The Mechanics of Terrestrial Locomotion and the Function and Evolutionary History Of

Home , Animal locomotion, Callipepla, Cassowary, Casuariiformes, Common starling, Cracidae

The Mechanics of Terrestrial Locomotion and the Function and Evolutionary History of

Head-bobbing in Birds

A dissertation presented to

the faculty of

the College of Arts and Sciences of Ohio University

In partial fulfillment

of the requirements for the degree

Doctor of Philosophy

Jennifer Ann Hancock

August 2010

The Mechanics of Terrestrial Locomotion and the Function and Evolutionary History of

Head-bobbing in Birds

JENNIFER ANN HANCOCK

has been approved for

the Department of Biological Sciences

and the College of Arts and Sciences by

______

Audrone R. Biknevicius

Associate Professor of Biomedical Sciences

______

Benjamin M. Ogles

Dean, College of Arts and Sciences 3 ABSTRACT

Hancock, Jennifer Ann, Ph.D., August 2010, Biological Sciences

The Mechanics of Terrestrial Locomotion and the Function and Evolutionary History of

Head-bobbing in Birds (194 pp.)

Director of Dissertation: Audrone R. Biknevicius

Head-bobbing is the fore-aft movement of the head exhibited by some birds during terrestrial locomotion. It is primarily considered to be a response to enhance vision. This has led some researchers to hypothesize that head-bobbing should be found in birds that are visual foragers and may be correlated with the morphology of the retina.

In contrast, other researchers suggest that head-bobbing is mechanically linked to the locomotor system and that its visual functions are secondarily adapted.

This dissertation explored the mechanics of terrestrial locomotion and head- bobbing of birds in both the lab and field. In the lab, the kinetics and kinematics of terrestrial locomotion in the Elegant Crested Tinamous (Eudromia elegans) were analyzed using high-speed videography and ground reaction forces. In the field, the kinematics of the terrestrial locomotion of charadriiform birds were analyzed using high- speed videography.

These biodynamic studies found that birds have two distinct locomotor transitions. The first transition occurs as they move from vaulting mechanics to bouncing mechanics, and the second transition occurs when they incorporate an aerial phase during running. Thus, many birds use grounded running during intermediate speeds. Also, this study found in general that head and neck movements are coordinated with limb 4 movements during terrestrial locomotion in charadriiform birds, but were not coordinated in the Elegant-crested Tinamou. Thus, the coordination is neither perfect nor obligatory.

Additionally, the evolutionary histories of retinal morphology, head-bobbing and foraging type in birds were investigated through a comprehensive review of avian literature. The data were compiled and mapped onto a composite avian phylogeny.

Then, the correlation of characters was analyzed using pairwise comparisons.

This study found that a nasal unifoveate retina with a band-shaped area centralis, non-head-bobbing and visual foraging appear to be the ancestral character states for birds. Additionally, there were no significant correlations between head movement, retinal pattern or foraging mode. Although some general trends were observed, most were within clades and, thus, a result of the independent evolutionary history of head movement, retinal pattern or foraging mode.

Approved: ______

Audrone R. Biknevicius

Associate Professor of Biomedical Sciences 5 ACKNOWLEDGMENTS

I would like to thank my advisor, Audrone R. Biknevicius, for all of her guidance, support and revisions. I also thank my dissertation committee members, Donald B.

Miles, Patrick M. O’Connor, Nancy J. Stevens and Nancy E. Tatarek, for the comments and support. I also thank Kay Earls for her assistance in data collection and LabVIEW programming; Josh Hill, Emily Bevis, Ozan Sauer, Kristin Stover and Jessica Freimark who also assisted in data collection and analysis; and the Ohio University Evolutionary

Morphology Group, especially Steve Reilly, Eric McElroy and Andy Lammers (now at

Cleveland State University) for discussions and support. Financial support was provided by National Science Foundation (IBN 0080158 to S. M. Reilly and A.R.B.), Sigma Xi,

Joseph Grinnell Student Research Award and the Ohio University College of Medicine

Direct Grant. 6 TABLE OF CONTENTS

Page

Abstract...... 3

Acknowledgments...... 5

List of Tables ...... 7

List of Figures...... 8

Chapter 1: Introduction...... 11

Chapter 2: Whole-body mechanics and kinematics of terrestrial locomotion in the

Elegant-crested Tinamou (Eudromia elegans) ...... 19

Chapter 3: The effect of head-bobbing on terrestrial locomotion of birds: A case study of the Elegant-crested Tinamou (Eudromia elegans) ...... 41

Chapter 4: The mechanics of locomotion and head-bobbing in charadriiform birds ...... 67

Chapter 5: The evolution of retinal morphology in birds ...... 91

Chapter 6: head-bobbing, retinal morphology and foraging in birds...... 121

Chapter 7: Conclusions and Future Directions ...... 171

References...... 179

7 LIST OF TABLES

Page

Table 2-1: Center of mass mechanics and footfall parameters in Tinamous ...... 36

Table 3-1: Relative durations between locomotor events and head-bobbing events in Tinamous...... 58

Table 4-1: Numbers of individuals and strides, presence or absence of head-bobbing and foraging ecology in charadriiform species ...... 80

Table 4-2: Results for the least squares regressions of log 10 stance and stride duration versus log 10 speed ...... 81

Table 5-1: The retinal patterns in birds...... 107

Table 6-1: The retinal patterns, head movements and foraging types in birds...... 143

8 LIST OF FIGURES

Page Figure 1-1: Diagram of a head-bob...... 16

Figure 1-2: The refraction of light by the fovea ...... 17

Figure 1-3: The different retinal configurations of the fundus oculi of birds...... 18

Figure 2-1: External mechanical energy profiles for representative Tinamous trials...... 37

Figure 2-2: Graphs of phase shift, duty factor and energy recovery in Tinamous ...... 38

Figure 2-3: Video images of a tinamou utilizing grounded running locomotion ...... 39

Figure 2-4: Graphs of peak vertical force, relative stride length and stride frequency in Tinamous ...... 40

Figure 3-1: Sagittal movements of the head relative to stance phases of the feet in a walking pigeon...... 59

Figure 3-2: The hypothesized effect of footfalls and head-bobbing on the pitching of the torso...... 60

Figure 3-3: Graphs of relative hold phase, relative duration from liftoff to the beginning of hold phase and relative duration of touchdown to the beginning of thrust phase in Tinamous...... 61

Figure 3-4: Ground reaction forces and torso pitch in four representative trials of Tinamous...... 62

Figure 3-5: Stance duration of hold versus thrust steps in Tinamous...... 63

Figure 3-6: Relative times of peak vertical force and fore-aft transition in hold versus thrust steps in Tinamous ...... 64

Figure 3-7: Phase shift and mechanical energy recovery for hold versus thrust steps in Tinamous...... 65

Figure 3-8: A diagram of COM movement during pitching of the torso...... 66

Figure 4-1: A phylogeny of Charadriiformes indicating head-bobbing and non-head- bobbing species ...... 82

Figure 4-2: Relative hold phase as a function of speed ...... 83 9 Figure 4-3: The relative duration from liftoff to the beginning of hold phase plotted against speed and box plots of the relative duration from liftoff to the beginning of hold phase for each species ...... 84

Figure 4-4: The relative duration from touchdown to the beginning of thrust phase plotted against speed and box plots of the relative duration from touchdown to the beginning of thrust phase for each species...... 85

Figure 4-5: Stance duration as a function of speed and log stance duration as a function of log speed for all head-bobbers versus all non-head-bobbers .....86

Figure 4-6: Stride duration as a function of speed and log stride duration as a function of log speed for all head-bobbers versus all non-head-bobbers .....87

Figure 4-7: Stance duration as a function of speed and log stance duration as a function of log speed in confamilial head-bobbing species (Black-bellied Plover) and non-head-bobbing species (Semipalmated Plover) ...... 88

Figure 4-8: Stride duration as a function of speed and log stride duration as a function of log speed in a head-bobbing species (Black-bellied Plover) and non-head-bobbing species (Semipalmated Plover) within the same family, Charadriidae ...... 89

Figure 4-9. The sagittal ( x-axis) coordinates for the eye in one trial of the Least Sandpiper plotted against time...... 90

Figure 5-1. Schematics of the retinal patterns in birds ...... 115

Figure 5-2. Diagram of the refraction of light by the fovea...... 116

Figure 5-3. Pie graph of the retinal patterns within birds ...... 117

Figure 5-4. Character reconstruction of retinal patterns in birds...... 118

Figure 5-5. Diagram of the evolution of retinal patterns in birds ...... 120

Figure 6-1. Diagram of the refraction of light by the fovea...... 157

Figure 6-2. Schematics of the retinal patterns in birds ...... 158

Figure 6-3. Pie graphs of the retinal patterns, head movements and foraging types in birds...... 159

Figure 6-4. Character reconstruction of retinal patterns in birds...... 160

Figure 6-5. Character reconstruction of head movements in birds...... 162

Figure 6-6. The proportion of retinal patterns within the total study population and within each head movement category...... 164

Figure 6-7. The proportion of head movement characters within the total study population and within each retinal pattern category...... 164

Figure 6-8. Avian phylogeny with the characters for retinal pattern, head movement, foraging type and foraging method...... 165

Figure 6-9. Character reconstruction of foraging type...... 167

Figure 6-10. The proportion of retinal patterns within the total study population and within each foraging type category...... 169

Figure 6-11. The proportion of foraging type within the total study population and within each retinal pattern category...... 169

Figure 6-12. The proportion of foraging type within the total study population and within each head movement category...... 170

Figure 6-13. The proportion of head movement characters within the total study population and within each foraging type category...... 170

Figure 7-1. Characters for retinal pattern, head movement, foraging type and foraging method...... 176 11 CHAPTER 1: INTRODUCTION

The ability to stabilize a visual image on the retina during terrestrial locomotion is critical in order to reduce motion blur. Consequently, several different mechanisms are used by animals to reduce erratic projections of images across the retina. In most mammals, the eyes can move within the orbit in response to the movement of an image over the retina (optokinetic response, Wallman and Letelier, 1993) or to changes in body posture detected by the vestibular system (vestibulo-ocular response, Wallman and

Letelier, 1993). However, the range of eye movements in birds is limited due to the large size of the avian eye relative to the orbit (Martin, 1985). Birds compensate by having exceptionally mobile heads and necks, and, as a result, the head may move in response to image drift on the retina (optomotor response, Wallman and Letelier, 1993; optocollic response, Gioanni, 1988) or in response to stimuli from the vestibular system

(vestibulocollic response, Wallman and Letelier, 1993). Therefore, birds, as well as other non-mammalian vertebrates, rely on head movements more than eye movements to stabilize vision (Haque and Dickman, 2005).

Head-bobbing, the fore-aft movement of the head during terrestrial locomotion in some birds, is predominately considered to be an optomotor response (Friedman, 1975).

Head-bobbing has two distinct phases: a thrust phase and a hold phase (Fig. 1-1; Dunlap and Mowrer, 1930). During the thrust phase, the velocity of the head is greater than the velocity of the body so that the head is translated to a point in front of the body. During the hold phase, the head is immobile (remains fixed in space) as the body travels forward, creating an illusive backward movement of the head. The thrust phase is thought to use motion parallax, the reality that objects closer to the viewer appear to move faster than 12 objects that are farther away, to generate depth perception (Frost, 1978), and also allows for the differentiation of stationary items on a background (Davies and Green,

1988). The hold phase functions to stabilize a moving image on the retina (Dunlap and

Mowrer, 1930).

Head-bobbing is not ubiquitous among birds. Although species known to head- bob represent only 21 (12.8%) families out of the approximately 187 families in the Class

Aves (Dunlap and Mowrer, 1930; Daanje, 1951; Friedman, 1975; Dagg, 1977; Frost,

1978; Pratt, 1982; Davies and Green, 1988; Wohlschlager et al., 1993; Troje and Frost,

2000; Fujita 2002, 2003; Cronin et al., 2005), sampling across Aves is quite sporadic.

Furthermore, non-head-bobbing species have been reported in 16 (8.6%) families, six of which also contain head-bobbers. For the remaining 153 (81.8%) of avian families, no data on head movements during locomotion have been reported.

The presence of head-bobbing in some birds and not in others has led some researchers to seek correlated features that may explain its distribution across birds. Two visually-oriented hypotheses have been proposed: one based on retinal morphology and the other on avian foraging behavior. The anatomically-based hypothesis suggests that birds with a single fovea in their retina will head-bob in order to keep the image stable on the fovea whereas those lacking a fovea or having two fovea will not head-bob

(Whiteside, 1966). The fovea is a depression in the retina found within the area centralis, the region of the fundus where the number of photoreceptors is increased. The area centralis in birds can be either circular or in the shape of a narrow band (Wood, 1917;

Walls, 1942). In some birds, the rods are eliminated within the area centralis and the region is completely composed of cones. The foveal depression may function to locally 13 magnify the image onto the layer of photoreceptors (Fig. 1-2; Walls, 1942): because the retina has a higher refractive index than the vitreous humor, any light rays that do not pass perpendicular to the retinal surface are refracted radially outward, thereby stimulating a greater number of photoreceptors and, hence, improving visual acuity. The fovea may also allow for fixation of the moving image because of increased sensitivity to motion in this area (Pumphrey, 1948; Martin, 1985). Twelve different retinal configurations have been described for birds, varying in the shape and location of the area centralis and fovea (Fig. 1-3; Wood, 1917; Walls, 1942). The area can either be circular and located in the nasal retina, temporal retina or in both retinal fields, or can be band-shaped, spanning between the nasal and temporal retinae. The foveal depressions range in number from none to two per eye, and may be circular or linear in shape.

An alternative ecologically based hypothesis states that head-bobbing is found in

visually-guided ground foragers because the thrust phase allows better differentiation of

food items on the ground (Davies and Green, 1988). It is clear that different foraging

modes place distinct demands on the visual system. For example, the width of the

binocular visual field, which is the area of overlap of the two monocular visual fields,

appears to vary according to foraging behavior (Martin and Katzir, 1999). Some non-

visual foragers have limited binocular field overlap and rely on tactile, auditory or

olfactory cues to locate and capture food items, whereas many birds that are visual

foragers possess a larger area of binocular field overlap that aids in the identification of

food items. Furthermore, the retinae of some birds with large binocular visual fields,

such as owls, have temporal foveae that are believed to function to increase visual acuity

within the binocular field (Wood, 1917). 14 In contrast to the visually-oriented hypotheses of head-bobbing, some researchers have advocated that head-bobbing may be linked mechanically to limb and body movements during locomotion (Daanje, 1951; Dagg, 1977). Head-bobbing may impact the craniocaudal movement of the center of mass, particularly if thrust phase of the head bob is synchronized with the onset of swing phase (both tending to shift the bird’s center of mass cranially; Dagg, 1977). Synchronization of head and leg movements have indeed been documented at slow speeds (Fujita, 2002, 2003; Fujita and

Kawakami, 2003), although these studies have noted that the thrust phase occurs during the double support phase when both limbs are in contact with the ground.

