatively fixed position during slow swimming.

The range of motion of the knee can however, be seen to increase with speed. Again, the foot and ankle angles when graphed prove to have curves that appear as mirror images of one another. These joints appear to be behaving almost 180° out of phase.

Differences in the movements of walking and swimming are explained as being the result of adaptations by the duck to the physical characteristics of the respective media involved. Of prime importance are those adaptations made to allow the animal to cope with the factors of weight

(on land) and drag (in water).

It is concluded that the similarities present in the movements of the walking and swimming duck can best be explained as being the result of the modification through time of a basic pattern of walking movements to meet the challenge of the aquatic medium. Comparisons made with data from the literature for other aquatic and cursorial indicate that these assumptions are valid in a variety of cases. INTRODUCTION

Birds have adapted to several different types of loco­ motion. Various birds walk, hop, run, swim, and fly.

Perhaps because man can do all of these activities except

fly, he has centered most of his attention on the flight of birds. Whereas, the of birds have been studied

in an attempt to understand their function, the hindlimbs have been examined mainly by taxonomists who have attempted

to use their structure to study evolutionary relationships

(see George and Berger 1966 for a review). Taxonomists have

paid little attention to the functional aspects of the avian

locomotor system. Only recently have functional morpholo-

gists begun to examine the dynamic qualities of the loco­ motor systems in living birds.

Most previous studies were based on data collected from dissections of dead animals, rather than from observations made of living animals. Those descriptions that are avail­

able are few and usually qualitative in nature. Until recently few recordings and measurements were made of nofl

flying birds.

The list of taxonomically oriented papers that deal with the avian hindlimb is enormous and beyond the scope of

this paper. The following papers are mentioned to give the reader a familiarity with the kinds of studies that have been conducted in the past. The questions asked in these studies are, for the most part, limited in scope, as are the investigative techniques. 5

Over the preceding decades, a great deal of data regard

ing avian locomotion has been collected by many people.

Early investigators were limited by the technology available

at the time. Functional anatomical and physiological tech­

niques have improved swiftly in recent years. Modern elec­

tronics and high speed photography as well as other record­

ing techniques have opened new vistas to those scientists

who dream of understanding the subtleties of the locomotor

systems of organisms. The rewards for the investigator who

pursues this field of inquiry will be the personal satisfac­

tion of providing data which other scientists will be able

to base their studies upon.

The avian locomotor system is particularly interesting

in that it addresses the problem of . Bipedalism

is a phenomenon peculiar to , and birds.

Bipedalism in man has been studied rather extensively for

evolutionary and clinical reasons. This paper will not

address itself to comparisons between bipedalism in man and

other organisms. One basic reason lies in that man is a

animal whereas birds are . While

the mode of locomotion is similar, the structures involved

' perform analogous roles not particularly appropriate to the

questions that will be asked here. The selective pressures

responsible for the adoption of bipedalism by man, birds,

and reptiles are very different. Three impetuses for the

adoption of a bipedal have been suggested: 1 ) for speed in the case of the reptiles; 2 ) for freeing of the fore­

limbs for manipulations and feeding as in man, and 3) the

adaptation of the forelimbs for flight as in birds. A more

detailed discussion is given by DuBrul (1962).

One of the first functional studies of locomotion

relevant to this paper was by Stolpe in 1932. Stolpe in­

vestigated avian and related his obser­

vations to the function of the leg. He discusses the adap­

tations of the hind for birds that live in a variety

of habitats. His work is significant in that it attempts

to explain the acquisition of a function in terms of the

environmental pressures with which this function allows the

animal to cope.

Miller (1937) studied the Hawaiian , a terrestri­

ally adapted bird. He describes the gait of this goose

from unaided observations and some simple snapshots. His

work illustrates some of the technical limitations under

which these men had to work and includes the wet volumes of

the goose's leg muscles along with their descriptions.

Miller's paper shows a classical approach to the study of

form and function in that he is primarily concerned with

' taxonomic questions.

In 1946, Fisher, recognizing the importance innerva­

tion has in the study of homologies, published information

regarding both the muscles and the nerves that supply them.

He was aware that though muscles may shift position and 7

function, the nerves that supply them may not (see

Hildebrand, 1974 for a discussion). He therefore included

this information in his descriptions of the and leg

muscles of cathartid vultures.

Other taxonomic studies on birds were conducted by

Burt (1930), on woodpeckers, and by Richardson (1942) who

studied modifications for tree trunk foraging in birds that

are unrelated. Richardson shows excellent insight into the

rationale behind functional studies, and his thoughts will

be commented upon later in this paper.

Wilcox, (1952) did a taxonomical study on another

aquatic bird, the . Berger (1953) worked with cuckoos,

and Hartman (1961) examined the wet weights of leg muscles

of some birds. Both employed dissection techniques and did

not perform any experiments on living birds.

In 1967, Owre watched the and the ,

both aquatic birds that dive, and reported his observations.

He used no underwater recording techniques, so no data are

available for the scrutiny of the reader.

Dissections provided most of the data for these early

studies. It was not until 1965 that Spring employed cinema­

tography, X-ray, photography, and dissections to make func­

tional anatomical observations. Spring studied the climb­

ing and pecking adaptations of some woodpeckers.

Raikow (1970) described both aquatic and terrestrial

locomotion in three species of ducks. The mallard was one of the ducks he studied and his observations were important

starting points for this paper. In his rather taxonomic

study, he describes the swimming movements of ducks that he

watched swim in a lake or marina. No recordings were made

of these movements, so his findings are subject to verifica­

tion. Raikow describes the legs of a swimming duck as being

held in an abducted position, the degree of which is charac­

teristic of the species being observed. His description of

walking is cursory and brief.

More recently, Rylander and Bolen (1974) investigated

the structures associated with walking, swimming and feed­

ing of three species of whistling ducks (Dendrocygnae).

They used cinematography to describe the of these

ducks. They then made taxonomic conclusions as to the rela­

tion between these species.

Prang and Schmidt-Nielson (1970), while measuring the

energetic cost of swimming in ducks, made measurements of

swimming speeds and briefly commented on the duck’s stroke

as it relates to the bird’s speed. More will be said of

their paper in the discussion section.

In 1971, Cracraft examined several interesting aspects

' of the avian locomotor system with specific reference to

the common pigeon (Columba livia). He examined the major

morphological features of the leg joints using modern tech­

niques of gross anatomy and microscopy and followed these

observations by a description of the hind limb muscles. 9

The fiber lengths and arrangements, attachments, angles of pinnation, and the histology of the muscles were also dis­

cussed. These finds were treated in a classical manner in

that they are mostly descriptive in nature. It is in the

latter section of Cracraft's paper that his work begins to

gain stature, and becomes more than just a collection of data. Cracraft presents a high speed motion picture analy­

sis of the pigeon's gait as it performs a slow walk, moder­

ate walk and a run. In this section, Cracraft discusses

the angular displacements of the joints and their angular

velocities. Some of the characteristics of landing are

also mentioned.

Cracraft assembles his data to construct an activity

sequence for the muscles used by the pigeon in locomoting.