The visual and mechanical explanations for head-bobbing may not be mutually- exclusive. The synchronization of the head and limbs may ease the stabilization of the head during the hold phase of head-bobbing. Whereas the transition from walking to running typically involves the inclusion of a suspension (aerial) phase between steps in tetrapods, many striding birds move at great speeds while retaining constant contact with the ground (Gatesy and Biewener, 1991; Rubenson et al., 2004). Grounded running in birds, as in humans (“Groucho running”; McMahon et al., 1987), may serve to decrease vertical movements of the center of mass. Therefore, limiting vertical displacement of the center of mass may facilitate the stabilization of a visual image during head-bobbing while grounded running. Interestingly, differences in stride characteristics were revealed in a comparison of head-bobbing and non-bobbing walking of the Black-headed Gull

(Larus ridibundus ; Fujita, 2006): stride length was longer and stride frequency was lower when Black-headed Gulls walked using head-bobbing. The lengthening of the stride may indicate a lengthening of the stance duration and a decrease in the vertical movements of 15 the center of mass, as seen in grounded running, and may improve the bird’s ability to stabilize the head and eyes during the hold phase.

This dissertation explores the terrestrial locomotion and head-bobbing of birds.

The second chapter discusses the mechanics of terrestrial locomotion in the Elegant- crested Tinamou ( Eudromia elegans ) over a range of speeds to analyze specifically the transitions between walking, ground running and aerial running. The third chapter discusses the mechanics of head-bobbing in the Elegant-crested Tinamou, and analyzes the synchronization between the movements of the limbs and the head during terrestrial locomotion. The fourth chapter compares the kinematics of terrestrial locomotion in head-bobbing and non-head-bobbing charadriiform birds, and analyzes the synchronization of limb and head movements during terrestrial locomotion in charadriiform birds. The fifth chapter evaluates the evolutionary history of head-bobbing within Class Aves and compares the evolutionary histories of head movements, foraging behavior and retinal pattern.

Figure 1-1. During the hold phase of head-bobbing the head is stable relative to the environment (signified by the dotted line) as the body moves forward during locomotion. Then, during the thrust phase of head-bobbing the head is accelerated in front of body as the body continues to move forward. 17

Photoreceptor layer Retina

Vitreous

humor

Light rays

Figure 1-2. The foveal depression within the retina creates a curved surface (black line). As light rays pass through the retina they are refracted radially due to the higher refractive index of the retina relative to the vitreous humor. As a result, images are magnified on the photoreceptor layer. 18

Figure 1-3: Schematics of the fundus oculi of birds showing the different retinal configurations known: black oval – pecten; grey shading – area centralis; black dot – circular fovea; black line – linear fovea. 19 CHAPTER 2: WHOLE-BODY MECHANICS AND KINEMATICS OF

TERRESTRIAL LOCOMOTION IN THE ELEGANT-CRESTED TINAMOU

(EUDROMIA ELEGANS )

Summary

Whereas humans and certain birds experience an abrupt change in locomotor

dynamics when shifting from walks to runs, a smooth walk-run transition characterizes

many ground-dwelling birds. This study defines the biomechanical distinction between

walks and runs in the Elegant-crested Tinamou, Eudromia elegans, using ground reaction

forces. Three birds were filmed at 250 Hz from a lateral view as they moved over a force

plate built into a trackway. Center of mass mechanics and kinematic variables were

analyzed in 81 steady-speed trials that represented a speed range from 0.66 to 2.78 ms -1.

E. elegans undergoes two speed-related changes in locomotor mechanics. The first is a

shift from walking strides that utilize vaulting mechanics to low-speed runs that exhibit

bouncing mechanics; this transition occurs at Froude numbers between 0.4 and 0.6. Such

low-speed runs exhibit duty factors exceeding 0.5 and, hence, lack an aerial phase

between steps. The second transition, from grounded running to aerial running, occurs

when duty factors decrease below 0.5. Grounded running in birds may enhance vision by

stabilizing visual stimuli over the retina. The eventual incorporation of an aerial phase

during running enables increased locomotor speeds primarily through longer stride

lengths.

20 Introduction

Birds differ from other vertebrates in many aspects of their biology. An unusual

feature of terrestrial locomotion in ground-dwelling birds is their capacity to move at

medium to high speeds without including an aerial phase (period of suspension when no

limbs are in contact with the ground) between steps. Whereas some species (Ostrich

[Struthio camelus ], Emu [ Dromaius novaehollandiae ], Greater Rhea [ Rhea americana ],

Brown Kiwi [ Apteryx australis ], Wild Turkey [ Meleagris sp. ], Helmeted Guineafowl

[Numida meleagris ], Painted Quail [ Coturnix sp. , formerly Excalfactoria ], Gatesy &

Biewener 1991; Gatesy, 1999; Abourachid & Renous, 2000) eventually incorporate an aerial phase at their highest speeds, others (Bobwhite [ Colinus sp. ], Gatesy & Biewener,

1991, Mallard and Indian Runner ducks [ Anas platyrhynchos ], Abourachid, 2000, 2001,

Southern Cassowary [ Casuarius casuarius ], Abourachid & Renous, 2000, Japanese Quail

[Coturnix japonica ], Reilly, 2000) remain grounded, alternating between periods of single- and double-limb supports, throughout their ranges of speed. The ratio of single limb support duration to stride duration is known as duty factor, a footfall-based parameter that has been used to distinguish walks from runs (Hildebrand, 1976): walks occur at duty factors greater than or equal to 0.5 whereas runs occur at lower duty factors.

In bipeds, an aerial phase occurs when duty factor falls below 0.5. By this criterion, walking would be considered to be the primary mode of terrestrial locomotion in ground- dwelling birds.

Yet walking and running dynamics involve much more than footfall patterns. The complex movements of the limbs, torso, neck and head during terrestrial locomotion can be summarized by the movement of the center of mass (COM). Two basic patterns of 21 COM mechanics have been observed in birds during terrestrial locomotion (Cavagna et

al., 1977; Heglund et al. , 1982; Muir et al., 1996; Rubenson et al., 2004; Griffin & Kram,

2000). At slow speeds, limbs function as stiff struts so that the external mechanical

energies of the bird’s COM fluctuate in a manner that resembles an inverted pendulum

(vaulting mechanics). Gravitational potential energy ( Ep) of the COM cycles out of phase with total kinetic energy ( Ek,tot ) so that the COM is at its highest position during

midstance when kinetic energy is at a minimum. Such dynamics allow for pendulum-like

exchange of external mechanical energy that provides an opportunity to reduce muscular

effort and, hence, improve locomotor efficiency. At faster speeds tetrapod limbs are

more compliant, such that the COM no longer rises during the first half of stance but

rather drops to its lowest position at midstance (spring-mass or bouncing mechanics).

The resulting in-phase fluctuations in Ep and Ek,tot are inconsistent with pendular mechanisms, yet some recovery of mechanical energy is still possible via the storage and release of elastic strain energy in the musculoskeletal system of the limbs.

There is a theoretical limit to how quickly tetrapods can move and still take advantage of pendular mechanics (Alexander, 1976; Usherwood, 2005). As the COM moves in an arc over a stiff limb it resembles a mass attached to the end of a string moving in a circle. For the mass to continue moving in a circle a centripetal force ( mv2/l, where m is mass, v is forward velocity and l is length of the string or, in the case of walking, hip height at midstance) must cause an acceleration of the mass towards the center of the circle. In walking the centripetal force is provided by gravitational force

(mg , where g is acceleration due to gravity). The COM will continue to move about the arc as long as the centripetal force required to maintain the movement does not exceed 22 gravitational force. The ratio of the required centripetal force to gravitational force is

2 known as a Froude number ( Fr = v /gl ); Fr also reflects the ratio of Ek,tot to Ep.

Therefore, animals must switch from pendular mechanics to spring-mass mechanics either at or before Fr = 1. In reality, animals switch mechanics when moving at much

lower Fr because of the prohibitive cost of swinging a limb faster than its natural

frequency (Usherwood, 2005). Among bipeds, humans transition from a walking gait to

a running gait at Fr ~ 0.4-0.6 (Alexander, 1977; Gatesy & Biewener, 1991) and,

similarly, crows [ Corvidae ] transition from a walking gait to a hopping gait at Fr ~ 0.5

(Hayes & Alexander, 1983). As duty factor is inversely related to Fr (Alexander &

Jayes, 1983; Hayes & Alexander, 1983), gaits used above these critical Fr values

typically display aerial phases (duty factor < 0.5).

It is well established that striding birds use pendular mechanics during walking

and bouncing mechanics during aerial running (Cavagna et al ., 1977; Heglund et al.,

1982; Muir et al ., 1996; Griffin & Kram, 2000), but what is the mechanical nature of

their grounded high-speed locomotion? In this study, we examine whether high-speed

grounded locomotion conforms more closely to mechanical expectations for walking

(vaulting mechanics) or running (bouncing mechanics). Using kinetic data Clark and

Alexander (1975) found bouncing mechanics in Japanese Quail across a speed range of

0.35 to 1.4 m s -1 and, because quail rarely incorporate an aerial phase (Gatesy &

Biewener, 1991), these are likely to be grounded runs. Kinematic support for grounded

running also exists (Gatesy & Biewener, 1991). Gatesy (1999) observed changes in limb

kinematics during grounded locomotion in Helmeted Guineafowl suggestive of a shift

from vaulting to bouncing mechanics prior to incorporation of an aerial phase. Similarly, 23 Rubenson and colleagues (2004) used changes in hip height during stance phase to infer that Ostriches transition from vaulting mechanics to bouncing mechanics before incorporating an aerial phase. However, Ostriches are highly derived cursors that do not have the capacity for flight; hence they may differ from the general avian condition.

Indeed, at high Fr values cursorial mammals have lower duty factors than non-cursorial

mammals (Alexander & Jayes, 1983). The goal of this study was to examine

relationships among gait, locomotor mechanics (based on kinetic data) and the presence

or absence of an aerial phase in a more generalized basal bird, the Elegant-crested

Tinamou, Eudromia elegans , family Tinamiformes. These Neotropical ground-dwelling

birds are related to the ratites but are capable of taking flight when startled.

Materials and methods

The kinetics and kinematics of terrestrial locomotion were analyzed in three

Elegant-crested Tinamous, Eudromia elegans (623-865 grams). The birds were filmed in

lateral view at 250 Hz using a NAC camera (HSV-500, Simi Valley, CA, USA) as they

moved over a Kistler force platform (plate type 9281B, Amherst, NY, USA) built into a

4.9 m trackway. The force platform longitudinal length (0.6 m) was adequate to capture

between two and five steps during each trial depending on the speed at which the birds

moved. The force platform recorded vertical, fore-aft (longitudinal) and mediolateral

(transverse) ground reaction forces (GRFs) at 1000 Hz. Video and force data were

synchronized using Motus version 7.2.6 (Peak Performance Technologies, Centennial,

CO, USA).

24 Locomotor Kinematics

Reflective markers were attached to the tip of the middle toe, synsacrum (between the femoral heads, approximating the acetabulum) and breast (between the furcula and keel). The markers were digitized using Motus version 7.2.6 software (Peak

Performance, Centennial, CO, USA) at touchdown, temporal midstance and liftoff of each foot. The forward speed of the birds was calculated videographically using the travel time of the breast marker across two 30-cm intervals marked on the back wall of the trackway and overlying the force platform. Only trials that differed in velocity by less than 10% between the initial and final intervals (steady speed) were analyzed further

[100* │(v1/vt) – ( v2/vt)│, where v1, v2 and vt are the forward velocity of the first half of the trial (30 cm), the second half of the trial (30 cm) and the whole trial (60 cm), respectively]. These trials also displayed balanced braking and propulsive components of their forward velocity profiles estimated from GRFs (see below).

Kinematic variables included duty factor (support duration/stride duration),

2 Froude number ( Fr = vt /gl, where vt is forward speed, g is gravitational acceleration or

9.81 m s -2 and l is hip height measured as the perpendicular distance from the platform

surface to the synsacrum marker at temporal midstance, ~ 0.17 m, Alexander & Jayes,

1983), relative stride length (stride length/ l) and stride frequency (stride duration -1).

Center of mass mechanics

A customized LabView program (National Instruments, Austin, TX, USA) was used to calculate fluctuations in kinetic (Ek) and gravitational potential ( Ep) energies of the COM from the GRFs following Cavagna et al . (1977) and Willey et al. (2004). 25 Vertical (minus body weight), fore-aft and mediolateral GRFs were divided by the body mass to determine acceleration in each direction. Each acceleration was then integrated once to obtain velocities of the COM in the three directions, and vertical velocity was further integrated to obtain vertical displacement of the COM. Average forward speed was used as the integration constant for fore-aft velocity, whereas vertical and mediolateral integration constants were estimated as the mean values for vertical and mediolateral records. Velocities were then used to calculate Ek for each direction as Ek =

½ mv 2, where m is the animal’s mass and v is velocity. Summing vertical, fore-aft and mediolateral Ek yielded total kinetic energy ( Ek,tot ). Gravitational potential energy was

computed as Ep = mgh where h is the vertical displacement of the COM. Lastly, total external mechanical energy ( Em,tot ) was computed as Ek,tot + Ep.

The phase shift between Ep and Ek,tot was calculated by dividing the time

difference between the minimum values of Ep and Ek,tot by the duration of the stride and

then multiplying that value by 360° (Cavagna et al ., 1977). Two external mechanical

energy patterns were identified based on the phase shift between Ep and Ek,tot minima:

vaulting mechanics ( Ep and Ek,tot fluctuate out of phase; phase shifts > 90°) and bouncing mechanics ( Ep and Ek,tot fluctuate in phase; phase shifts < 90°). Trials with bouncing mechanics were further subdivided into grounded runs (duty factors ≥ 0.5) and aerial runs

(< 0.5).

The amount of external mechanical energy recovered via pendular mechanics was calculated following Blickhan and Full (1992):

%R = 100*[( ∆Ep + ∆Ek,tot ) - ∆Em,tot ]/( ∆Ep + ∆Ek,tot ) 26 where ∆Ep, ∆Ek,tot and ∆Em,tot are the sums of the positive increments of the Ep, Ek,tot and Em,tot profiles, respectively.

A repeated measures ANOVA was performed using SYSTAT (version 11) software to evaluated differences in locomotion between the individual birds. The independent variable was bird identity and the dependent variables were speed, Froude number, duty factor, phase shift, energy recovery, relative stride length, stride frequency and peak vertical force. Because some significant differences were found a post hoc analysis was performed to determine which individuals were different.

To illustrate trends in the data on the graphs both linear ( y = ax +b) and nonlinear

(y = aX b) regressions were executed using SYSTAT (version 11) software for phase shift,

percent recovery, duty factor, peak vertical force, relative stride length, and stride

frequency versus Froude number and phase shift versus duty factor. For each variable

pair, the adjusted R-square ( R2) from the linear regression was compared to the mean-

corrected R2 from the nonlinear regression to determine the best fit curve. The majority of trends were nonlinear, except the linear trend of peak vertical force versus Froude number for which the adjusted R2 for the linear regression was greater than the mean- corrected R2 for the nonlinear regression.

Results

Eighty-one steady-speed trials were captured with speeds ranging from 0.66 to

2.78 ms -1; the tinamous refused to move at lower or higher speeds in the laboratory.