This is an interesting attempt at using information gained

from non-living specimens to make definite statements re­

garding the muscular actions as they may be occuring in

live birds. Although his conclusions are logical, based on

his data and assumptions, the validity of his conclusions

are certainly subject to question and further experimental

investigation. The results of an electromyographic analysis

of the hind limb musculature of the pigeon will someday

confirm or dispute Cracraft's conclusions as to the func­

tions of these muscles in a living pigeon. Nonetheless,

Cracraft's paper gives us some good data for making more

comparisons. 10

Dagg (1977) provides us with more data with which we

may compare the locomotor activity of many more birds. In

her study of the Silver Gull, Dagg filmed and examined the

angles made by the tibiotarsal-tarsometatarsal joint of

each leg during walking. The angles were then correlated

with the bird's speed. An attempt is made to find a rela­

tionship between the speed of a step with the size of the

bird being studied. Head bobbing as exhibited by some birds

and not others is also commented on though the conclusions

are not definite. Dagg also comments on hopping, a form of

bipedalism used by some birds. In addition to work done on

the Silver Gull, 24 other species of small and large birds

were examined as to their gaits. The data are not very

detailed, yet allow us to make comparisons with the pigeon

and the duck.

Clarke and Alexander in 1975 did an elegant analysis

of the forces involved in running by the quail. They com­

ment on the locomotor process in terms of the physicist.

Their data provide little information of use to this paper

in regards to determining leg movements. However, their

observations do provide useful information from which I

'will draw some conclusions.

It can be seen that early studies usually attempted to

compare the morphological features of several species in an

attempt to draw evolutionary and taxonomic conclusions.

More recently single species studies have become popular. 11

In single species studies, the characteristics of a given animal as it locomotes are studied in detail. Conclusions are then made as to the relationship between the observed morphological features and the functions they perform. The result of all the previous studies is that we now have a considerable amount of data from which we may make new comparisons.

In this study, interspecific and intraspecific com­ parisons were made. The common wild type mallard duck,

(Anas platyrhynchos) was studied as it walked and swam.

Comparisons were also made between the data presented in the paper and the data presented in the literature.

Richardson (1942) discussed the basic rationale behind most functional anatomical studies. He stated:

Functional anatomical studies may deal with just one kind of animal, but ordinarily several forms are compared. Such comparisons offer a convenient and necessary check on interpretation and disclose the evolutionary significance of the specialized struc­ tures. In general, there are two types of comparative functional-anatomical studies: one of the closely related forms, the other of distantly related forms. The first type leads to an understanding of the phylogeny of a well-circumscribed group or of the evolution of its adaptations. This is most often accomplished by analyzing varying degrees of the same adaptations in the related forms... In the study of distantly related animals, one of the chief aims is to establish the validity of qualitative and obvious adaptations by showing their occurrence in unrelated species of the same habits. This is essentially a consideration of convergent or of parallel adaptation.

Another interesting case exists. We may also compare distantly related species that do not share common ecologi­ cal characteristics. If we should find, despite these large 12

differences in habits and morphology, that the species still

maintain strong functional similarities, then we may con­

clude that the selective advantage granted the animal by

the original adaptation was, and is, so strong that modifi­

cations of the basic pattern have been selected against.

Species do evolve, and birds have adapted to many

different factors. In order to maintain a given function,

such as a gait, despite other morphological changes in say

the neuromuscular or skeletal system, we must imagine that

the entire complex is relatively flexible: that is, it will

adapt to changing conditions over a long period of time.

On the other hand, once a pattern is established, as biped­

alism has been established long before birds that swim

evolved, the conservative nature of the nervous system

would tend to maintain the original patterns of movement.

For these reasons I believe that differences in movements

within a species or between species will be relatively

slight, or at least modifications of a basic pattern.

Two ways of testing this thesis come to mind. First,

one can take an animal such as the duck that is adapted to

two media and examine its movements in each. Whether swim-

^ming or walking, the duck must use the same morphological

components. We are comparing the duck with itself, so in a

sense, it becomes its own constant. By changing the media

in which it locomotes from land to water and back, we are

forcing the same anatomical and physiological systems to 13 operate under different conditions. We can thus expect to gain some insight into the relative plasticity of the system.

Another approach is to take the first animal, the duck, and compare it with other birds. Birds can be chosen that are entirely terrestrial, or have adapted to swimming in varying degrees. Using the data supplied by the literature, especially that provided by Dagg and Cracraft, I will test the thesis that the basic locomotor pattern is so entrenched in the avian line that only slight modifications of the basic pattern will be found. MATERIALS AND METHODS

Animals

The early phases of this study consisted of visual and

photographic observations of domesticated white ducks.

Various techniques were employed in an attempt to find a method of recording and analyzing both walking and swimming.

At the same time, both newly hatched ducklings and adult ducks were used as subjects in an attempt to discern whether young ducklings would prove a more convenient animal to

study. It was hoped that ducklings could be used because of their small size and minimal food requirements and would not tax the spacial or monetary resources of our laboratory.

The ducklings were not used in the final analysis; they were too fragile. Final data collections were made using

six adult male wild type mallards (Anas platyrhynchos) obtained from a local propagator. These animals were pin­

ioned, which did not change their behavior in any way.

Only male ducks were used because they survived caging better than females and no discernable differences could be detected in regards to the locomotor behaviors of the

two sexes.

14 15

Recording techniques

Initially 16mm films at 32 frames per second (fps) were made of the original dozen ducklings and ten adults.

Some still 35mm photographs were taken of some adult ducks utilizing a xenon flash. Both of these recording methods were abandoned for several reasons.

Motion pictures proved too inconvenient, time consum­ ing, and costly. The images that resulted from this method did not give sufficient resolution, and several days would elapse between the time of exposure and final development.

Perhaps the most inconvenient aspect of conventional motion picture photography is the lack of any immediate means to monitor the quality of the data collected. If a sequence was not taken correctly or if there was a mistake made at the developing laboratory, days would elapse before the error or oversight could be rectified. Such delay was a serious drawback because one had to consider the set-up time consumed for a session and the scheduling of need assistants.

Nor were the resulting photographic images of suffi­ cient quality, because 32 fps was not fast enought to stop the movements of the lower leg. Even Cracraft who filmed the pigeon at 250 fps reported that he could not adequately discern the rapid movement of the phalanges. Synchronizing a stroboscope to the camera may have eliminated the blur­ ring problem, but the aforementioned delay was a problem 16 large enough that I decided against investing in this technique.

The cost of film and developing were also prohibiting factors. Ducks were found to be less than ideal subjects.

Many attempts were made to acquire just one useful sequence of movements due to the stubborn nature of the animals.

Great quantities of film would have had to be exposed be­ fore the desired data could have been collected.

Another handicap inherent to conventional 16mm cinema­ tography is lighting. To achieve maximum clarity and depth of field, it is desireable to use small aperture, but the bright (hot) lamps required for this posed two problems.

First, the lights disturbed the subjects. Second, the animals quickly overheated in the hot light during walking experiments, showing clear signs of distress from the heat in a very short time.

Also found to be unsatisfactory were 35mm still photo­ graphs. Though the images obtained by using a short dura­ tion xenon flash were sharp, they gave only isolated frag­ ments of data from the series of movements responsible for the act of locomoting.

Video recording offered a solution to some of the above problems. Video tapes can be played back instantly elimi­ nating long and costly developing times. The tapes can also be erased and reused; only useful data need be saved.

With conventional motion picture methods you pay for your mistakes. Television cameras also allow continuous monitor­ ing during recording. This provides immediate feedback as to what is going onto the tapes. In addition, the sound track can be used for verbal notes and comments. Electro- physiological data may also be recorded in synchrony with the visual image providing an opportunity to use electro- mygraphy, electrocardiography or any other from of electro- physiological. Attempts were made to use these techniques, but they will not be considered. The main drawback to con­ ventional video recording is the lack of resolution of the images.

In black and white television cameras, the visual image is retained on a phosphorescent screen for a rela­ tively long period even after the electron beam has swept across the screen to change the light image into electrical impulses. The retention time inherent to the phosphors causes blurring of the image when the tape is played back at slow speeds. Fortunately a new technique was developed to overcome this problem.

Ordinarily, a television camera is used with contin­ uous lighting. The image that is focused on the vidicon tube's phoshorescent surface, unless stationary, moves along the screen's surface. Because of the long retention time of the vidicon tube's coating, when the electron beam

sweeps back for the next image, it not only sees the image of the subject in its new position, it also transmits the 18 previous "ghosts" image. The electron beam scans the image tube thirty times per second. As in the case of cine films, the illusion of motion is a product of our tendency to re­ tain visual images and blend them into what we interpret as one smooth continuous movement if the images are presented about the human fusion frequency of 10 fps.

Photographers have known for years that image blurring can be eliminated by using the light from a bright discon­ tinuous light source such as that from a spark or xenon flash tube. This technique I discovered also works when applied to video records, and greatly enchances its capabilities.

The light from a xenon stroboscope served this purpose.

The stroboscope was set for a flash rate of 1800 flashes per minute (fpm). In this way only one brief flash of light was emitted every 33.33 milliseconds (ms) (thirty times per second) . The television vidicon tube "saxv" noth­ ing in between flashes. Therefore after an image was formed on the screen, the electron beam was then scanning it during a period of darkness. Only the position of the subject during the very brief period for which the light was on was recorded on the screen. By the time the light flashed again, the original image had some time to decay, and the subject moved to a new position. Ghosts are still visible, yet now they consist of distinct images at regular inter­ vals. Instead of detracting from the data they enhance it. 19

They allow one to see the limb in several positions at

regular time intervals on the same frame.

The camera and light source were not synchronized as

the technical complexity of this procedure was prohibitive.

Fortunately the brief duration of the flash (less than

.0005 seconds) gives one a relatively long period of time

per frame in which to provide the flash for the next sweep

of the vidicon tube's electron beam. Occasionally the

timing of the two events was such that the resulting images

were not as bright as one would wish. This would occur if

the electron beam were to sweep the screen long after the

stroboscope had already flashed.

Observation methods; walking

I discovered early in my experiments that ducks could

not be expected to behave in a predictable manner suitable

for my purposes. The animals never become tame despite

frequent handling. Some remained relatively vicious, and I

learned that the , claws, and bill of an enraged or

frightened duck are formidable weapons. A fleeing duck is

very elusive and can cause considerable destruction.

Walking movements were monitored using a custom build

'walkway consisting of a 121.92 cm by 182.88 cm fibreboard

back wall, a 30.48 cm wide pine walking surface, and a

121.92 cm by 182.88 cm clear plexiglass observation window

which served as the front wall (see Figure 1). Recordings

could be made from the front, back and side. The animals 20

were controlled by an assistant from behind the opaque back

wall. An unwilling subject could be reached from above

since the two walls (front and back) were braced by simple

notched braces which were small, movable and removable.

The dimensions allowed the duck freedom of movement only

within the bounds prescribed by the dimensions of the

walkway.

The walkway's length was determined by several factors.

Spacial restrictions dictated that the walkway be mobile.

The walkway had to be long enough so that the subject could

be taped after it had "warmed up" with a few steps. I

wanted to record animals moving at a constant velocity.

It was decided that the camera remain fixed while the

subject passed through its field of view. Moving the cam­

era to follow the subject would have introduced other vari­

ables. For instance, to keep the subject at a constant

distance from the camera, a curved pathway would have been

needed, and the camera would have been placed at the focus

of the curve. This would, however, require the duck to

constantly accelerate due to the continuous directional

changes it would have to make to remain on the walkway. To

'minimize this effect, a curve with a very large radius,

with a very gradual curve, would have been required. Space

was limited so this technique was out of the question.

By using a camera in a fixed position with a telephoto

lens, I was able to record extreme closeups. Parallax was 21

kept to a minimum because the field of view was very small.

Although only a few steps could be recorded out of any

series of movements, it was deemed that a sufficiently

large number of representative steps used in the final

analysis would negate this drawback.

Observation methods; Swimming

All swimming was; done in a custom built plate glass

tank 203.2 cm by 40.6 cm by 121.9 cm as shown in figure 2.

Although the tank was structurally sound using RTV rubber

sealing compound alone, an aluminum framework was con­

structed and fitted around the edges of the tank. Addi­

tional sealing compound was placed between the framework

and glass, providing additional support to make the tank

portable when full of water. The stand was of an open

frame design so that observations could be made from beneath

the tank.

The tank was built to satisfy the same basic require­

ments as the walkway. That is, the animal had to be free

to move within a defined space that would permit the animal

to make natural and unencumbered movements while restrict­

ing its path. In addition, the tank had to be deep enough

' so that the duck could not reach the bottom of the tank

with its foot, or it would walk rather than swim.

The water was cleaned frequently during filming ses­

sions because the ducks quickly fouled the water, and the

slightest amount of turbidity had marked effects on picture

clarity. 22

The water in the tank was kept static in preference to a moving water system such as that used by Prang and Schmidt-

Nielson (1976). My own work with cats, (Messinger, 1974) and that of Wetzel (1975) and Tucker (1968) indicate that movements made by an animal on a treadmill, or by an animal trying to maintain its position in a moving water system must be very different from those of an animal trying to propel itself through water that is not moving.

Contrary to popular belief, a swimming animal or human does not try to move its limbs through the water for pro­ pulsion. The act is better thought of as being an attempt by the swimmer to stabilize a distal and movable part of its body in the water and then draw its body forward using that stabilized part of its body as a fixed point. Any movement of the "paddle" area through the water is called

"slippage". It is essential to keep slippage to a minimum since it results in wasted effort. Slippage may be toler­ ated, however, if it places the limb in a position that gives it some form of postural advantage, perhaps by allow­ ing a larger muscle group to come into play, or increasing the limb's mechanical advantage (see Cousilman, 1968).

A moving current would upset the balance of forces normally established by the swimmer in relation to the water. Indeed, the entire milieu of eddys and currents normally encountered (see Hildebrand, 1974 for a review) would be upset. For these reasons, in addition to the cost 23 and complexity of a moving water system, the simple and more natural static water system was employed.

Analytical methods

For analysis, the video tapes were played back at slow speed, stopped, and the desired representative sections were photographed frame by frame using Kodak Plus X panchro­ matic film. Contact prints were then made from these nega­ tives. The measurements presented were made from these prints and converted to stick figures 4 and 5 (see pages 56 and 58). Attempts to trace the images directly from the television screen were abandoned. The screen of a televi­ sion is curved, introducing measurement errors since tracing what appeared as a straight line required following an arc over the screen. The image on the screen also tended to lack definition and contrast. It was also deemed desireable to show actual photographs since they lent themselves to the use of a new stereoscopic technique discovered during the course of this investigation and described below.