Similar locomotor patterns were observed between individuals in relation to speed,

Froude number, duty factor, relative stride length and peak vertical force. However, 27 individual differences were observed in phase shift (p = 0.014), stride frequency (p <

0.001) and energy recovery (p = 0.006). Post hoc analyses revealed that Tinamou A differed from Tinamous B and C, but Tinamous B and C were similar. These differences could be a result of Tinamou A preferentially moving at slow speeds (0.76 – 1.53 ms -1), whereas Tinamous B and C moved at greater ranges of speed (0.64 – 2.78 ms -1 and 0.66 –

1.83 ms -1, respectively). Accordingly, because Tinamou A generally moved slower, its phase shift and energy recovery values were generally higher and its stride frequency values were generally lower than Tinamous B and C.

Figure 2-1 illustrates the vertical, fore-aft and mediolateral GRFs and their associated Ek,tot , Ep and Em,tot profiles for representative trials with vaulting and bouncing

(grounded and aerial) mechanics. Vaulting mechanics were observed only at the slowest

speeds, whereas tinamous were capable of utilizing bouncing mechanics throughout the

speed range observed (Fig. 2-2A). Bouncing mechanics (phase shift < 90°) were used in

the majority of the trials (87.6%). Because phase shifts deviating substantially from

perfect vaulting (180°) or bouncing (0°) mechanics were commonly present, we represent

these trials as displaying intermediate mechanics in Figures 5 and 7 (45-135°, following

Ahn et al., 2004).

Tinamous exhibited duty factors ranging between 0.39 and 0.70 (Fig. 2-2B, Table

1). Aerial phases (duty factor < 0.5) were observed in only 7.4% of trials, all of which

followed phase shift expectations for bouncing mechanics. Only two of the three birds

ran with aerial phases; aerial trials in these birds occurred at distinct Fr values (Tinamou

A displayed aerial trials at Fr ~ 1, whereas Tinamou B exhibited aerial phases Fr > 2, see

Fig. 2-2B). Grounded trials (duty factors ≥ 0.5, Fig. 2-3) constituted the remainder of the 28 sample, including all strides with vaulting mechanics as well as 92% of trials for which bouncing mechanics was observed. As such, duty factor cannot be regarded as a clear predictor of mechanics for this sample: although vaulting seems invariably associated with relatively high duty factors ( ≥ 0.5), only 8% of the trials exhibiting bouncing

mechanics were accounted for by low (< 0.5) duty factors. Furthermore, there was broad

overlap between duty factor values for walking (vaulting) and grounded running trials

(Fig. 2-2C).

Froude numbers ( Fr ) in this study ranged from 0.25 to 4.48 (Table 1). Tinamous

using vaulting mechanics tended move with Fr at or below the 0.4-0.6 mechanical

transition zone determined empirically for terrestrial tetrapods (Alexander, 1977;

Alexander & Jayes, 1983; Gatesy & Biewener, 1991); only one vault-like trial exhibited a

higher Fr (0.74). In comparison, bouncing mechanics (grounded or aerial) virtually

always occurred at or above the 0.4-0.6 transition zone; no aerial runs and only 6% of

bouncing trials fell below 0.4 Fr .

Trials with vaulting mechanics recovered the greatest amount of external

mechanical energy via pendulum-like exchange of Ep and Ek,tot (Fig. 2-2D, Table 1). The maximum energy recovered (55.1%) was found in the slowest walk (0.64 ms -1). Because

the tinamous were unwilling to walk slower, we did not obtain the expected bell-shaped

distribution of energy recovery (low at slow and fast walking speeds and high at

intermediate walking speeds) and therefore we cannot verify whether this value is the

absolute maximum amount of energy that Tinamous are capable of recovering. Much

lower energy recoveries were obtained for trials exhibiting bouncing mechanics

(maximum 29.1% and 4.8% for grounded and aerial runs, respectively). 29 Peak vertical force increased linearly with Fr and similarly across all gaits, whereas relative stride length and stride frequency followed a more curvilinear pattern

(Fig. 2-4). No discrete change in peak vertical force was observed between vaulting trials, grounded runs and aerial runs. Trials with aerial runs exhibited longer relative stride lengths but less frequent strides compared with grounded runs of similar speed and

Froude number. Thus, neither the peak vertical force nor the limb stride kinematics predictably distinguished the vaulting (walking) versus grounded running trials, although stride kinematics do discriminate grounded versus aerial runs.

Discussion

Gait transitions are traditionally identified by changes in limb kinematic parameters such as phase relationships between footfalls or the absence or presence of an aerial phase between steps (Hildebrand, 1965). This study suggests that striding birds switch gaits (as measured by COM mechanics) long before they begin to use aerial running (duty factor < 0.5) and the presence of grounded runs obfuscates these mechanical gait transitions. Some birds, such as the Black-billed Magpie ( Pica pica ),

show changes in kinematic parameters as they switch from a striding walk to a hop

(Verstappen & Aerts, 2000), but many birds move solely by striding and do not show

discrete changes of kinematic parameters over most of their speed range (Gatesy &

Biewener, 1991; Abourachid, 2000; Gatesy, 1999; Abourachid & Renous, 2000; Fujita,

2004). Even among striding birds, however, kinematic signals exist that suggest changes

in COM mechanics with speed (Gatesy & Biewener, 1991; Gatesy, 1999; Fujita, 2004).

For example, discontinuities in the relationship of stride frequency, swing phase duration 30 and maximum knee joint angle with speed occur before the onset of aerial phases in

Helmeted Guineafowl (discontinuities at 0.5 Fr versus aerial phase at 2.3 Fr ; Gatesy,

1999).

Tinamous move with grounded locomotion over all but the highest speeds observed. The slower grounded trials conform to pendular mechanics and are rightly considered to be walks, as has been found in ratites (Greater Rheas, Ostriches) and galliform birds (Wild Turkeys, Chickens [ Gallus gallus ], Cavagna et al., 1977; Muir et al., 1996; Rubenson et al., 2004). Although high speed grounded locomotion has been reported in other striding birds (Gatesy & Biewener, 1991; Reilly, 2000), the mechanical nature of this locomotor behavior has been established only in Ostriches (Rubenson et al.,

2004) and now Tinamous. High speed grounded locomotion is invariably governed by bouncing mechanics, yet these species differ in the degree to which they utilize grounded running. Tinamous prefer to run without an aerial phase over ~50% of their running Fr

range (and one tinamou did not ever incorporate an aerial phase) whereas Ostriches limit

grounded running to the slowest 20% of their running Fr range. Furthermore, Tinamous

begin to employ aerial running at higher Froude numbers (0.97) than Ostriches (0.68).

Similarly, non-cursorial mammals exhibit larger duty factors than cursorial mammals at

similar Fr values as they increase speed (Alexander & Jayes, 1983). Although two of the

tinamous eventually incorporate an aerial phase with increasing speed, their long reliance

on grounded running is reminiscent of the resistance to aerial running observed in other

small ground-dwelling birds (Gatesy & Biewener, 1991).

Across vertebrates grounded locomotion is most commonly associated with

walking. Vaulting mechanics used during walking are believed to reduce muscular effort 31 through a pendulum-like exchange of gravitational potential energy and kinetic energy

(Cavagna et al., 1977). The theoretical limit for when terrestrial animals must switch out of vaulting mechanics occurs at Fr = 1, because at this speed gravitational force becomes inadequate for providing the centripetal force necessary to continue moving the COM along a curved path (Alexander, 1976). Yet changes in COM mechanics reflective of a walk-grounded run transition occurs at a range of 0.4-0.7 Fr in Tinamous and 0.54 Fr in

Ostriches, paralleling gait transitions based on limb kinematics in birds (~0.5 Fr ; Gatesy,

1999) and footfall changes in quadrupedal tetrapods (0.4-0.6 Fr ; Alexander, 1977;

Alexander & Jayes, 1983; Hayes & Alexander, 1983). The Tinamous’ limit for vaulting

mechanics is consistent with predictions of a collisional model of energy costs during

bipedal terrestrial locomotion (0.7 Fr ; Ruina et al., 2005).

Why don’t animals continue to move with vaulting mechanics at higher speeds?

As walking speed increases, animals reach a maximum step length at which vaulting mechanics can occur (Usherwood, 2005). In order to increase speeds further, animals can either increase step frequency or choose to switch from vaulting mechanics to bouncing mechanics. It is now appreciated that the swing phase of locomotion is energetically costly: a substantial fraction (about 26%) of a Helmeted Guineafowl’s total energy is consumed during this phase of locomotion (Marsh et al ., 2004), so energetic factors may act to limit step frequencies. Thus, animals are eventually forced to abandon pendular mechanics when step frequencies become unattainable (Usherwood, 2005). Gait transitions at lower Fr values may reflect a strategy to reduce metabolic costs of locomotion (Hoyt & Taylor, 1981; Kram et al ., 1997; Griffin et al ., 2004). Indeed, the capacity of birds to recover external mechanical energy through pendular mechanics 32 peaks at intermediate walking speeds and then falls precipitously at higher speeds

(Cavagna et al., 1977; Rubenson et al., 2004); unfortunately, tinamous in this study did

not move at slow enough speeds to display this pattern. Thus, birds may switch from

vaulting mechanics to bouncing mechanics at speeds at which pendulum mechanics no

longer provides an energetic benefit. For example, Ostriches display a drop in net

metabolic work when transitioning from vaulting to bouncing mechanics, whereas at

higher speeds Ostriches incorporate an aerial phase with no decrease in net metabolic

work (Rubenson et al., 2004). By switching to bouncing mechanics, some of the

gravitational potential energy and kinetic energy absorbed by the limbs during the first

half of the stance phase is recovered as elastic strain energy in the second half of stance

as a result of the stretch and recoil of elastic elements within the limb (e.g., tendons,

Cavagna et al., 1977). Passive elastic mechanisms thereby reduce the work that muscles must accomplish during running. Muscle-tendon springs within the gastrocnemius muscle of Wild Turkeys have been identified as important energy-saving devices

(Roberts et al., 1997), although the effectiveness of passive elastic mechanisms during

running in smaller vertebrates, such as Tinamous, is debated (Biewener & Blickhan,

1988; Bullimore & Burn 2005).

Another factor that may affect gait choice is locomotor force magnitude. Because

increased speed is related to higher locomotor forces (GRFs), animals may switch gaits in

order to limit overloading their bones or overexerting their muscles (Farley & Taylor,

1991; Fewster & Smith, 1996). Findings of this study suggest that force magnitudes may

not represent a potent stimulus for the walk-run transition in striding birds, as no discrete

difference was found in peak vertical GRFs between walks and grounded runs (Fig. 2- 33 4A). Indeed, the compliant limbs with which striding birds move may serve to dampen peak force magnitudes across the walk-run transition in a manner similar to bent-knee

(Groucho) running in humans (McMahon et al ., 1987).Although Tinamous can run at

intermediate speeds either with grounded runs of relatively short strides and high stride

frequency or aerial runs of longer strides and lower frequency (Figs. 7B, C), they rely

exclusively on aerial runs at speeds greater than 2.2 Fr . Smaller ground-dwelling birds

increase speed primarily by increasing stride length (Gatesy & Biewener, 1991,

Abourachid & Renous, 2000). Although their crouched posture provides great effective

limb length, birds eventually reach the extension limit of their hindlimbs (with or without

femoral involvement) and must “go aerial” in order to continue to increase speed.

Therefore, aerial runs have a greater potential for speed than do grounded runs.

Furthermore, both external and internal kinetic energy fluctuations may be reduced with

the inclusion of an aerial phase at high speeds which would moderate muscular effort in

running (Clark & Alexander, 1975). What possible advantage does grounded running

provide over aerial running in birds? At slower running speeds, double limb support

phases may help counteract the tendency of the body to pitch when the line of action of

the GRF does not pass through the COM at the end of stance phase (Clark & Alexander,

1975). Also, grounded running may provide a visual benefit to striding birds (Gatesy &

Biewener, 1991). The transmission of impact spikes from the foot to the head is greatly

reduced when humans run with bent knees compared with aerial running (McMahon et

al ., 1987). If a compliant hindlimb in birds similarly serves to dampen jarring impact

forces then the visual signal will be less perturbed with each step. This may be especially

important to birds (like Tinamous) that bob their heads during terrestrial locomotion. 34 During the hold phase of head-bobbing, the head is held steady relative to the environment (Dunlap & Mowrer, 1930), a mechanism believed to function in stabilizing visual stimuli on the retina (Friedman, 1975). Head-bobbing and locomotion are well synchronized at slow speeds, with a bob occurring at every step (Davies & Green, 1988;

Fujita, 2002; Fujita, 2003) but are poorly synchronized at intermediate speeds (J. A.

Hancock unpublished), speeds at which Tinamous move in grounded runs. Absence of an aerial phase, in combination with the flexed limb system, during grounded walks and runs in birds may help stabilize vertical movements of the COM during the hold phase of head-bobbing. Therefore, locomotor mechanics, in conjunction with head-and-neck mechanics, may enable a steadier retina and visual signal in birds.

Although the avian visual apparatus may benefit from grounded running, it is not

necessarily true that grounded running evolved in conjunction with a need to stabilize

visual images on the retina. Whereas Groucho running is a contrived gait in humans,

grounded running is a natural part of the locomotor repertoire of many tetrapods. It has

been observed in lizards (McElroy et al ., 2004), frogs (Ahn et al ., 2004), opossums

(Parchman et al ., 2003) and rats (A. R. Biknevicius, unpublished) and has been inferred

in primates (Schmitt, 1999). The sprawled and crouched postures of these quadrupeds

may provide a degree of limb compliancy permitting habitual grounded running as a

solution to a variety of mechanical challenges unrelated to vision. The use of grounded

running in such a broad range of terrestrial tetrapods suggests that striding birds may

simply be exploiting a relatively common capacity for grounded running as a means of

enhancing visual stimuli. 35 In conclusion, Elegant-crested Tinamou undergo two speed-related gait changes. The first is a shift from vaulting mechanics (walking) to bouncing mechanics

(grounded running) occuring at a Fr ~ 0.5. This shift from vaulting mechanics to bouncing mechanics represents a change in energy-saving strategies from a pendulum- like exchange of external mechanical energies during walking to a greater reliance on elastic energy storage and recovery during grounded running. Because the birds continue to move with duty factors > 0.5 this shift is not readily obvious without performing

COM-based studies. The second shift, from grounded running to aerial running, is easily observed as it occurs when duty factors fall below 0.5 and an aerial phase is incorporated between each step; this represents a strategy to further increase locomotor speed through longer stride lengths while holding down stride frequencies. In addition, grounded running in birds may enhance vision by permitting smoother strides that allow for the vertical stabilization of visual stimuli upon the retina.