Stereoscopic viewing

Comparisons can be made between successive frames of the contact prints using the principle of stereoscopy. The principle of the stereoscope is a very old one. According

to Asher (1961), Leonardo da Vinci came very close to in­ venting the stereoscope. He ably described the concept of both eyes seeing different views of the same scene. It remained for Sir Charles Wheatstone to invent the 24 stereoscope in 1832, he used a mirror device that was very effective, but not without certain drawbacks. For example, a special room had to be constructed to make them. Photo­ graphy came into being within a short time of Wheatstone's invention, and he had early stereoscopic photographs made.

Later inventors, among them Oliver Wendell Holmes, refined the stereoscope to its present form. Two photographs or drawings of the same scene, each taken from a different location in time or space, are placed at the focal points of two convex lenses. In this way, the rays of light coming from any one point in either picture emerge from the lenses parallel to each other. This gives the effect of placing the subject at a point at an infinite distance from the observer. Less stress is placed upon the viewer's eyes since accommodation for the distance of the subject is no longer a factor.

A steroscope camera is really two cameras in one. The two cameras take their respective pictures of a scene simul­ taneously. Due to the spacial separation of the two lenses, each photograph is taken from a slightly different position.

Aerial photographs on the other hand, employ a single camera with a single lens. Exposures of the landscape below are taken sequentially in time as the airplane flies a straight course over the terrain. Each photograph records the terrain from a slightly different position. 25

While examining the accompanying contact prints (see

Plates I-VII) it occurred to me that these exposures were not unlike those made by an airplane taking sequential ex­ posures of the countryside. Each frame in a series of steps presented on the accompanying sheets is taken from the video tapes used for the original collection of data.

Frames of the television occur at a rate of 30 fps or once every 33.33 ms. If the camera is still, and the subject moves, then the view of the subject is one taken 33.33 ms later and consequently from a slightly (assuming the sub­ ject moves at a relatively slow speed) different angle.

One can then treat any two successive frames as comprising a stereoscopic pair.

The advantages of viewing the data stereoscopically are more a qualitative rather than quantitative asset.

Conventional photography and stick diagrams like those used by previous authors are useful tools only if we acknowledge some of their inherent disadvantages. Locomotion, and the movements of the anatomical structures used by any animal to accomplish it, occurs in three dimensions. While a body or limb moves foreward, it may simultaneously be moving in other directions. We can talk about pitch, roll, or torques, but our understanding of the phenomenon will de­ pend on our ability to integrate all the movements that are occurring in the real three dimensional world. A three dimensional appreciation is highly desireable from an 26 intuitive even more than an objective viewpoint. It takes a determined and highly sophisticated reader to assimilate the separate descriptions of a limb moving anteriorly and posteriorly, the descriptions of its movements medially and laterally, and the simultaneous movements of the limb up and down. Even then the reader's conception may be con­ fused. The stereoscopic method, via the illusion of depth allows one to gain a more natural understanding for what is happening. The illusion on motion is also heightened using this technique.

One can develop the ability to view stereoscopic pictures without the aid of any kind of viewer. For in­ stance, by converging the images of the letters 0 and X, an

X can be made to appear inside the 0, if the two letters are spaced properly on a page. Construct a small card printed with these letters spaced approximately three cm apart, and then produce a fused image with the card near the contact sheets by simply shifting your gaze to the photographs, and the desired illusion will be produced.

During the course of the text, I will constantly be referring to actual photographs of ducks performing the

'movements being discussed. These pictures constitute the data base, for this study and can be found in the appendix

(Plates I-VII).

Since there are 160 individual photographs presented, a system for locating individual frames was developed. 27

Each frame is located and referred to by using three

numbers. The first number indicates which photographic

sheet (one of seven located in the appendix) that contains

the frame. The second number refers to the horizontal

strip, counting from top to bottom, on which the frame is

on. The third number refers to the number of the frame on

its strip, counting from left to right. For example, 543

means the fifth contact sheet, the fourth strip of film

from the top of the page and the third frame from the left

hand margin of the sheet.

An entire series of frames may also be referenced. In

this case, four numbers will be used. The first two refer

to the same information as explained above, however the

last two numbers will be hyphenated. These two numbers

refer to the first and last frame numbers of the series.

Therefore 543-5 is a series of three frames on the fifth

sheet, fourth strip from-the top, frames three through five.

Stick figures

The stick figures presented in the results section of

this paper were constructed as follows. Joint angles v?ere

constructed from the contact prints using basic geometric

<■ construction techniques. Because the images were so small,

tracings were impractical. Instead, an "average" limb was

synthesized. Raikow (1970) presents ratios for the rela­

tive lengths of the limb's skeletal components. Because he

presents data from a large number of samples, I used his 28

ratios for calculating the lengths of an ideal limb. I

measured the tibiotarsus of three of my own ducks and used

the average length of this bone as a base from which to use

Raikow's ratios.

The phalangeal-tarsometatarsal joint (foot), the

tarsometatarsal-tibiotarsal joint (ankle) and tibiotarsal-

femoral joint (knee) (see figure 3 for a diagram) could be

seen on the photographs since the feathers of the left limb

were shaved off. White "Krylon" spray enamel paint, a non­

toxic product, was sprayed on the subject’s limb to improve

the contrast of the images. These preparations did not

change the duck's behavior to any noticeable degree.

The hip joint could not be seen distinctly, so the

acetabulum was located by extrapolation. It could be seen

that the acetabulum did not oscillate in the vertical plane,

so the joint was located using the method described by

Rylander and Bolen (1974). In brief, their method utilizes

successive frames. The length of the femur is used as the

radii of two circles drawn around the knee in each frame.

The acetabulum lies where the two circles intersect. This

method was modified for the walking sequences since small

'vertical displacements of the body were apparent. The knee

joint was used as a reference point again. However its

distance from the walkway was measured in successive frames

on the photographs, and then using a proportionate distance,

its position was drawn on the stick diagram. From this

calculated knee position, the hip joint was also located. RESULTS

Walking: Knee, ankle and foot angles

Both walking and swimming at a constant velocity on or

in a homogeneous medium can be considered cyclic events.

We can think of walking as consisting of two main phases, a propulsive phase, and a nonpropulsive phase. The propul­

sive phase exists when the animal's limb is on the ground and is supporting the body's weight. This may also be called the stance phase. During the stance phase, the limb

first makes contact with the ground, and the shock of the descending body is absorbed. The limb then supports the body as it passes over the limb, and then supplies muscular energy to propel the animal forward. This cycle repeats itself for each step. The nonpropulsive phase of the step begins as the foot leaves the substrate, and may be called the swing phase. During the swing phase, the limb is brought forward in relation to both the body and the ground

for the next step which will begin at foot contact with the ground. Swimming also contains propulsive and nonpropul­

sive phases. The propulsive phase begins as the limb is brought posteriorly in relation to the body, and the nonpro­ pulsive phase, or recovery phase, begins as the limb is brought forward in relation to the body. 30 In the accompanying stick figures, an arbitrary begin­ ning to the step cycle was chosen. Each step cycle begins just after the limb has completed its propulsive phase of the cycle (the swim cycle also). This point was chosen because at this time both the walking and swimming limbs look very similar.

Figure 6 illustrates the angles measured from the stick figures. Each graph shows the measurements for more than a complete cycle: the graph being expanded by simply repeating points in succession from the cycle. If we first look at the foot (tarsometatarsal-phalangeal joint) angles,

(stick figure 4, figure 6) we see that the foot's distal end is rising away from the ground. This movement appears to be initiated by the ankle joint flexing. The phalanges remain pointing downwards, and the web is closed. As the foot is brought forward, the phalanges are flexed, tucking the foot underneath the body. Photographs (334-344) show that the foot will be brought parallel to the ground as the foot angle approaches 180°. At this time, the knee and ankle are both extending. The knee extends slower than the ankle until about stick 10 in Figure 4 when the knee reaches 115°. The knee then extends faster than the ankle until the knee reaches a peak angle of 204°. Just after this (approximately 33 ms) the ankle extends to a peak angle of 155°. 31

While the knee is reaching its greatest extension

during the swing phase, the ankle is extending and the foot

is being brought closer to being horizontal. The knee

stops extending while the foot and ankle continue to extend.

The foot and ankle increase their rates of extension until

the limb is fully extended (between stick figures 1 and 2).

The limb is never completely straight (see frames 334, 344,

and 345).

Frames 215, 523-4, 613 and 632 all illustrate the

phalanges striking the ground before the "heel" (hallux).

This observation will be commented upon within the discus­

sion. It is also apparent from frames 212-4, 331-2, 336,

523j 531-2 and 543 that the duck's foretoes strike the

ground first, and that the middle, and the longest ,

is the last part of the foot to leave the ground. This

digit appears to provide a final push during the last part

of the stance.

At the contact with ground, the foot becomes horizon­

tal and serves as a fixed point over which the body pivots.

The foot angle remains relatively constant for a time (see

figure 4, stick figures 6 and 7) as does the knee until the

- limb passes through the vertical axis. As the body pivots

over the limb, the foot angle begins to decrease and both

the knee and ankle extend.

Just before lift off, the knee and ankle (figure 4,

stick figures 8 and 9) begin to flex. This begins to raise 32 the foot from the ground. In order to maintain toe contact with the substrate for the push off, (see frames 616, 633,

535, 523, 543, 223) the foot must now extend.

The ankle and knee continue to flex into the beginning of the swing phase. During the swing we see that the foot continues to flex. This tucks it close to the body and presents its dorsal surface forward. The foot flexes until the limb again passes through the vertical plane. Now however, the body serves as the "fixed" reference point for the swinging limb. From its vertical tucked position, the foot is quickly extended past the plane horizontal to the ground, and the ankle, which continues to flex, raises the foot still further from the ground. The knee maintains a relatively fixed position. This indicates that the hip is flexing and drawing the leg forward as a unit. Indeed the photographs 525, 526 ana especially 342-6 show the tarsomet- atarsus being thrust forward along the horizontal plane with very little vertical displacement. The tarsometatarsus and phalanges thrust forward, presenting little anterior surface area.

Frontal view of walking

.To add another dimension to the study of the duck's walk, recordings were taken as the subject approached the camera. From this position, an observer can see that as the foot leaves the ground, the web closes, and the foot swings in an arc toward the midline of the body, to about 14° from the vertical (see 62 1-6 and 63 1-6). These move­ ments occur during flexion, that is the foot is still being lifted away from the substrate. As extension begins, the foot begins to move laterally, until at contact, its axis lies about 19° off the vertical (see 113, 215). As the body pivots over the leg, the leg approaches within 8° of being vertical (just as the other foot makes contact).

Thus one leg moves in an S shaped path as the body rocks medially and laterally over the supporting leg. It is this rocking motion of the body that gives the duck's gait its characteristic waddle.

Swimming: Knee, ankle and foot angles

Figure 5 shows the basic movements of the limb during swimming from a lateral view. The stick figures again represent simplified drawings constructed from the photo­ graphs in the appendix.

Figure 7 illustrates the knee, ankle and foot angles during swimming. The cycle is begun during maximum flexion of the ankle joint. Like the walk cycle, the swim cycle has a propulsive phase and a nonpropulsive phase also called the recovery phase.

As the foot is brought forward during recovery, the ankle extends from a minimum value of 27° to a maximum value of 146° (see frames 725-6). This extension continues until the end of the propulsive part of the stroke. During this time the foot is quickly extending, thus bringing the 34 phalanges closer to being in line with the tarsometatarsus ana forming a foot angle of 180°. Since the ankle is still extending, the now straight lower limb begins to exert a force having both a vertical and horizontal component.

This begins the propulsive part of the stroke.

It is the reaction forces exerted on the limb during the early part of the propulsive phase as described above, that are most probably responsible for the foot angle going beyond 180° as in the stick figure number 6. There is no corresponding position to this one during the walk.

The ankle and foot appear to be the points most in­ volved in the swimming form of locomotion. Knee angles remain relatively constant (figures 7 and 8) during the cycle, the range of motion being only about 25°. This is very slight in contrast with the 125° excursions of the ankle and the 135° excursions of the foot.

As the foot and ankle extend during the propulsive phase, so does the knee. This motion drives the foot down­ ward, thus extending the limb and providing a larger sur­ face area to the surrounding water. By tremendously decreasing the ankle and knee angles, the moment arm of the leg is decreased. The torque that needs to be generated by the musculature is consequently reduced. The limb can then be brought back to its forward position with a minimum expenditure of energy while also producing a minimal amount of drag. DISCUSSION

In the early stages of this investigation I formed the hypothesis that walking and swimming are very similar move­ ments (see Introduction). While making some preliminary observations of a duck swimming, I was struck by the fact that if one did not know which form of locomotion was being observed, one could easily be confused. One brief film sequence,made accidently when something scared one of the animals, dramatically illustrated this fact and led to a series of inquiries. The frightened animal in question swam across the tank in panic. When the video tape was played back at slow speed, the duck appeared to be running through the water rather than swimming.

Unfortunately, an intensive study of high speed locomo­ tion for the duck was not possible. Nonetheless, a relatively high speed sequence is included with the data (see frames

411-6) and will be discussed later on in this paper.

Comparison of the joint angles during walking and swimming

Figure 8 shows that the knee is subject to greater move­ ment in walking. Its range of motion is at least 126 degrees, as compared to only 33° of motion during swimming.

When a duck is swimming at a slow speed, the ankle joint lies just at the level of the water's surface. The ankle

35 36 does not move horizontally,in relation, to the body at slow speeds. This observation is another indication that there is a lack of movement by the knee. The swimming gait was observed to change with an increase in the animal's velocity and this phenomenom will be discussed below.

Figure 9 can be used to compare the actions of the ankle joints. The two curves are remarkably similar. The range of motion for the ankle during walking is 125° and 119° for * swimming. Although the walking ankle extends to 161° , 15° more than the ankle during swimming, the ankle during the walk only flexes to a minimum of 36° , whereas the swimming duck's ankle flexes to a minimum of 27°. Therefore one curve is offset vertically from the other curve by a few degrees.