Table 2-1. Center of Mass Mechanics and Footfall Parameters in Elegant-Crested Tinamous. Froude Mechanics Phase shift N Duty Factor % Recovery Number 0.64 ± 0.05 0.38 ± 0.15 36.7 ± 12.4 Vaulting > 90° 10 (0.55-0.70) (0.25-0.74) (14.6-55.1) Bouncing 0.57 ± 0.05 1.08 ± 0.50 6.41 ± 6.53 < 90° 65 (grounded) (0.50-0.70) (0.26-2.20) (0.32-29.1) Bouncing 0.45 ± 0.04 2.33 ± 1.41 2.54 ± 1.87 < 45° 6 (aerial) (0.39-0.49) (0.97-4.48) (0.57-4.83) Values are means ± S.E.M. (min-max range) 37

Figure 2-1. External mechanical energy profiles for representative trials exhibiting vaulting mechanics (forward velocity, vt = 0.64; duty factor, d = 0.69; Froude number, Fr = 0.25), grounded bouncing mechanics ( vt = 1.56; d = 0.56; Fr = 1.57) and aerial bouncing mechanics ( vt = 2.78; d = 0.39; Fr = 4.48). Illustrated are (top to bottom) ground reaction forces, gravitational potential energy, total kinetic energy and total external mechanical energy during a single step. The vertical dotted lines in the ground reaction force profiles indicate the touchdown of the contralateral foot; the vertical dashed lines in the energy profiles indicate the minima of the gravitational potential and total kinetic energy profiles.

Figure 2-2. (A) Phase shift versus Froude number (nonlinear regression line: a = 25.2, b = -1.11; mean corrected R 2 = 0.577). With increasing Fr , Tinamous generally shift from moving with vaulting mechanics (phase shift > 135 °) through intermediate mechanics (vaulting-like mechanics 135-90 °; bouncing-like mechanics 90-45 °) to bouncing mechanics (< 45 °). (B) Duty factor versus Fr (nonlinear regression line: a = 0.56, b = - 0.14; mean corrected R 2 = 0.689). (C) Phase shift versus duty factor (nonlinear regression line: a = 530, b = 4.95; adjusted R 2 = 0.295). (D) Percent energy recovery versus Fr (nonlinear regression line: a = 5.68, b = -1.34; mean corrected R 2 = 0.517). 39

0 s 0.03 s 0.06 0.10 s 0.13 s

Figure 2-3. Video images of a tinamou utilizing grounded running locomotion ( v = 1.58 m/s, Fr = 1.46, d = 0.57). The first image (0 s) is touchdown, the second (0.03 s) approximates the time of peak vertical force and the last (0.13 s) is liftoff of the left limb. Reflective markers represent the base of the neck, breast and synsacrum (approximating the hip joint) on the torso, the mid-tibiotarsus and ankle joint on the left limb and the tarsometatarsal-phalangeal joint and the tip on the middle toe on the left and right limbs. 40

Figure 2-4. Kinetic and kinematic parameters plotted against Froude number. (A) Peak vertical force (linear regression line: a = 0.28, b = 1.21; adjusted R 2 = 0.748), (B) relative stride length (nonlinear regression line: a = 2.14, b = 0.26; mean corrected R 2 = 0.863) and (C) stride frequency (nonlinear regression line: a = 3.56, b = 0.25; mean corrected R2 = 0.842). Symbols follow Fig. 2-2. 41 CHAPTER 3: THE EFFECT OF HEAD-BOBBING ON TERRESTRIAL

LOCOMOTION OF BIRDS: A CASE STUDY OF THE ELEGANT-CRESTED

TINAMOU ( EUDROMIA ELEGANS )

Summary

Head-bobbing is the fore-aft movement of the head relative to the body during terrestrial locomotion in birds. It is considered to be a visual response, yet some studies have suggested that the movement of the head is correlated with the movement of the hind limbs. However, these studies only analyzed locomotion at slow speeds. This study analyzed terrestrial locomotion and head-bobbing of the Elegant-crested Tinamou

(Eudromia elegans ) at a range of speeds by synchronously recording high-speed video and ground reaction forces in a laboratory setting. Results indicate that the timing of the movement of the head is dissociated from the timing of the movement of the hind limbs in tinamous. However, head and neck movements do affect the body pitch and, hence, the movement of the center of mass.

Introduction

Head-bobbing, the fore-aft movement of the head during terrestrial locomotion in some birds, is typically considered to be an optomotor response (Friedman, 1975). Head- bobbing has two distinct phases: a hold phase and a thrust phase (Dunlap and Mowrer,

1930). During the hold phase, the head is immobile (remains fixed in space) as the body travels forward, creating an illusive backward movement of the head. During the thrust phase, the speed of the head is greater than the speed of the body so that the head is 42 translated to a point in front of the body. Hold phase is believed to function in stabilizing a moving image on the retina (Dunlap and Mowrer, 1930), whereas thrust phase is thought to facilitate motion parallax to generate depth perception as objects closer to the viewer appear to move faster than objects that are farther away (Frost,

1978). Also, the thrust phase may allow for the differentiation of stationary items on a background (Davies and Green, 1988).

In addition to the visual hypothesis of head-bobbing, some researchers have

suggested that head-bobbing is linked mechanically to aspects of locomotor biodynamics.

Daanje (1951) recognized that the beginning of hold phase occurred with the liftoff of a

hind limb and that the beginning of thrust phase occurred with the touchdown of a hind

limb. From this Daanje correlated the movement of the head during walking to the

movement of the head during jumping or hopping in birds. When jumping or hopping, a

bird flexes its neck prior to the jump and extends the neck during the jump. Daanje

suggested that the movement of the head during head-bobbing is derived from the

movement of the head during jumping, and that the visual function of head-bobbing is

secondarily adapted. Dagg (1977) further suggested that movement of the head during

thrust phase of the head-bob assists with shifting the animal’s center of mass (COM)

forward as each hind limb begins to swing forward.

The kinematics of head-bobbing have been most extensively studied in pigeons

(Columba livia ; Dunlap and Mowrer, 1930; Frost, 1978; Davies and Green, 1988;

Wohlschlager et al., 1993; Troje and Frost, 2000; Fujita, 2002). The study of the

kinematics of head-bobbing in other species is limited to Chickens ( Gallus gallus ;

Dunlap and Mowrer, 1930; Pratt, 1982), Starlings (Sturnidae; Dunlap and Mowrer, 43 1930), Ring Doves ( Streptopelia risoria ; Friedman, 1975), Little Egrets ( Egretta

garzetta ; Fujita, 2003), and Whooping Cranes ( Grus americana ; Cronin et al., 2005).

Although the mechanics of terrestrial locomotion have been evaluated in a variety of

birds (Clark and Alexander, 1975; Gatesy 1999; Abourachid and Renous, 2000; Reilly,

2000; Vertappen and Aerts, 2000; Abourachid, 2001), only four studies integrate

locomotor mechanics with data on head-bobbing. Both hind limb and head kinematics

are reported for pigeons (Fujita, 2002), egrets, stilts and herons (Fujita, 2003; Fujita and

Kawakami, 2003), and gulls (Fujita, 2006). Only two of these studies (Fujita, 2002,

2003) analyzed the synchronization of hind limb and head movements. These studies

found that the beginning of hold phase begins slightly after the liftoff of the hind limb,

and that the beginning of thrust phase occurs slightly before the touchdown of the hind

limb (Fig. 3-1). Also, these studies suggested that the beginning of hold phase occurred

after the positioning of the body’s COM over the supporting hind limb and that the

beginning of thrust phase occurred as the COM was translated in front of the supporting

hind limb. Furthermore, these studies suggested that the movement of the head during

thrust phase contributed minimally to shifting of the body’s COM. As a result, Fujita

stated that head-bobbing may be biomechanically constrained by other locomotor

movements.

Additionally, differences in stride characteristics were revealed in a comparison

of head-bobbing and non-head-bobbing walking of the Black-headed Gull ( Larus

ridibundus ; Fujita, 2006), with an increase in stride length and decrease in stride

frequency when Black-headed Gulls bobbed their heada while walking. 44 Finally, head-bobbing frequency is speed dependent (Davies and Green, 1988).

As the forward speed of the bird increases, the proportion of hold phase decreases linearly. At relatively high speeds flexion and extension of the neck still occur, but there is no ‘hold phase’ when the head is stable relative to the environment. Finally, at the fastest speeds head movement is absent and the neck is extended in a constant thrust phase.

Previous studies assessing the precise relationship between head movements and footfalls have only analyzed locomotion at low speeds. As such the first objective of this study was to evaluate the relationship between head-bobbing and footfall pattern over a range of speeds. The second objective was to assess the effect of head-bobbing on the body’s COM. These objectives were examined by studying the locomotion and head- bobbing of Elegant-crested Tinamous, a basal bird species.

Materials and methods

The dynamics of terrestrial locomotion and head-bobbing were analyzed in three

Elegant-crested Tinamous (Family Tinamidae, Eudromia elegans ; 623-865 grams). The birds were filmed in a lateral view at 250 Hz using a NAC camera (Simi Valley, CA,

USA) as they progressed over a Kistler force platform (plate type 9281B; Amherst,

NY,USA) built into a 4.9 m trackway. The force platform longitudinal length (0.6 m) was adequate to capture two to five steps during each trial depending on the speed at which the birds moved. The force platform recorded vertical, fore-aft (longitudinal) and mediolateral (transverse) ground reaction forces (GRFs) at 1000 Hz. Video and force 45 data were synchronized using Motus motion analysis software (version 7.2.6, Peak

Performance Technologies, Centennial, CO, USA).

Locomotor Kinematics

Reflective markers were attached to the tip of the middle toe, synsacrum (between the femoral heads, approximating the acetabulum) and breast (between the furcula and keel). The markers and the left eye were digitized in every frame within a trial using

Motus software (version 7.2.6, Peak Performance Technologies, Centennial, CO, USA).

The forward speed of the birds was calculated videographically using the travel time of the breast marker across two 30-cm intervals marked on the back wall of the trackway and overlying the force platform. Only trials that differed in velocity by less than 10% between the initial and final intervals (steady speed) were analyzed further [100* │(v1/vt)

– ( v2/vt)│, where v1, v2 and vt are the forward velocity of the first half of the trial (30 cm),

the second half of the trial (30 cm) and the whole trial (60 cm), respectively].

The movement of the head was analyzed using the movement of the eye in space.

The onset of hold phase was defined as the first frame in which the x and y coordinates of

the eye did not vary from the coordinates in the previous frame, and the onset of thrust

phase was defined as the first frame in which the x and y coordinates varied from the previous frame after a hold phase. Relative hold phase was computed as the ratio of hold duration to head-bob cycle duration (hold phase + thrust phase). A least squares linear regression equation assessed the relationship between relative hold phase and speed.

The specific timing of events was used to assess the relationship between head and hind limb movements and GRFs. Head-bobbing events were the onset of hold phase 46 and the onset of thrust phase; footfall events were touchdown and liftoff for both the left and right hind limbs; and GRF events were the timing of peak vertical force, peak braking force (fore-aft GRF minimum), peak propulsive force (fore-aft GRF maximum) and the transition between braking and propulsive forces (fore-aft GRF = 0). Time durations between head-bobbing events and footfall and GRF events were calculated.

Relative time durations were then calculated by dividing time durations by stride duration. To determine the correspondence of head movements with hind limb movements and GRF events, relative time durations were then regressed against speed using a MANOVA (SYSTAT 11 software).

The pitch of the torso was approximated by torso angle, measured as the angle between the vertical axis passing through the synsacrum marker and a line passing from the breast marker to the synsacrum marker. We hypothesize that torso pitch may be affected by two factors: locomotor movements of the body and the movements of the head and neck. Based on work by Lee et al. (1999), a moment will be exerted about the pitch axis of the body when the fore-aft forces cause acceleration or deceleration of the

COM. The torso is expected to pitch ventrally (breast marker moves ventrally) during the braking phase of each step (from touchdown to the time of the fore-aft GRF transition); then the torso is expected to pitch dorsally (breast marker moves dorsally) during the propulsive phase (from fore-aft GRF transition to touchdown of the next hind limb; Fig. 3-2A). Therefore, a maximum torso angle is expected at each hind limb touchdown and a minimum torso angle at the time of fore-aft transition.

Moreover, the need to vertically stabilize the head relative to the environment during the hold phase is expected to result in a downward movement of the breast 47 (ventral pitch) as the torso proceeds forward under the stabilize head and the increasingly flexed neck. During the subsequent thrust phase, the head is free to move in both the vertical and fore-aft directions (Cronin et al., 2005). As a result, the extension of the neck during the thrust phase of a head-bob may cause an upward movement of the breast which would result in dorsal pitching of the torso (Fig. 3-2B). Thus, head-bobs alone may lead to a maximum torso angle at the beginning of hold phase and a minimum torso angle at the beginning of thrust phase. A synchronization of hind limb and head movements is expected to create a regular pattern of torso angle against time (Fig. 3-2C), whereas an uncoupling of these movements will create unique torso angle patterns.

COM mechanics

The movement of the COM was quantified using the vertical, fore-aft and mediolateral GRFs. COM movements can be used to distinguish the COM mechanics being employed (Cavagna et al., 1977). During slow speed locomotion the kinetic and gravitational potential energies of the COM cycle out-of-phase with one another, allowing for external mechanical energy to be recovered via pendular mechanics.

Conversely, during fast speed locomotion the kinetic and gravitational potential energies cycle in-phase and external mechanical energy is poorly recovered via pendular mechanics. Instead, spring-mass mechanics are employed and external mechanical energy is recovered by using the muscles and tendons in the hind limbs to store and release elastic energy during each support duration. In the present study, COM mechanics were compared in steps in which a hold phase occurred and steps entirely within a thrust phase. 48 A customized LabView program (National Instruments, Austin, TX, USA) was used to calculate fluctuations in kinetic (Ek) and gravitational potential ( Ep) energies of the COM from the GRFs following Cavagna et al. (1977) and Hancock et al. (2007).

Vertical (minus body weight), fore-aft and mediolateral GRFs were divided by the body mass to determine acceleration in each direction. Each acceleration was then integrated once to obtain velocities of the COM in the three directions, and vertical velocity was further integrated to obtain vertical displacement of the COM. Average forward speed was used as the integration constant for fore-aft velocity, whereas vertical and mediolateral integration constants were estimated as the mean values for vertical and mediolateral records. Velocities were then used to calculate Ek for each direction as Ek =

½ mv 2, where m is the animal’s mass and v is velocity. Summing vertical, fore-aft and mediolateral Ek yielded total kinetic energy ( Ek,tot ). Gravitational potential energy was

computed as Ep = mgh where h is the vertical displacement of the COM. Lastly, total external mechanical energy ( Em,tot ) was computed as Ek,tot + Ep.

The phase shift between Ek,tot and Ep was calculated by dividing the time

difference between the minimum values of Ek,tot and Ep by the duration of the stride and

then multiplying that value by 360° (Cavagna et al., 1977). Following Ahn et al. (2004),

birds were considered to use pendular mechanics when the phase shift for a step was

between 135° and 180°, spring-mass mechanics when the phase shift for a step was

between 0°and 45°, and mixed mechanics (a combination of pendular and spring-mass

mechanics) in steps with phase shift values between 45° and 135°.

The amount of external mechanical energy recovered via pendular mechanics was

calculated following Blickhan and Full (1992): 49 %R = 100*[( ∆Ep + ∆Ek,tot ) - ∆Em,tot ]/( ∆Ep + ∆Ek,tot )

where ∆Ep, ∆Ek,tot and ∆Em,tot are the sums of the positive increments of the Ep, Ek,tot and

Em,tot profiles, respectively.