The most obvious difference between the two curves appears during the propulsive phase. When walking, the duck flexes its ankle during the time it passes its body over the foot, but during the corresponding propulsive phase in swimming, the ankle extends to a maximum value and then flexes. This also occurs while the body is passing over the foot. Of course, this can be explained by the fact that the physical demands of the two forms of locomotion differ. For instance, t when the body passes over the foot in walking, its weight contributes vertical forces that must be counteracted. One way to counteract these forces is to use the skeletal com­ ponents of the body to bear the weight of the body. The limb could be kept straight throughout the stance phase. 37

Dagg (1977) found that the Ostrich exhibits this form

of gait. She provides a graph of its gait, and briefly

comments that there is no mid-stance trough for this large

walking bird. When she talks about the mid-stance trough,

she is referring to a phenomenon she observed during the

early stance. The ankle angle of most birds decreases dur­

ing this time. On a graph it reveals itself as a trough.

Such was not the case for the ostrich.

It would seem that this style of gait would cause unde-

sireable vertical displacement of the body. Undesireable

that is, if we assume that these nonpropulsive motions con­

sume energy that is not contributing to the animals forward

progress.

To minimize vertical displacements of the body, joints

such as the ankle can be flexed at the appropriate moments

in time. Gravity can be used to flex the joints in prefer­

ence to energy consuming muscular contractions. Neither

the duck, gull nor the pigeon ever completely straighten

their limbs during the stance.

Another hypothesis concerning the origin of this mid-

stance trough concerns basic muscle physiology. Flexing j the joint during the stance might serve to stretch the ex­

tensor muscles that will soon be called upon to provide

tension. Prestretching these muscles allows them to con­

tract with greater force (see Close (1972) for a review). 38

During swimming, the ankle can be seen to extend and

flex in a smoother manner than the corresponding movement

in walking as shown by Figure 9. This is possible because

the body's bouyancy effectively negates the effect of the

animal's weight. The animal's limb is thus relieved of its

weight bearing responsibility, so the body's center of gra­

vity moves smoothly along a horizontal path. Therefore no

work in the physical sense is done in accelerating the

body's mass. Relieved of these stresses, the limb need not

give, and the mid-stance trough disappears.

The most marked similarity between aquatic and terres­

trial locomotion occurs when one compares the curves obtain­

ed for the foot angle values (see figure 10). The curve

begins just prior to foot placement for walking, and just

prior to the propulsive phase for swimming. Both curves

show that at this time the phalanges are extended. As

weight is being placed on the foot, during walking, and

water resistance is being met during swimming, both inter-

tarsal joints show a tendency to flex. The walking foot

remains planted as the tarsometatarsus pivots over it, and

the swimming foot "gives" slightly due to the stresses

'exerted upon the water by the web.

By hyperextending in this fashion, the web stays per­

pendicular to the path of motion of the body for a longer

period of time. This tends to optimize the ability of the

leg to provide a backwards thrust against the water 39

(observe frames 736-8 for a better appreciation of the actual movements being referred to). The shapes of the two foot angle curves are remarkably similar although we can see that the respective foot positions are actually differ­ ent. The foot remains parallel to the line of motion dur­ ing the propulsive part of the walking cycle, and perpendi­ cular to the line of motion during the swimming cycle.

Comparison of the swing phases

When comparing the walking and swimming cycles of the duck, it is the two swing phases that show the most simi­ larity. Compare the swimming series 721-6 and 731-6 with the walking series 521-6. Notice the exaggerated lifting of the foot during walking. Especially apparent is the closing of the web and the flexion of the digits during protraction. These walking movements are very similar to those that occur during the recovery (swing) phase of the swim cycle.

Why does the web open and close during ? Why do the phalanges flex as much as they do?

If we view the duck walking from the front and the rear

(see 111-6), we gain a good vantage for observing the action of the webbed foot, and the motion of the limb side­ ways. One can see that the leg does not swing medially to a large enough extent to make interference between the swinging leg and the supporting leg a factor necessitating the closing of the web. One might argue that by closing 40 the web, air resistance is decreased for the swinging limb.

This is true, yet the amount of air resistance at these low velocities is probably negligible. It is my thesis that the similarities observed, such as the closing of the web, the extreme flexion of the digits, and the raising of the foot, can be attributed to the fact that the duck has adopted basic locomotor patterns that can be used in the water and on land. These patterns are mainly the result of selection for movements that either reduce drag, or are involved in supporting the duck's weight. Retaining move­ ments for walking from swimming that tend to reduce drag is developmentally economical since common neural pathways can be used for both patterns of movement. Drag reducing movements, such as closing of the web, and a high step offer no disadvantages to the walking animal, while water resistance is an important factor to the swimming animal.

Water resistance is responsible for the observation that the recovering foot during swimming forms a more acute angle with the tarsometatarsus than the swinging foot during walking. This is due to the frontal drag the foot encounters. This is probably a passive movement since the

forces are supplied by the resistance of the water rather than by intrinsic muscles.

Comparing the gaits of aquatic and terrestrial birds of different species reveals findings that are predictable

in light of the thesis stated above. That is, walking 41

movements of swimming birds have been modified to such an

extent that they now show similarities in walking gaits not

inherent to more cursorial species.

For example, in Figure 6 we see that the mallard, an

awkward walker, shows the most flexion of the ankle during

the swing phase. Dagg supplies the four reference angles

for eight other birds in her Table II (pg. 533). The non-

aquatic pea fowl has a maximum flexion of only 75°, while

the magpie goose, a presumably more aquatic species, shows

the greatest flexion during the swing phase of any of the

other birds on the table. These findings suggest that

adaptation to swimming does influence the gait of these

animals. It can be seen that the swimming birds reduce

drag during the recovery by flexing the ankle joint. This

brings the limb closer to the body thus reducing the length

of the lever arm, and the frontal surface area presented to

the water. Since drag is a key factor in aquatic locomo­

tion, swimming ducks (and other birds) have modified their

movements to mitigate its effects.

At this point it seems appropriate to comment on some observations made and recorded concerning the change in

gaits that occurs as the velocity of the swimming duck in­

creases. These data are being discussed at this time be­

cause the conclusions drawn, and the principles involved

concern both swimming and walking. 42

Frames 753-6 show a duck moving at a relatively fast rate of speed. The added extension of the limb is obvious, and is also predictable from Prang's (1970) work.

While studying the metabolic cost of swimming in ducks,

Prang noticed that:

The duck, however, was observed to paddle at a constant stroke rate at all swimming speeds. Since the web of the duck's foot appeared to be fully opened at all speeds, increases in swimming speed were probably accomplished by an increase in the length of the path followed by the foot.

Hildebrand (1974) states that the speed of an animal is determined by its stride length and its stride rate for terrestrial animals. Arshavsky (1975) and Goslow (1973) showed that for the dog and the cat respectively, (see

Grillner (1972) for a review) the swing phase remains al­ most constant at all speeds.

Therefore the duck would be expected to increase the extension of the leg at higher speeds. Since the foot and ankle are fully extended even at slow speeds, the added extension of the little used knee joint and hip joints is expected in both walking and swimming.

In summary, we can say that walking and swimming are basically very similar movements for the mallard duck. The differences that we can observe are specific adaptations of a basic pattern of movements to specific locomotor require­ ments. These adaptations, such as increasing the role of the knee joint, decreasing the amount of extension of the 43 foot during walking, and increasing the amount of flexion of the foot during the recovery phase during swimming, are adjustments necessary to counteract the effects of weight and water resistance.

Comparison of the intertarsal (ankle) angles of the walking duck, pigeon and gulT

Locomotor data for the walking Silver Gull (Dagg 1977) and the pigeon (Cracraft 1971) are available from the literature.

In her study of the Silver Gull, Dagg presents the ankle angles of the gull at five intervals during one step.

She plots these angles against the corresponding motion picture frames on which they appear. I have regraphed these data after converting the frame references into time in milliseconds by using the formula Time (in ms) equals the number of frames times 1000 ms divided by 24 fps. This shows that the cycle took 296.29 ms. Figure 11 includes the plotted data for the gull as well as Cracraft*s moder­ ate gait data for the pigeon. Again, the frame references have been converted into real time in milliseconds.

Cracraft's pigeon, moving at a moderate gait, took 384 ms to complete its cycle. My data for the mallard ankle are also plotted on Figure 11. The duck's walk cycle is about

366 ms in duration. All the cycles were shifted along the abscissa so as to be roughly in phase with one another. In this way the leg positions of the birds closely coincide 44 with each other.

All of the ankle angle data available were plotted on the same graph. Unfortunately Dagg used only five widely spaced data points from the entire cycle. The leg posi­ tions corresponding to Dagg’s graph’s data points in her own Figure #2 are not very clear. It was assumed that

Dagg's minimum point corresponds to the ankle position dur­ ing the swing phase as shown by the left leg of gull number five in her Figure #1. This position is easily distin- quished in the duck and pigeon data and is used as a common point of reference for all three curves.

Similarities are obvious. Even with the differences in stride time, the slopes and variations of. all three curves are remarkably alike. Each shows two maximum and two minimum values. Dagg discusses two maximum and two minimum points for the ankle angles of the walking gull.

None of the three birds exhibited a straight limb

(ankle angle of 180°) during their respective step cycles

(upward deflections of the curve represent extensions of the ankle and downward deflections represent flexion). All three birds show maximum ankle flexion during the swing phase. This helps raise the foot off the ground. All three birds show another minimum point during the early stance phase. These results may have been predicted on the basis of Grillner's work in 1972. He states that during the early stance, (he calls it the yield) maximum stress is 45 placed upon the limb. At this time, the extensor muscles giving support, may even be stretching (while actively supplying tension) because the joints are in fact closing under the large forces encountered. Fortunately these stresses were later measured by Clarke and Alexander (1975) in their study of the forces encountered by running quail.

They found that when they graphed the force exerted the ground against time, a graph with two force maximums was obtained. The first and largest maximum occurs as the foot touches the ground for the stance (which verifies Grillner's yield hypotheses) phase. This is logical since in addition to the body’s weight, the kinetic energy of the body would be transferred to the ground through the suddenly deceler­ ated limb. Later on during the stance phase, as the limb extends to accelerate the body forward, another force in­ crease occurs due to the body's weight and the forces gen­ erated by the bird's leg muscles. There seems little reason to doubt that Clarke's force graph should not fit most any alternating-gait biped. It lends a great deal of validity to the observed similarities in all four birds.

Differences obviously do occur between the gaits of

"these species. For instance, the duck's ankle extends as far as 161° and flexes to an angle as close as 36°, a range of 125°. The gull extends its ankle as far as 185° and only flexes it to about 72° during a gait of moderate speed

(a range of 103°). Cracraft's pigeon, at a moderate gait 46 exhibited a maximum ankle extension of only 146° and a flexion of 65° giving it an excursion of only 101° through one stride.

Frames 215, 523-4, 613 and 632 all illustrate the phalanges of the duck striking the ground before the heel.

This does not appear to be a method common to all the birds examined in this study. Cracraft (1971) tells us that for the pigeon, the hallux and foretoes touch the ground "at very nearly the same time" and that the toes are extended as the leg prepares to touch the ground again. This is unlike Dagg’s gull in which the heel strikes the ground before the toes (see her Figure #1, sketch 3). The quail also makes contact with the heel before the toes according to Clarke and Alexander (1975). Why there should be such variation during this part of the step between species is uncertain. We may speculate that it has something to do with the type of terrain each species encounters most often, or perhaps with an important function of the foot during some other activity that might influence the step much as swimming influences the walk cycle. CONCLUSIONS

1. Walking movements and swimming movements of the duck are

very similar, especially at high speeds.

2. The mallard' s knee and ankle move synchronously during

walking and swimming.

3. The knee joint plays a large role in walking, and a rela­

tively minor role in swimming, at slow speeds. During

swimming, bouyancy frees the knee joint from its weight

supporting role.

4. The knee joint is involved to an increasing extent as

swimming speeds increase.

5. Birds that use muscular tension to maintain limb stability

during the stance phase (for example, ducks, pigeons and

) do not completely straighten their limbs during

walking. Analysis of their ankle angles will show a

slight angular decrease during the stance.

6 . Birds that use skeletal elements to maintain limb stabil­

ity during the stance phase (for example the ostrich)

will completely straighten their limbs during walking.

No decrease will be seen in the ankle angle during the

early stance.

47 The movements of the ankle joints of the duck, pigeon, and seagull show similarities during walking that indi­ cate the existance of neuromuscular patterns common to all walking birds.

Swimming birds use the basic movement patterns of walking birds. Swimming birds show differences in their gaits resulting from modifications pertaining to the effects of drag and gravity.

Birds that swim will show modifications in their walking gaits. These modifications, such as an exaggerated lift­ ing and tucking of the foot, show the influence that swimming adaptations have made on the basic bird gait pattern. FURTHER INVESTIGATIONS

The findings of this investigation suggest that more work needs to be done. I would like to see subjects studied with cine radiography and flouroscopy. Using X-rays would elimi­ nate most of the errors that arise when trying to ascertain the exact location of the skeletal elements in the limbs.

There also exist at this time television cameras that are interfaced with computers. The computers are programmed to use the electronic signals produced by the television camera to analyze quantitative measurements made while the subject locomotes. Velocities, accelerations, forces, angles, and torques can all be measured, and graphed automatically by having the subject, appropriately prepared, pass in front of the cameras.

Electromyographic studies would prove of interest in determining the activity sequences used by the muscles involved in locomotion. These findings would also shed light on the evolutionary development of locomotor patterns.

It would be interesting to perform rigorous tests to determine if and how a moving water system would change the swimming gait of a swimming bird as compared to a static water system. Perhaps a non-swimming bird could also be tested in such tanks. The observations might prove informative. 49 BIBLIOGRAPHY

Arshavshii, Y.I., Kots, Y.M. , Orlovskii, G.N., Rodionov, I.M., and Shik, M.L. 1965 Investigation of the bio- mechanics of running by the dog. Biofizika, 10:737-746..

Asher, H. 1961 Experiments in Seeing, Basic Books Inc., N.Y.

Berger, A.J. 1952 The comparative functional morphology of the Pelvic appendage in three genera of Cuculidae. Amer. Midland Nat., 47:513-605.

Burt, W.H. 1930 Adaptive modification in the woodpeckers. Univ. California Publ. Zool., 32:455-524.