The timing of the peak vertical force and fore-aft force transition from braking to propulsion were compared in both steps with a hold phase and steps entirely within a thrust phase. Relative times were calculated for both GRF events by dividing the duration from touchdown to the event by the stance duration.

The relationship between locomotor mechanics parameters (phase shift, percent recovery, relative times) and head-bobbing were evaluated for steps in which a hold phase occurred and steps entirely within a thrust phase. Each parameter was regressed against speed using a least squares regression, and t-tests were used to compare the slopes and elevations of the equations obtained for the hold and thrust phases.

Results

Effect of speed and footfall on head-bobbing

We obtained 67 steady speed trials with speeds ranging from 0.43 to 1.60 ms -1

(1.05 ± 0.28 ms -1, mean ± standard deviation). Head-bobbing patterns changed

predictably with speed. The relative hold phase duration decreased from 0.51 to 0.06

(0.26 ± 0.10) as speed increased (regression line: y = -0.31x + 0.59, adjusted R 2 = 0.708,

P > 0.001; Fig. 3-3A). Less predictable relationships were obtained between head- bobbing and footfall or GRF events (Table 3-1). For example, the relative duration from hind limb liftoff to the beginning of hold phase ranged from -0.85 to 0.36, i.e. -85% to

36% of stride (-0.12 ± 0.24; Fig. 3-3B). Also, the relative duration of time between hind 50 limb touchdown to the beginning of thrust phase ranged from -0.93 to 0.68, i.e. -93% to 68% of stride (0.14 ± 0.31; Fig. 3-3C). An extremely low adjusted R 2 (0.083) for the

regression of the relative time duration from hind limb liftoff to the beginning of hold

phase against speed was also observed (Table 3-1). These results show that the hold

phase in tinamous can be initiated and completed at any time within a step. A similar

dissociation was found between all other footfall and GRF events and head-bobbing

events (Table 3-1): the magnitudes of the ranges were similar and all regressions against

speed had low adjusted R 2 values (< 0.3). Therefore, the timing of head-bobbing events

in tinamous was independent of the timing of footfall and GRF events.

Effect of head-bobbing on ground reaction forces

Ground reaction force profiles were most strongly influenced by locomotor speed,

as noted by comparing fast and slow trials (Fig. 3-4). High-speed locomotion resulted in

single-peaked vertical GRFs for each step, characteristic of animals moving with

bouncing mechanics (Fig. 3-4A); in these fast trials the necks of the tinamous were

continually extended and no head-bobbing occurred. At the slowest speed, vertical GRFs

were double-peaked, typical of inverted pendular mechanics (Fig. 3-4D); frequent head-

bobbing in these trials resulted in a hold phase with every step.

Overlaid on this general pattern are modifications to the GRFs due to head-

bobbing. Distinct differences in the GRFs were noted between steps in which a hold

phase occurred and steps entirely within a thrust phase. This effect was most evident in

intermediate speed locomotion: the vertical GRFs of steps in which a hold phase occurred

were either double-peaked or plateaued (second step in Fig. 3-4C), whereas steps entirely 51 within a thrust phase displayed single-peaked profiles (first and third steps in Fig. 3-

4C). With an increase in speed, the effect of head-bobbing on GRF profiles was more subtle but still evident (Fig. 3-4B), namely, steps in which a hold phase occurred were only slightly plateaued compared to steps occurring entirely in the thrust phase.

Another aspect of locomotor biodynamics affected by head-bobbing is stance duration. There is a tendency for stance durations of steps in which a hold phase occurred to be slightly longer than the steps within a thrust phase (Fig. 3-5). This effect was more evident at slower speeds, although the difference is not significant over the entire sample.

Effect of head-bobbing on torso pitching

Both locomotor dynamics and head-bobbing affected dorsoventral pitching of the torso. The effect of locomotor dynamics is best noted in the fastest trials without head- bobbing (Fig. 3-4A). In these trials, the torso angle decreased (ventral pitch) from the touchdown of a hind limb until the transition of the fore-aft force from braking to propulsion, and the torso angle increased (dorsal pitch) from the transition of the fore-aft force until the touchdown of the opposite hind limb. In other words, the torso pitches ventrally then dorsally during the each support duration, as we hypothesized due to locomotor dynamics alone (Fig. 3-2A).

Whereas the influence of locomotor dynamics on torso pitch is also evident in slower trials, head-bobbing interrupts the normal cycling pattern so that torso angle magnitudes do not precisely mirror the GRFs in fast and intermediate speed locomotion with head-bobbing (Fig. 3-4B and C). Our hypothesis that the hold phase of head- 52 bobbing should correspond to a ventral pitching of the torso (Fig. 3-2) is borne out by these trials. When a hold phase occurred during the middle or end of a step (Fig. 3-4C, second step; Fig. 3-4B, first step), the dorsal pitching of the torso due to hind limb movements is moderated by the need to accommodate the flexing neck and vertically stabilized head. This effect is observed to a lesser degree in slow speed locomotion with head-bobbing (Fig. 3-4D). During the thrust phase of a head-bob, the head moves both in the horizontal and vertical planes. As a result, the pitching of the torso caused by locomotor dynamics is not restricted during thrust phase. Ultimately, the pitching of the torso and the GRF profile are primarily affected by locomotor dynamics and secondarily affected by the hold phase of head-bobbing.

Effect of head-bobbing on COM mechanics

The time of peak vertical force relative to stance duration ranged from 0.09 to

0.40 (0.25 ± 0.07; Fig. 3-6). Steps with a hold phase and steps entirely within a thrust phase had different values for the relative time of peak vertical force (P < 0.02), in particular, the steps with a hold phase had earlier peak vertical forces. The time of the fore-aft force transition relative to stance duration ranged from 0.26 to 0.50 (0.37 ± 0.04).

Steps with a hold phase and steps entirely within a thrust phase had similar values for the relative time of fore-aft force transition (P > 0.05).

Across the entire sample, the phase relationship between the minima of the profiles of gravitational potential energy and kinetic energy ranged from 0° to 165.7°

(Fig. 3-7A). Tinamous in this study predominately moved with bouncing mechanics

(phase shift < 45° in 58% of the trials). Thirty-six percent of trials were mixed 53 mechanics (phase shift between 45° and 135°), whereas only 4 trials were vaulting or pendular mechanics (phase shift > 135°). No differences were found in phase shift between steps with a hold phase and steps within a thrust phase (P > 0.05). All of the trials with vaulting mechanics had a hold phase occurring with every step; as a result, there were no vaulting mechanics steps that were entirely within a thrust phase.

The amount of mechanical energy recovered using pendulum-like mechanics ranged from 0.3% to 57.6% (Fig. 3-7B). Trials with bouncing mechanics had low energy recoveries (range: 0 – 17.5%; 5.9 ± 4.1), trials with mixed mechanics tended to have intermediate energy recoveries (2.3 – 44.1%; 20.8 ± 12.8) and trials with vaulting mechanics had high energy recoveries (33.9 – 57.6%; 48.1 ± 10.7). Steps with a hold phase and steps entirely within a thrust phase had similar values for energy recovery (P >

0.05). Therefore, COM mechanics were not demonstrably affected by head-bobbing.

Discussion

Although visual cues appear to be the primary trigger for head-bobbing in birds, some birds also synchronize the movement of the head with the movements of the hind limbs. Synchronization has been observed at slow speeds in members of the

Columbiformes (Fujita, 2002), Ciconiiformes (Fujita, 2003; Fujita and Kawakami, 2003), and Charadriiformes (Fujita, 2006; Fujita and Kawakami, 2003; Hancock, Chapter 4).

Most birds bob their head once per step (twice per stride). Among these birds, hold phase occurs during the single-limb support phase (swing phase of the contralateral hind limb), with thrust phase being initiated during the later part of the single-limb support phase and continuing throughout the period of double-limb support. The coordination of the hold 54 phase with the single-limb support may be related to visual stability, as it would avoid any potential jarring effect when the contralateral hind limb contacts the ground. Yet another function may be to enhance postural stability during terrestrial locomotion, as theorized by Dagg and (1977) and Fujita (2002, 2003). The backward movement of the head and neck (relative to the COM) during the hold phase may help moderate how far forward the COM shifts relative to the supporting foot during the single-limb support phase. Similarly, the forward thrusting of the head and neck (relative to the COM) during the double-limb support phase may help reposition the COM over the supporting foot for the beginning of the subsequent single-limb support phase.

Regardless of the potential visual or postural benefits, the Elegant-crested

Tinamous in this study did not always synchronize head-bobbing with hind limb movements in the majority of the trials. Instead, the timing of head movements was found to be independent of the timing of hind limb movements and GRF events. We offer three possible explanations for why head-bobs need not be tightly coupled with hind limb movements, including anatomical, behavioral and phylogenetic factors.

First, there appears to be no anatomical constraint that would require

synchronization in birds, as illustrated by the following situations in which head and hind

limb movements occur independent of one another. When head-bobbing birds are

moving at their fastest speeds and with their highest stride frequencies the hold phase is

absent and the neck is extended in a constant thrust phase (Davies and Green, 1988).

Furthermore, head-bobbing birds walking on treadmills do not bob (Frost, 1978), nor do

blindfolded birds (Necker et al., 2000), presumably because a streaming visual signal is

suppressed. Finally, head-bobbing has been observed during landing when the hind 55 limbs are held against the body (Green et al., 1994). Hence, head-bobbing birds can and do have hind limb movements without head-bobs as well as head-bobs without hind limb movements.

Second, head-bobbing may be induced by a need to process particular types of visual signals. Davies and Green (1988) hypothesized that recognition of moving objects occurs during the hold phase whereas the thrust phase functions to provide depth perception and thereby enhance recognition of objects that are stationary. Foraging has been hypothesized to play an important role in the initiation of head-bobbing. Those studies that have found head-bobbing to be synchronized with hind limb movements were performed in field settings, with birds able to move freely and forage. Furthermore,

Black-headed Gulls ( Larus ridibundus ) head-bob while foraging but do not head-bob

while engaged in non-foraging walking (Fujita, 2006). Similarly, Pacific Reef Egrets

(Egretta sacra ) and Gray Herons ( Ardea cinerea ) cease head-bobbing when they are not foraging (Fujita and Kawakami, 2003). In the current study, tinamous were filmed on a featureless trackway in the laboratory setting and were not foraging during locomotor bouts. Future studies could incorporate a more natural setting in order to assess whether tinamous similarly synchronize head and hind limb movements during foraging.

Whereas the first two factors illustrate situational disconnections between head and hind limb movements, there is a third possible factor for the lack of head-hind limb synchronization in tinamous that is rooted in the evolutionary history of birds. The bodies of modern birds have been suggested to be segregated into different locomotor modules (Gatesy and Dial, 1996). Over evolutionary time, the movements of the forelimb module in nonavian theropod dinosaurs (the hypothesized clade from which 56 birds are derived) were functionally decoupled from those of the hind limbs and tail; subsequently, tail movements were decoupled from hind limb movements.

Consequently, in most modern birds (Neornithes) the forelimb (wing) and tail modules are used in flight whereas the hind limb module is used primarily during terrestrial locomotion (and typically during take-off and landing features associated with flight).

Situational coupling of the hind limb module with the head and neck during terrestrial locomotion (head-bobbing) in the form of precise synchronization of head and hind limb movements has only been reported in neognathe birds (Columbiformes [Fujita, 2002],

Ciconiiformes [Fujita, 2003; Fujita and Kawakami, 2003] and Charadriiformes [Fujita,

2006; Fujita and Kawakami, 2003]). Tinamous are a paleognath bird, thus it is possible that head-hind limb synchronization evolved primarily in neognaths. Additional data on head and hind limb movements in other paleognaths is necessary to verify this hypothesis.

Despite the irregularity of head-bobs relative to footfalls, head-bobbing in tinamous has a predictable effect on the mechanics of the body’s COM. The influence of the head alone is likely to be trivial, since the head and neck constitutes only 5.4% of total body mass (Hancock, pers. ob.). Rather, the COM of the body is much more affected by the compensatory pitching of the torso that accompanies head-bobs (Fig. 3-

4B). The torso must pitch ventrally during the hold phase so that the head is held stable both horizontally and vertically. Rotation of the torso during pitching occurs about the acetabulum. Because the torso of tinamous is largely horizontal in orientation, pitching will displace the body’s COM more vertically than horizontally (Fig. 3-8).

Consequently, the hold phase has a greater effect on the vertical ground reaction force 57 record than on the fore-aft force record (as seen in Fig. 3-4B, C). Yet even this is trivial: there is a difference in the timing of peak vertical forces between steps within a hold phase versus a thrust phase (Fig. 3-6B), but this does not translate into significant differences in COM mechanics (Fig. 3-7).

In conclusion, tinamous in the lab did not always synchronize head-bobbing and hind limb movements. Compensatory pitching of the torso to stabilize the head during the hold phase has a predictable, but quantitatively trivial, effect on whole body ground reaction forces and mechanics. To better understand the coordination of head-bobbing with hind limb movements further studies need to be performed on paleognaths and neognathes both in laboratory and field settings. Also, the effect of head-bobbing on the mechanics of locomotion should be analyzed in birds whose head and neck represent a larger proportion of their total body mass. 58 Table 3-1: Relative Durations between Locomotor Events and Head-bobbing Events.

Relative durations (duration/stride duration) Range Mean Adjusted R 2 Touchdown to the beginning of hold phase -0.98 – -0.02 -0.33 0.018 Touchdown to the beginning of thrust phase -0.93 – 0.68 0.14 0.205 Liftoff to the beginning of hold phase -0.84 – 0.36 -0.12 0.083 Liftoff to the beginning of thrust phase -0.45 – 0.69 0.33 0.049 Time of peak vertical force to the beginning of hold phase -0.78 – 0.31 -0.09 0.059 Time of peak vertical force to the beginning of thrust phase -1.73 – -0.09 -0.62 0.277 Time of fore-aft force transition to the beginning of hold phase -0.57 – 0.38 0.09 0.000 Time of fore-aft force transition to the beginning of thrust phase -1.37 – 0.11 -0.44 0.153 Time of peak braking force to the beginning of hold phase -0.76 – 0.11 -0.16 0.003 Time of peak braking force to the beginning of thrust phase -1.71 – -0.18 -0.69 0.173 Time of peak propulsive force to the beginning of hold phase -0.29 – 0.87 0.54 0.063 Time of peak propulsive force to the beginning of thrust phase -0.93 – 0.39 0 0.037

Figure 3-1. Sagittal movements (in the x-axis) of the head relative to stance phases of the feet in a walking pigeon (modified from Fujita, 2002).