Clarke, J., and Alexander, R.M., 1975 Mechanics of Running in quail. (Coturnix), J. Zool. London, 176:87-113.

Close, R.I. 1972 Dynamic properties of skeletal muscles. Physiol. Rev. 52:129-197.

Councilman, J.E. 1968 Science of Swimming Prentice Hall, N.J. 457 pgs.

Cracraft, J. 1971 The functional morphology of the hind limb of the domestic pigeon Columba livia. Bull. Amer. M u s . Natur. Hist. 144:173-268"!

Dagg, A.I. 1977 Walk of the Silver Gull (Larus-novihollandiae) and of other birds. J. of Zool. 182:529.

DuBrul, E.L. 1962 The general phenomenon of bipedalism. Am. Zool. 2:205-208.

Fisher, H.I. 1946 Adaptations and of the locomotor apparatus of new world vultures. The Amer. Midland Nat., 35:545-727.

George, H.C., and A.J. Berger 1966 Avian Myology New York, Academic Press.

Goslow, G.E., Jr., Robert M. Reinking and D.G. Stuart 1973 The cat step cycle: hind limb joint angles and muscle lengths during unrestrained locomotion. J. Morph. 141:#1 1-42.

50 51

Grillner, S., 1972 The role of muscle stiffness in meeting the changing postural and locomotor requirements for force development by the ankle extensors, Acta Physiol. Scand, 86:92-108.

Hartman, F.A. 1961 Locomotor mechanisms of birds. Smithsonian Misc. Collections, 143 #1.

Hildebrand, M. 1974 Analysis of Vertebrate Structure, John Wiley & Sons Inc., U.S. 710 pgs.

Messinger, D.S. 1974 Ankle extensor activity in the walking cat. Am. Zool. 14:Abstract 1267.

Miller, A.H. 1937 Structural modifications in the Hawaiian goose (Nesichen sandvicensis), a study in adaptive evolu­ tion. Univ. of Calif. Pub. Tn Zool. 42:1-80.

Owre, O.T. 1967 Adaptations for locomotion and feeding in the Anhinga and the Double-crested Cormorant. Amer. Ornith. Union, Ornith. Monogr., #6 1-138.

Prang, H.D., and K. Schmidt-Nielsen, 1970 J. Exp. Biol. 53:763-777.

Raikow, R.J. 1969 Evolution of diving adaptations in stifftail ducks. Ph.D. Thesis Univ. Calif., Berkeley. Univ. micro­ films. Ann Arbor, Mich. (Dess. Abstr. 31B:484).

Richardson, F. 1942 Adaptive modifications for tree-trunk foraging in birds. Univ. of Calif. Pub. in Zool., 46:317-368.

Rylander, M.K. and E.C. Bolen 1974 Analysis and comparison of gaits in whistling ducks (Dendrocygna). The Wilson Bull. 86:237-245.

Spring, L.W., 1965 Climbing and pecking adaptations in some North American woodpeckers. Condor, 67:457-488.

Steinbacher, G. 1835 Funktionell-anatomische Untersuchwngen an vogel fussen mit wendeze'nen und ruckzehen. J. fur Ornithologie. 33:214-282.

Stolpe, M. 1932 Physiologisch-anatomische untersuchungen liber die hintere extremitat der vogel. J. fur Ornithologie 80:161-247,

Tucker, V.A. 1968 Respiratory exchange and evaporative water loss in the flying budgerigar. J. Exp. Biol. 48:67-87. 52

Wetzel, M.C., A.E. Atwater, J.V. Wait, and D.G. Stuart. 1975 Neural implications of different profiles between tread­ mill and overground locomotion timings in cats. J. Neurophysiol. 38:#3.

Wilcox, H.H., 1952 The pelvic musculature of the loon, Gavia immer. Amer. Midland Nat., 48:513-573. Walkway for Observing Ducks Walkin Tank for Observing Ducks Swimming Ankle Angle

Foot Angle

Fig. 3 Diagram of Ducks's Leg with Joint References 204°

5 4 . 3 2

Fig. 4 Stick Diagrams of the Walking Duck Cn cr>

r- Figure 4 (Continued)

70°

73°

115°

235

193'

10 9

Ln '-J 108° 108° 133° 125° 125°

146

104° 67° 101° 139° 122°,

160, 158' 148° 180°1 140.

Fig. 5 Stick Diagrams of the Swimming Duck F ig u r e 5 (C ontin ued)

117 112 105 100 100

81°

m k27°

66 °

147° 05°

13 12 11 10 9 8

Ln v£> 200

v> .—I0) on c

00

• A nkle

0 Knee

0 Fool-

I 5 1.0 S t i c k F i g u r e 15 20

Fig. 6 Ornph of Knee, Ankle nnd Foot Angles During the Wnlk Cycle

O 2 5 0 °

200°

o--

10 0 °

5 0 °

20 0 A nkle Stick Figure Fig. 7 Graph of Knee, Ankle, and Foot Angles During Swimming n Knee 0 F o o t

O' 300'

150°

■D-'

100°

70°

1 5 Stick Figure 10 15

0 Walking Knee Fig. 8 Graph Comparing Knee Angles of the Walking and Swimming Duck • Swimming Knee

ro A n g le s 200 ' g. Grp Co aig ke es he Waki n Simn Duck Swimming and g in alk W e th t a s le g n A nkle A paring om C raph G 9 . ig F • Ankle Walk Walk Ankle • Ake Swim Ankle O 5 gure r u ig F k c i t S IU to CT> 0 F o o t Swim Swim t o o F 0 Walk W t o o F • 1 0 i Angl of the Wal ng a wmig Duck Swimming d an g in lk a W e h t f o s le g n A t o o F g rin a p m o C h p a r G 10 - g i f

A n g le s 250 200 150° 10CP ° 5 10 64 A n g le s 200 ' geon n o e ig P □ 1 Duck Duck 1 * Gul 1. Gul * g. ah rng Ankl Angl of the Wal ng Duk Il n Gull u G and n o e I’lg uck, D g in lk a W e h t f o s le g n A le k n A g arin p m o C raph G I I . ig F T S 166. 3 47 633 467 333 .5 6 6 1 UT,S' m ' O Plate I. Rear and Lateral Views of Walking

With Lateral Views of Swimming

66 Plate Plate II. Rear View of Walk with Lateral View of Swim

68 Plate II Plate III. Lateral View of the Walk Plate III Plate IV. Moderately Fast Swim Plate IV

V -y.

HB o-*< m H I* - ) • .« f .« t -y S^'" ‘ ''V HfiSt ' JV *V<

/ - —

4 -bJ 9 >-. g t-> -L«*a gg§i llil Iggiiil HHHjLL vBB ^ ^ T T T ? ?♦ WBBKMmI

> ' W ■ ■ p - „ \i ’- - ^ - HHSKr^ ...... ,.T jgfeSiF Jll pp!piSiSfM ^ ^M5»;.^ s ,^tmm ’- / — '^A*' ■•„ '£T* ? Ji ”? Plate V. Closeups and Multiple

Exposures of the Walking Duck

74 Plate V Plate VI. Frontal and Lateral Views of the Walk Plate VI Plate VII. Fast and Slow Swimming

78 Plate VII