Figure 3-2. A) The hypothesized effect of footfalls on torso pitch and torso angle over the time of a single trial. Open gray and black bars represent the stance phases of the right and left hind limbs, respectively. Line drawings depict the expected torso pitch due to accelerative and decelerative forces generated by the limbs (Lee et al., 1999). Torso pitch is represented by the orientation of the line between the synsacrum and breast on the birds. The curve represents the torso angle expected due to orientation of the fore-aft ground reaction forces relative to the COM. B) The hypothesized effect of head-bobbing on torso pitch and torso angle over the time of one trial. Filled gray and black bars represent thrust phase and hold phase, respectively. Line drawings depict the expected torso pitch due to the movements of the head and neck. The curve represents the expected torso angle due to movements of the head and neck. C) Hypothesized changes in torso angle resulting from the summed influences of limb and head movements. 61 0.6 A 0.5

0.4

0.3

0.2 Relative holdphase 0.1

0 0 0.5 1 1.5 2 Speed (m/s)

1 0.8 B 0.6 0.4 0.2 0 -0.2 -0.4 -0.6 beginning of hold ofhold phase beginning -0.8

Relative duration from liftoff to the the to liftoff from duration Relative -1 0 0.5 1 1.5 2 Speed (m/s) 1 0.8 C 0.6 0.4 0.2 0 -0.2 -0.4 -0.6

the the beginningthrustof phase -0.8

Relative durationfromtouchdown to -1 0 0.5 1 1.5 2 Speed (m/s)

Figure 3-3. A) Relative hold phase as a function of speed. The solid line represents the regression line (y = -0.31x + 0.59, adjusted R 2 = 0.708, P > 0.001) in tinamous. The dashed line represents the regression equation for pigeons (relative hold phase = 0.67- 0.88*absolute speed, Pearson r = 0.97, P > 0.01) given by Davies and Green (1988). B) Relative duration of time from the beginning of hold phase to liftoff as a function of speed. 62

Figure 3-4. The top graphs in each section depict the vertical (black curve) and fore-aft (gray curve) GRFs in four representative trials. The bottom graphs in each section show the corresponding torso angles. Gray and black bars above the graphs represent thrust phase and hold phase, respectively. A) In a fast speed trial (1.74 ms -1) without head- bobbing, the torso angle magnitude inversely mirrors the vertical GRFs. B,C) In fast (1.56 ms -1) and intermediate (1.13 ms -1) trials with head-bobbing, the torso angle does not mirror the GRFs. D) In a slow speed trial (0.76 ms-1) with head-bobbing, hold phases can occur with every step. However, the relationship between torso angle and GRF is determined by the timing of hold phase within an individual step. 63

0.3 Hold 0.26 Thrust

0.22

0.18

Stance duration (s) duration Stance 0.14

0.1 0.5 0.75 1 1.25 1.5 1.75 Speed (m/s)

Figure 3-5. Stance duration as a function of speed. Black squares represent stance durations for steps in which a hold phase occurs and gray squares represent stance durations for steps entirely within a thrust phase. Stance duration is slightly longer in steps that include a hold phase.

64 0.6 A Hold 0.5 Thrust

0.4

0.3

RelativeTime Fore-aftof Transition 0.2 0.58 0.78 0.98 1.18 1.38 1.58 Speed (m/s)

0.45 B

0.35

0.25

0.15

Relative Vertical Relative ofPeak Time Force 0.05 0.58 0.78 0.98 1.18 1.38 1.58 Speed (m/s)

Figure 3-6. A) Relative timing of the fore-aft ground reaction force transition as a function of speed. Black squares represent steps in which a hold phase occurs and gray squares represent steps entirely within a thrust phase. Relative timing of the fore-aft force transition of steps with and without a hold phase are not significantly different. B) Relative timing of peak vertical force as a function of speed. Black squares represent steps in which a hold phase occurs and gray squares indicate steps entirely within a thrust phase. No significant differences are observed in the timing of peak vertical force between steps with and without a hold phase.

65 180 A Hold 135 Thrust

45 Phase shift (degrees) shift Phase

0 0.5 0.75 1 1.25 1.5 1.75 Speed (m/s)

60 B

20 Percent energy recovery (%) recovery energy Percent 0 0.5 0.75 1 1.25 1.5 1.75 Speed (m/s)

Figure 3-7. A) Phase shift between the minima of Ep and Ek,tot curves as a function of speed. Black squares represent steps in which a hold phase occurs and gray squares represent steps entirely within a thrust phase. Phase shifts of steps with and without a hold phase are not significantly different. B) Percent of mechanical energy recovered via pendulum-like mechanisms as a function of speed. Black squares represent steps in which a hold phase occurs and gray squares represent steps entirely within a thrust phase. Mechanical energy recoveries of steps with and without a hold phase are not significantly different. 66

Figure 3-8. A diagram of COM movement during pitching of the torso. The solid line drawing represents a ventrally pitching bird during braking and the dashed line drawing represents a dorsally pitching bird during propulsion. Pitching of the bird torso results in more vertical than horizontal movement of the COM. 67

CHAPTER 4: THE MECHANICS OF LOCOMOTION AND HEAD-BOBBING IN

CHARADRIIFORM BIRDS

Summary

Head-bobbing is the fore-aft movement of the head relative to the body during terrestrial locomotion in birds. Head-bobbing is predominately considered to be an optomotor response, in which the hold phase functions to stabilize an image on the retina.

However, some studies have suggested that head-bobbing is linked mechanically to aspects of locomotor biodynamics. The goals of the study were (1) to assess whether head-bobbing was synchronized with hind limb movements in charadriiform birds and

(2) to compare the kinematics of locomotion among head-bobbing and non-head-bobbing charadriiform species. Freely-moving birds were filmed in the field. Lateral views of the birds were analyzed by recording the timing of head-bobbing and locomotor events and by calculating kinematic variables in both head-bobbing and non-head-bobbing birds.

The results suggest that head movements and hind limb movements are moderately coordinated in charadriiform birds. The onset of hold phase relative to liftoff was similar between species whereas the beginning of thrust phase relative to touchdown showed more variability. Also, results show that stride duration and stance duration differed significantly in head-bobbing and non-head-bobbing species. As footfall patterns may affect the stability of the visual system, these results provide additional support for the optomotor hypothesis of head-bobbing.

Introduction

Head-bobbing is an optomotor response during terrestrial locomotion in some birds (Friedman, 1975). The head-bobbing cycle has two phases, a hold phase and a thrust phase (Dunlap and Mowrer, 1930). During the hold phase the head remains stable relative to the environment as the body moves forward, and during the thrust phase the head is accelerated in front of the body. The hold phase of head-bobbing is suggested to allow stabilization of a visual image on the retina during terrestrial locomotion in birds.

If movements of the hind limbs (including the cyclical collisions with the ground at touchdown) disrupt stability of the visual system, then some coordination of head movements with hind limb movements may be expected in order to allow for better image quality.

Some previous studies reported a predictable pattern between head and hind limb movements (Fujita, 2002, 2003): the beginning of hold phase occurs slightly after hind limb liftoff and the subsequent thrust phase begins slightly before the touchdown of that same hind limb. However, this pattern is not ubiquitous across birds. One study found that the timing of head and hind limb movements was more irregular, i.e., the onset of hold or thrust phases could occur at any time during the stride (Hancock, chapter 3).

Differences in the design of these studies may account for their conflicting results. The studies that reported synchronization were performed in the field as the birds were naturally moving, whereas the study in which synchronization rarely occurred was performed in the laboratory as birds moved over a trackway upon encouragement of researchers. The richer visual and substrate variations in the natural setting are likely to 69 provide a very different stimulatory environment for both the optomotor response and locomotor function compared to the lab. Secondly, the birds that have been shown to synchronize head and hind limb movements are neognathes (Pigeons [ Columba livia ] and

Little Egrets [ Ergetta garzetta ]), whereas the birds that did not always synchronize are paleognaths (Elegant-crested Tinamous [ Eudromia elegans ]), therefore, there may be a

phylogenetic basis for the head-hind limb synchronization patterns reported.

Additionally, some birds are facultative head-bobbers, capable of walking both

with and without head-bobbing (Fujita, 2006; Ortega, 2005). Studies on these birds

provide insight into the function of head-bobbing as well as the link between head-

bobbing and locomotor cycles. For example, Black-headed Gulls ( Larus ridibundus ) display head-bobbing when foraging and non-head-bobbing walking when not foraging – strong evidence for the visual function of head-bobbing during foraging – and these gulls move with longer stride lengths when they do head-bob – evidence that head and hind limb movements may not be completely independent (Fujita, 2006).

The present study analyzed kinematics of terrestrial locomotion and head-bobbing of charadriiform birds in a natural setting. The first goal of the study was to assess whether head-bobbing was synchronized with hind limb movements in multiple charadriiform species as the birds moved through their natural environment. The second goal was to compare the kinematics of locomotion of head-bobbing and non-head- bobbing charadriiform birds.

70 Materials and Methods

Fifteen species of charadriiform birds were examined in this study (Table 4-1;

Fig. 4-1): Black-bellied Plover ( Pluvialis squatarola ), Black-necked Stilt ( Himantopus mexicanus ), Black Skimmer (Rynchops niger ), Greater Yellowlegs ( Tringa melanoleuca ), Killdeer ( Charadrius vociferous ), Laughing Gull ( Larus atricilla ), Least

Sandpiper ( Calidris minutilla ), Lesser Yellowlegs ( Tringa flavipes ), Marbled Godwit

(Limosa fedoa ), Ring-billed Gull ( Larus delawarensis ), Royal Tern ( Sterna maxima ),

Sanderling ( Calidris alba ), Semipalmated Plover ( Charadrius semipalmatus ), Whimbrel

(Numenius phaeopus ) and Willet ( Tringa semipalmatus ). Head-bobbing and terrestrial locomotion were evaluated by direct observations of birds in their natural habitat. These birds were investigated opportunistically at three locations: Anahuac National Wildlife

Refuge, Texas; St. Marks National Wildlife Refuge, Florida; and Ottawa National

Wildlife Refuge, Ohio.

Individual birds were located, identified and observed for a maximum period of ten minutes. All observations were recorded at 60 Hz using a JVC digital video camera

(model DVL 9800). Video sequences were uploaded using Motus motion analysis software (version 7.2.6, Peak Performance Technologies, Centennial, CO, U.S.A.).

Three points were digitized on the birds: one eye, the most anterior point on the breast and the dorsal surface of the base of the tail. Body length was approximated as the distance between the breast and tail base. The raw x coordinates of the breast and tail base were used to calculate speed as a function of body length.

The raw x coordinates of the eye were plotted against time and used to determine the beginning of hold phase and the beginning of thrust phase in each head-bobbing 71 cycle. The onset of hold phase was defined as the first frame in which the x

coordinates of the eye did not vary from the x coordinates in the previous frame, and the

onset of thrust phase was defined as the first frame in which the x coordinates varied from

the previous frame after a hold phase. Relative hold phase was calculated as the duration

of hold phase divided by the duration of the head-bob cycle (hold phase plus the thrust

phase). Relative hold phase was regressed against speed using a least squares linear

regression for all species combined, as well as for each individual species. The time of

the touchdown and liftoff of each hind limb were also recorded.

The timing of head-bobbing events (beginning of hold phase and thrust phase)

were compared to the timing of liftoff and touchdown of the hind limbs in order to

determine the degree of synchronization of head and hind limb movements. The relative

duration from liftoff to the beginning of hold phase was calculated by subtracting the

time at the beginning of hold phase from the time at the liftoff of a hind limb and then

dividing the difference by stride duration. Similarly, the relative duration from

touchdown to the beginning of thrust phase was calculated by subtracting the time at the

beginning of thrust phase from the time at the touchdown of a hind limb and then

dividing the difference by stride duration. Absolute synchronization of the head-bobbing

and footfall events was evaluated by performing t-tests to determine if the means of the

relative durations equaled zero. Additionally, an ANOVA was performed to determine

whether the relative durations differed among species.

Finally, locomotor kinematic parameters were compared between head-bobbing

and non-head-bobbing birds. First, stride duration (time from the touchdown of a hind

limb to the subsequent touchdown of the same hind limb) and stance duration (time from 72 the touchdown of a hind limb to the liftoff of the same hind limb) were computed.

Then the log 10 -transformed kinematic variables for head-bobbing and non-head-bobbing

species were regressed against log 10 -transformed speed using a least-squares linear

regression algorithm, and slopes and elevations of regression equations were compared

between head-bobbing and non-head-bobbing birds using t-tests.

Results

Head-bobbing and locomotor kinematics

The birds that displayed head-bobbing in this study were the Sanderlings, Least

Sandpipers, Lesser Yellowlegs, Greater Yellowlegs, Willets, Marbled Godwits,

Whimbrels, Black-necked Stilts and Black-bellied Plovers (Table 4-1). The relative hold

phase of head-bobbing birds ranged widely, from 0.09 to 0.86 (Fig. 4-2A; 0.44 ± 0.16

sec, mean ± SD), and there was no significant change in relative hold phase with speed

across all species. Within species, however, a significant speed-dependent decrease in

relative hold phase was found only for the Least Sandpipers (Fig. 4-2B; y = -0.05x +

0.6086; R 2 = 0.2339; P < 0.05) and Black-bellied Plovers (y = -0.1777x + 0.6493; R 2 =

0.2273; P < 0.05).

Hind limb and head movements were found to be fairly well coordinated but not

precisely synchronized in the charadriiform birds. The onset of hold phase and the

timing of hind limb liftoff corresponded closely: the relative duration from liftoff to the

beginning of hold phase ranged from -0.30 to 0.30 (0.02 ± 0.07; Fig. 4-3A) but, on

average, hold phase did not initiate precisely at liftoff (mean relative duration did not

equal zero in t-test; P<0.001). There was a slight bias toward initiating the hold phase 73 before the hind limb liftoff (55% of the trials) although in 18% of the trials these events occurred at the same time and in 27% of trials the beginning of hold phase occurred after the liftoff of a hind limb. There was no difference between the relative duration from liftoff to the beginning of hold phase of different species (Fig. 4-3B; F =

1.227; P > 0.2).

The relative timing of hind limb touchdown and the onset of thrust phase had

similar results. Thrust phase could begin either before or after touchdown (Fig. 4-4A;

relative duration from touchdown to the beginning of thrust phase ranged from -0.24 to

0.29; 0.09 ± 0.08) but, on average, precise synchrony was not found (mean relative

duration did not equal zero in t-test; P<0.001). The tendency for thrust phase to precede

hind limb touchdown was much stronger that the tendency for hold phase to begin before

liftoff. In 87% of the trials the beginning of thrust phase occurred before the touchdown

of a hind limb, whereas these events occurred simultaneously in only 5% of the trials and

in 8% of trials the beginning of thrust phase occurred after hind limb touchdown. A

difference between species was found in the relative duration from touchdown to the

beginning of thrust phase (Fig. 4-4B; F = 7.322; P < 0.001). Post-hoc t-tests revealed that

the relative touchdown to thrust differed significantly between Whimbrels and all other

species (P < 0.0001 in all comparisons) as well as between Greater Yellowlegs and four

other species: Black-bellied Plovers (P < 0.01), Black-necked Stilts (P < 0.05), Lesser

Yellowlegs (P < 0.5), and Willets (P < 0.01). All other species comparisons were not

significantly different.

74 Locomotor kinematics in head-bobbing and non-head-bobbing species

Six species of charadriiform birds were never observed to head-bob (Ring-billed

Gulls, Laughing Gulls, Royal Terns, Black Skimmers, Semipalmated Plovers and

Killdeer) and these were compared to the 9 head-bobbing species with regard to locomotor speed and hind limb kinematics (Table 4-1). The speeds of head-bobbing birds ranged from 0.72 to 20.2 body lengths per second (4.07 ± 3.13 body lengths per sec), and the speeds of non-head-bobbing birds ranged from 0.92 to 15.7 body lengths per second (3.59 ± 3.04 body length per sec). These speeds were not statistically different (P

= 0.055).

Across these speeds, there was a broad overlap in stance and stride durations between the head-bobbing and non-head-bobbing birds, but differences are apparent

(Figs. 4-5B and 4-6B). Regression lines were significantly different (both at P<0.001;

Table 4-2), and the steeper slopes in the head-bobbing birds resulted in longer, on average, stance and stride durations at the slower speeds. The result is that head-bobbing birds tended to have longer stance durations (range 0.05-1.73 sec; 0.38 ± 0.25 sec) than do non-head-bobbing species (range 0.07-0.97 sec; 0.33 ± 0.18 sec) but were distinctive only at slower speeds. A similar pattern was observed for stride duration (Fig. 4-6): stride durations were longer in head-bobbing species (range 0.13-2.45 sec; 0.56 ± 0.32 sec) than non-head-bobbing species (range 0.15-1.3 sec; 0.47 ± 0.23 sec) but were distinctive only at the slower speeds. Differences in stance and stride durations between head-bobbing and non-head-bobbing species were not evident at faster speeds.

This order-wide tendency for stance and stride durations to be elongated in head-

bobbing birds was mirrored in the narrower comparison of the locomotor kinematics 75 between two species within the family Charadriidae: the head-bobbing Black-bellied

Plovers and the non-head-bobbing Semipalmated Plovers. Stance durations tended to be longer in the Black-bellied Plovers (range 0.1-0.97 sec; 0.31 ± 0.19 sec) compared with the Semipalmated Plovers (range 0.07 to 0.57 sec; mean 0.18 ± 0.12 sec) due to the particularly long stance durations at the slowest speed and significantly different regression lines of stance duration on speed (P<0.001; Fig. 4-7A; Table 4-2). Similarly, stride durations were longer in the head-bobbing Black-bellied Plovers (range 0.21-1.3;

0.47 ± 0.25) compared with the non-head-bobbing Semipalmated Plovers (0.15-0.72;

0.28 ± 0.14; Fig. 4-8A).

Discussion

When Pigeons and Elegant-Crested Tinamous increase speed during terrestrial

locomotion, they predictably change their head-bob cycle by reducing the duration of the

hold phase (Davies and Green, 1988; Hancock, Chapter 3). A similar pattern was found

for the Least Sandpipers and Black-bellied Plovers in this study. The reductions in

relative hold phase for these species are statistically significant, yet there is substantially

greater variability than was observed in the Elegant-crested Tinamous (Fig. 4-2C; only

means were plotted for the Pigeons, thus the degree of variability is unknown).

Furthermore, the remaining seven species of head-bobbing charadriiform birds failed to

display significant speed-related reductions in relative hold phase. Why are head-bob

cycles so poorly correlated with speed among charadriiforms? The answer is revealed by

viewing the head-bobbing cycles within any one trial (Fig. 4-9). Whereas thrust phase

durations (rapid forward progressions of the eye during brief periods of double limb 76 support) tend to remain largely consistent across a trial, hold phase durations can vary substantially. In the illustrated example (Fig. 4-9), hold phases B and F are more than twice as long as the other hold phases. Consequently, at a similar speed, a bird may elongate the hold phase during one step and shorten it in the subsequent step. The impetus for such variability in hold phase durations may be a periodic need to elongate the duration of head, and thus visual, stabilization. Indeed, a key difference from the

Pigeon and Tinamous studies is that the present study was conducted in the birds’ natural environment and, in most cases, the birds were recorded while foraging.

Further evidence to support the visual foraging hypothesis is found by considering head-bobbing cycles together with footfall patterns. In charadriiform birds, most hold phases begin just before liftoff of a hind limb and continue until just before touchdown.

In Pigeons and Little Egrets the hold phase spans the period from slightly after the liftoff of a hind limb to just before the touchdown of the same hind limb (Fujita, 2002, 2003).

Whereas there is a slight shift in the onset of hold phase and liftoff between these studies, the net result is the same: in each case, the bird’s head is largely stabilized in space during the period of single-limb support (gray boxes in Fig. 4-9). Horizontal and vertical stability of the head (hold phase) during the single-limb support phase occurs by progressively flexing the superior neck and hyperextending the inferior neck as the torso continues to move forward. Vertical stability is further enhanced by two means: the predominant use of grounded locomotion in birds and the delaying of touchdown until the end of hold phase.

Many birds transition seamlessly between walks and runs by using grounded runs, i.e., running mechanics without an aerial phase (Gatesy and Biewener, 1991; Rubenson et 77 al., 2004; Hancock et al., 2007). This locomotor behavior is accomplished in birds by implementing a more crouched posture and increasing hind limb compliance (Gatesy and

Biewener, 1991). It is dynamically similar to Groucho running, a gait that occurs when humans move with flexed knees and elongated stride lengths (McMahon et al., 1987).

During Groucho running, vertical oscillations of the center of mass and head are dampened. This vertical stabilization of the head also occurs during grounded running in birds (Gatesy and Biewener, 1991; Hancock et al., 2007). Therefore, because single-limb support phase during grounded runs limits the vertical displacement of the head and because it coincides with hold phase, the visual system will minimally shift (if at all) in the vertical plane during the hold phase.

The end of the single-limb support phase occurs when the swinging hind limb touches the ground. Touchdown of the hind limb is considered to be akin to a collision with the ground that redirects the movement of the center of mass and thus the head, imparting a vertical acceleration away from the ground (McMahon et al., 1987; Ruina et al., 2005). If the hold phase functions to maximize visual stability, then birds should attempt to avoid the potentially jarring effects of this collision by ending the hold phase before touchdown. Indeed, charadriiform birds (this study), Pigeons (Fujita, 2002) and

Little Egrets (Fujita, 2003) typically begin thrusting their heads forward over the torso just before or at hind limb touchdown. In only 8% of the trials in the present study did hind limb touchdown occur while the head was still in hold phase.

Finally, the hypothesized visual function of the hold phase is bolstered by the comparison between head-bobbing and non-head-bobbing birds. In the present study, head-bobbing charadriiform species had longer stride and stance durations than non- 78 head-bobbers. This difference may have a phylogenetic basis, since the majority of the non-head-bobbing species are within three closely related families (Laridae, Sternidae,

Rynchopidae) and the majority of the head-bobbing species are within the Family

Scolopacidae (Fig. 4-1). However, the Charadriidae contains both head-bobbing and non-head-bobbing species, and the analysis between the two charadriid species showed the same pattern as was observed across Charadriiformes: the head-bobbing Black-bellied

Plover had longer stride and stance durations than the non-head-bobbing Semipalmated

Plover. If visual stability during the hold phase is important to head-bobbing birds, then longer stance durations would correspond to longer periods of single-limb support and longer hold phases. A parallel situation may be occurring in birds that are facultative head-bobbers, e.g., the longer stride lengths of the Black-headed Gulls when they head- bob may provide a means to maximize the period of visual input when foraging (Fujita,

2006).

It has been suggested that visually-guided ground foraging birds will head-bob because the hold phase of the head-bob allows the birds to better differentiate food items on the ground (Davies and Green, 1988). By this logic, birds that are primarily tactile foragers should not head-bob, nor should birds that primarily forage in flight, dive, surface dip, filter-feed or stalk. The foraging ecology of charadriiform birds differs among species: some are visual foragers that stride, stalk, dive, surface dip or forage in flight, other species are tactile foragers that probe, and some species use both visual and tactile methods (del Hoyo, 1996; Table 4-1). Charadriiform birds that are primarily visual foragers include both head-bobbing and non-head-bobbing species. Generally, the species that primarily forage by walking on the ground and looking for food were head- 79 bobbers, and the species forage by surface dipping, plunge diving, catching prey in flight and skimming the surface of water were non-head-bobbers. An exception to this pattern is the Ring-billed Gulls that primarily forage by walking but do not head-bob.

The only primarily tactile species, Black Skimmers, were non-head-bobbers. Finally, birds that forage by waiting for a prey item to emerge and then running after it consisted of both head-bobbing (Black-bellied Plovers) and non-head-bobbing species

(Semipalmated Plovers and Killdeer). Therefore, although there are some exceptions, the foraging type is well related to whether a bird is a head-bobber or non-head-bobber.

In conclusion, precise synchronization of head-bobbing cycles and footfall patterns was rarely observed in Charadriiform birds, nonetheless, head-bobbing movements did corresponded loosely to footfall patterns. Coordinating the timing of hold phase with single-limb support has clear advantages for visual stability. Yet a degree of variability in timing was observed and is likely due to the conflicting challenges of the locomotor and visual systems. Uneven substrates may necessitate adjustments in the stride cycle. Conversely, a visual interest might inspire the bird to elongate its hold phase. In either case, the single-limb support phase and head-bobbing cycle could be adjusted by the birds in order to optimize horizontal and vertical stability of the visual system. Though some imprecision of head-bobbing and locomotor events occurs (perhaps because of the lag between sensory stimuli and motor responses), it appears that avoiding the potentially jarring effect of hind limb touchdown on visual stability is of paramount importance to the head-bobbing birds. 80 Table 4-1. Numbers of Individuals and Strides, Presence or Absence of Head-bobbing and Foraging Ecology (del Hoyo et al., 1996) in Charadriiform Species. Number of Number of Species Head-bobbing Foraging (vision) Foraging (type) Food type individuals strides Ring-billed Gull 11 39 NHB visual & tactile walking, plunge diving omnivorous

Laughing Gull 13 45 NHB visual surface dipping, walking, in flight omnivorous plunge diving, surface dipping, Royal Tern 6 28 NHB visual & tactile carnivorous skimming Black Skimmer 5 26 NHB tactile skimming carnivorous

Sanderling 21 60 HB visual & tactile walking, probing, pecking omnivorous

Least Sandpiper 20 59 HB visual & tactile walking, probing, pecking carnivorous

Lesser Yellowlegs 9 38 HB visual walking carnivorous

Greater Yellowlegs 8 36 HB visual walking carnivorous

Willet 23 65 HB visual walking carnivorous

Marbled Godwit 7 32 HB visual & tactile walking, probing, pecking carnivorous

Whimbrel 8 31 HB visual & tactile walking, probing, pecking omnivorous

Black-necked Stilt 10 38 HB visual & tactile walking, probing, pecking carnivorous

Black-bellied Plover 17 65 HB visual waiting and running omnivorous

Semipalmated Plover 13 45 NHB visual waiting and running carnivorous

Killdeer 9 31 NHB visual waiting and running omnivorous

HB = head-bobber; NHB = non-head-bobber 81 Table 4-2. Results for the Least Squares Regressions of Log 10 Stance and Stride Duration (sec) Versus Log 10 Speed (body lengths per second). Regression Slope Intercept R2 P Log 10 stance versus log 10 speed All Head-bobbers -0.9414 -0.0151 0.809 <0.001 All Non-head-bobbers -0.7111 -0.2399 0.6285 <0.001

Log 10 stride versus log 10 speed All Head-bobbers -0.7729 0.0778 0.7625 <0.001 All Non-head-bobbers -0.5138 -0.1508 0.5283 <0.001

Log 10 stance versus log 10 speed Head-bobbing Black-bellied Plover -0.8086 -0.1347 0.932 <0.001 Non-head-bobbing Semipalmated Plover -0.9374 -0.1137 0.8683 <0.001

Log 10 stride versus log 10 speed Head-bobbing Black-bellied Plover -0.6792 -0.0086 0.911 <0.001 Non-head-bobbing Semipalmated Plover -0.632 -0.1022 0.8067 <0.001

Ring-billed Gull (NHB) Laughing Gull (NHB) Laridae Sternidae Royal Tern (NHB)

Rynchopidae Black Skimmer (NHB)

Stercorariidae Dromadidae

Alcidae

Glareolidae Sanderling (HB) Jacanidae Least Sandpiper (HB) Lesser Yellowlegs (HB) Rostratulidae Thinocoridae Greater Yellowlegs (HB) Pedionomidae Willet (HB)

Marbled Godwit (HB) Scolopacidae Pluvianellidae Whimbrel (HB)

Chionidae

Burhinidae Haematopodidae Black-necked Stilt (HB)

Recurvirostridae Black-bellied Plover (HB) Charadriidae Semipalmated Plover (NHB)

Killdeer (NHB)

Figure 4-1. A phylogeny of Charadriiformes (on left), adapted from Thomas et al. (2004), to show hypothesized evolutionary relationships between the species within this study (on right). The species are indicated as either being head-bobbers (HB) or non- head-bobbers (NHB). 83

1 A

0.8

0.6

0.4 Relativeholdphase

0.2

0 0 2 4 6 8 10 Speed (body lengths/s)

0.6 0.8 B C y = -0.05x + 0.6086 0.5 0.7 R2 = 0.2339 0.6 0.4 0.5 0.3 0.4 0.3 0.2 Relative hold phase hold Relative

Relative hold phase hold Relative 0.2 0.1 0.1 0 0 0 2 4 6 8 10 0 0.5 1 1.5 2 Speed (body lengths/s) Speed (m/s)

Figure 4-2. A) Relative hold phase as a function of speed for all head-bobbing species. B) Relative hold phase as a function of speed for the Least Sandpiper (Calidris minutilla ). This species had the highest R 2 value for the linear regression of relative hold phase against speed (y = -0.05x + 0.6086; R 2 = 0.2339; P < 0.05). C) Relative hold phase as a function of speed for Elegant-crested Tinamous ( Eudromia elegans , Hancock, chapter 3). 84 1 A 0.8 0.6 Beginning of hold before liftoff 0.4 0.2 0 -0.2 -0.4 beginning of hold phase hold of beginning -0.6 Beginning of hold after liftoff Relative duration from liftoff to the the to liftoff from duration Relative -0.8 -1 0 2 4 6 8 10 Speed (body lengths/s)

0.35 B 0.25

0.15

0.05

-0.05

-0.15

-0.25 the beginning of hold phase hold of beginning the Relative duration from liftoff to to liftoff from duration Relative -0.35

g r s s it el r pe g w ve lin leg illet d er pi w wle o lo d d W imbr ed Stilt P an llo G d an S e Wh ck S Y ne st r r Yello rbled - ea e te a L ss k-bellie e rea M lack c L B la G B Species

Figure 4-3. A) Relative duration from liftoff to the beginning of hold phase plotted against speed. B) Box plots of the relative duration from liftoff to the beginning of hold phase for each species. An ANOVA of the relative duration from liftoff to the beginning of hold phase for each species did not show significant differences among species (F = 1.227; P > 0.2).

1 e A 0.8 Beginning of thrust before touchdown 0.6 0.4 0.2 0 -0.2 -0.4

beginningof thrust phase Beginning of thrust after touchdown -0.6 -0.8 Relative duration Relative fromtouchdown tothe -1 0 2 4 6 8 10 Speed (body lengths/s)

0.35 B

0.25

0.15

0.05

-0.05

-0.15

-0.25 to the beginning of thrust phase thrust of beginning the to

Relative duration from touchdown from duration Relative -0.35

er t er gs gs v ling e e ille dwit wl o mbrel Stilt o W i d andpip llowl ll G h ke ed W c Sander Ye Ye l ne llied plo ast s r r rb - e te a ck Le M a Lesse rea Bl G Black-b Species

Figure 4-4. A) Relative duration from touchdown to the beginning of thrust phase plotted against speed. B) Box plots of the relative duration from touchdown to the beginning of thrust phase for each species. An ANOVA of the relative duration from touchdown to the beginning of thrust phase for each species did show significant differences among species (F = 7.322; P < 0.001).

86 2 A 1.8 1.6 Head-bobbers 1.4 Non-Head-bobbers 1.2 1 0.8 0.6 Stance duration (s) 0.4 0.2 0 0 5 10 15 20 25 Speed (body lengths/s)

0.5 B y = -0.9414x - 0.0151 R2 = 0.809 0

-0.5

y = -0.7111x - 0.2399

Log stance duration stance Log -1 R2 = 0.6285

-1.5 -0.25 0 0.25 0.5 0.75 1 1.25 Log speed

Figure 4-5. A) Stance duration as a function of speed for all head-bobbers versus all non- head-bobbers. B) Log stance duration as a function of log speed for all head-bobbers versus all non-head-bobbers. The gray line is the linear regression line for the head- bobbers and the black line is the regression line for the non-head-bobbers. Regression lines differed significantly (P < 0.001). 87 3 A

2.5 Head-bobbers

2 Non-Head-bobbers

1.5

1 Stride duration (s) duration Stride 0.5

0 0 5 10 15 20 25 Speed (body lengths/s)

0.75 B y = -0.7729x + 0.0778 R2 = 0.7625 0.25

-0.25

Log stride duration stride Log -0.75 y = -0.5138x - 0.1508 R2 = 0.5283 -1.25 -0.25 0 0.25 0.5 0.75 1 1.25 Log speed

Figure 4-6. A) Stride duration as a function of speed for all head-bobbers versus all non- head-bobbers. B) Log stride duration as a function of log speed for all head-bobbers versus all non-head-bobbers. The gray line is the linear regression line for the head- bobbers and the black line is the regression line for the non-head-bobbers. Regression lines differed significantly (P < 0.001). 88

1.2 A

1 Black-bellied plover (HB) Semipalmated plover (NHB) 0.8

0.6

0.4 Stance duration (s) 0.2

0 0 5 10 15 20 Speed (body lengths/s)

-0.5-0.1 0 0.5 1 1.5 B -0.3 y = -0.8086x - 0.1347 -0.5 R2 = 0.932 -0.7

-0.9

-1.1

Log stance duration (s) y = -0.9374x - 0.1137 -1.3 R2 = 0.8683 -1.5 Log speed (body lengths/s)

Figure 4-7. A) Stance duration as a function of speed in a head-bobbing species (Black- bellied Plover) and non-head-bobbing species (Semipalmated Plover) within the same family, Charadriidae. B) Log stance duration as a function of log speed in a head-bobbing species (Black-bellied Plover) and non-head-bobbing species (Semipalmated Plover). The gray line is the linear regression line for the Black-bellied Plover and the black line is the linear regression line for the Semipalmated Plover. Regression lines were significantly different (P < 0.05). 89

1.4 A 1.2 Black-bellied plover (HB)

1 Semipalmated plover (NHB) 0.8

0.6

0.4 Strideduration (s) 0.2

0 0 5 10 15 20 Speed (body lengths/s)

0 -0.5 0 0.5 1 1.5

B -0.2 y = -0.6792x - 0.0086 R2 = 0.911 -0.4

-0.6

Log stride duration (s) -0.8 y = -0.632x - 0.1022 R2 = 0.8067 -1 Log speed (body lengths/s)

Figure 4-8. A) Stride duration as a function of speed in a head-bobbing species (Black- bellied Plover) and non-head-bobbing species (Semipalmated Plover) within the same family, Charadriidae. B) Log stride duration as a function of log speed in a head-bobbing species (Black-bellied Plover) and non-head-bobbing species (Semipalmated Plover). The black line is the linear regression line for the Black-bellied Plover and the gray line is the linear regression line for the Semipalmated Plover. Regression lines were significantly different (P < 0.01). 90

Figure 4-9. The sagittal ( x-axis) coordinates for the eye in one trial of the Least Sandpiper plotted against time. Hold phases correspond to periods of no change in the x- coordinates; thrust phases occur during brief periods of rapid change in position of the eye. The gray and black bars at the top of the plot represent the support phases of the left and right hind limbs, respectively. The gray shaded boxes represent the periods when only one hind limb is in contact with the ground (single-limb support phase). It is evident in this trial, as well as other trials, that the duration of hold phase can change between steps and that they largely correspond to the single-limb support phase. 91 CHAPTER 5: THE EVOLUTION OF RETINAL MORPHOLOGY IN BIRDS

Summary

The morphology of the retina differs among bird species. The area centralis is a region of increased in retinal cell density, and its location and shape vary in birds. It may be circular and located in the nasal or temporal retina, or it may be a horizontal band that extends from the nasal to the temporal aspect of the retina. Also, there may be two circular areas, one in the nasal retina and one in the temporal retina. The fovea is a depression within the retina that may increase visual acuity and the ability to sense motion, and its location and number also vary. In birds, the number of foveae ranges from 0 to 2. Also, the fovea may be located in the nasal, temporal or both retinal fields.

This study collected the known data on avian retinal morphology, mapped the characters onto a phylogeny and reconstructed the evolution of the characters. The ancestral character state for birds was a nasal unifoveate retina with a band-shaped area, and from this nine other retinal patterns evolved. The evolution of these retinal patterns appears complex, involoving retentions of the primitive condition, directional changes toward different configurations and reversals.

Introduction

Photoreceptor distribution across the fundus of the eye varies among birds (Wood,

1917; Walls, 1942; Martin, 1985). A key feature of some species is the localized region of increased retinal cell density known as the area centralis (Fig. 5-1). The area centralis in birds can be circular in shape and located on the nasal aspect of the fundus (area 92 nasalis) or on the temporal aspect (area temporalis). Additionally, some species of birds have both nasal and temporal circular areas. The area centralis may also be in the shape of a band that runs horizontally from the nasal aspect of fundus to the temporal aspect (area centralis horizontalis). Although area centralis is sometimes used interchangeably with area nasalis, in this study “area centralis” or simply “area” will refer to all regions of increased retinal cell density found in birds. In some species, the rods are eliminated entirely within the area and the region is completely composed of cones.

The fovea is a depression in the retina found within the area of most birds (Walls,

1942; Martin, 1985). It is believed to function to locally magnify the image onto the photoreceptor layer (Fig. 5-2; Walls, 1942; Martin, 1985). This occurs because the retina has a higher refractive index than the vitreous humor, so that any light rays that do not pass perpendicular to the retinal surface are refracted radially outward, thereby stimulating a greater number of photoreceptors and, hence, improving visual acuity. The fovea may also allow for fixation of the image because of increased sensitivity to motion in this area (Pumphrey, 1948; Martin, 1985). Foveal depressions range in number across birds from none to two per eye; when present, the fovea can be circular or linear in shape

(Fig. 5-1).

In this paper, “retinal pattern” refers to the configuration of both the area centralis

and fovea. The goals of this study were to assemble the retinal patterns of different

species reported in the literature, map the characters on a composite phylogeny and assess

the evolutionary history of retinal patterns within birds.

93 Materials and Methods

Retinal pattern data were obtained from literature sources (Table 1). Birds display twelve discrete retinal patterns, varying in the presence, shape and location of the area and fovea: no area centralis, afoveate with a circular nasal area, afoveate with a circular temporal area, afoveate with a band-shaped area, unifoveate nasal with a circular area, unifoveate nasal with a band-shaped area (circular and linear fovea), unifoveate temporal with a circular area, unifoveate temporal with a band-shaped area, bifoveate with circular areas and bifoveate with a band-shaped area (circular and linear fovea; Fig.

5-1). In this study, retinae with linear foveae were lumped with the same retinal pattern displaying circular fovea, assuming that linear and circular fovea are functionally similar.

This resulted in ten different character states.

The avian phylogeny used in this study is a supertree constructed from multiple sources in the literature. The higher-level organization of the phylogeny was based on a class-wide analysis (Hackett et al., 2008) that used 19 loci from 169 species that represented the major extant clades of birds. Multiple analyses within avian orders, families and genera were then added to the higher-level phylogeny. If there was more than one analysis of a clade, then analyses were chosen based on the inclusive of taxa for which retinal data was known. In some cases the use of more than one analysis of a clade was necessary in order to incorporate all of the taxa. If an analysis included more than one tree, than the tree that incorporated more data was chosen. If the analysis used multiple methods to generate trees, then the trees based on either Bayesian or maximum likelihood methods were chosen. The majority of analyses used to construct the phylogeny were based on molecular data and used either Bayesian or maximum 94 likelihood analyses (Alström et al., 2006 [Sylvioidea]; Barker et al., 2004

[Passeriformes]; Benz et al., 2006 [Picidae]; de Kloet and de Kloet, 2005

[Psittaciformes]; Ericson et al., 2005 [Corvidae]; Ericson and Johansson, 2003

[Passerida]; Feldman and Omland, 2005 [ Corvus ]; Fuchs et al., 2006 [Passerida];

Grapputo et al., 2001 [Emberizidae]; Johnson and Clayton, 2000 [Columbiformes];

Klicka et al., 2005 [Turdinae]; Moyle, 2006 [Alcedinidae]; Riesing et al., 2003 [Buteo];

Sheldon et al., 2000 [Ardeidae]; Sheldon et al., 2005 [Hirundinidae]; Spicer and

Dunipace, 2004 [Passeriformes]; Yuri and Mindell, 2002 [Fringillidae]). Two analyses

were based on molecular data and used parsimony methods (Oates and Principato, 1994

[Anatidae]; Wink et al., 1996 [Falconiformes]). One analysis was based on molecular,

morphological and behavioral characters and used parsimony methods (Patten and

Fugate, 1998 [Emberizidae]). Two analyses were based on morphological data and used

parsimony methods (Dyke et al., 2003 [Galliformes]; Livezey, 1998 [Gruiformes]).

Finally, one analysis was a supertree (Thomas et al., 2004 [Charadriiformes]).

It is important to note that in the Hackett et al. (2008) phylogeny some orders

were not monophyletic. First, the gruiforms were not monophyletic and, as a result, the

three gruiform species (Kagu, Great Bustard and Brazilian Seriema) within this analysis

were not grouped together. The Kagu instead was placed at the base of a group that

included the Caprimulgiformes and Apodiformes, the Great Bustard was placed at the

base of a group that included the Cuculiformes and Ralliformes and the Brazilian

Seriema was at the bas of the group that include the Passeriformes, Psittaciformes and

Falconidae. Second, the Falconiformes were split into two clades, the first, as stated

above, included the Falconidae, and the second included the Accipitridae and 95 Sagittariidae. Third, Pelecanus conspicillatus was group with the Ciconiiformes

instead of with the other Pelecaniformes. Finally, the Piciformes were nested within the

Coraciiformes.

The out-group for this study was Crocodylia, represented by a single retinal

pattern (nasal unifoveate with a band-shaped area centralis) as reported for both

Crocodylus intermedius and Alligator mississippiensis (Chievitz, 1889). Retinal pattern

characters were mapped onto the phylogeny using Mesquite software (version 2.72;

Maddison and Maddison, 2009). Then, the evolution of the characters was analyzed

using an unordered parsimony method in Mesquite.

Results

From the literature, retinal patterns of 165 species of birds were recorded (Table

1; Fig. 5-3). Nearly 70% of reported retinal patterns in birds were nasal unifoveate, with

either a circular area (66 species) or band-shaped area (37 species). The second most

common patterns were bifoveate, with either a circular area (16 species) or a band-shaped

area (7 species). The least common retinal patterns included: temporal unifoveate with a

circular area (7 species), no area centralis (6 species), afoveate with a circular nasal area

(4 species) and afoveate with a circular temporal area (2 species). Finally, two species,

the Common Swift and Oilbird, exhibited unique retinal patterns: temporal unifoveate

with a band-shaped area and afoveate with a band-shaped area, respectively.

The Mesquite-based character reconstruction of retinal patterns suggests that the

ancestral character state of the Class Aves is a nasal unifoveate retina with a band-shaped

area (Fig. 6-4). This retinal pattern, which matches that found in crocodylians, was 96 retained in most of the Anseriformes and Charadriiformes. Occasional records of nasal unifoveate retinae with band-shaped areas were also scattered across the avian phylogeny

(Great Crested Grebe, Great Bustard, American Coot, Dark-bodied Shearwater,

American Crow, Whinchat, Wheatear, Common Linnet, White Wagtail and Yellow

Wagtail).

A nasal unifoveate retina with a circular area was the most common configuration found among extant birds (~45%). This is the dominant pattern noted for the highly specious Passeriformes. It is also widely distributed across neognathic birds, including multiple species within Galliformes, Ciconiformes, Pelecaniformes and Piciformes.

Finally, all paleognathes had nasal unifoveate retinae with circular areas.

Species with the other retinal patterns containing foveae were both scattered throughout the avian phylogeny and concentrated within individual clades. The majority of species within Falconiformes and Coraciiformes had a bifoveate retina with circular areas. Additionally, species with a bifoveate retina with circular areas were found within the Cariamidae, Hirundinidae (Passeriformes) and Trochilidae (Apodiformes). All terns

(Charadriiformes) exhibited a bifoveate retina with a band-shaped area, as did the

Common Flamingo (Phoenicopteriformes), Red-tailed Hawk (Falconiformes) and

European Chimney Swallow (Hirundinidae). Species with a temporal unifoveate retina with a circular area were limited to the Strigiformes.

Finally, retinae lacking a fovea were found in numerous clades across the neognathic birds. There were two species within Psittaciformes, two within Galliformes, one within Columbiformes and one within Ralliformes that had a retina without an area centralis. Two species within Galliformes, two within Columbiformes and one within 97 Cuculiformes had an afoveate retina with a circular nasal area. Only two species, the

Owl-parrot (Psittaciformes) and Jackass Penguin (Sphenisciformes), had an afoveate retina with a circular temporal area.

Discussion

The ancestral character state of the retina in Class Aves appears to be the nasal

unifoveate retina with a band-shaped area. This is the only retinal pattern reported for the

other extant archosaur clade, Crocodylia (Chievitz, 1889), the outgroup used in this

analysis. The pattern is also found within Lepidosauria (Ross, 2004).

The evolution of the retina from the primitive condition (nasal unifoveate retina

with a band-shaped area) to the nine other configurations appear to have followed a

complicated path in birds (Fig. 5-5), including retentions of the primitive condition,

directional changes toward different configurations and reversals. One major directional

change toward a bifoveate retina with circular areas seems to have occurred within the

common ancestor to the clade within Neoaves (Falconiformes, Strigiformes,

Coraciiformes, Piciformes, Cariamidae, Psittaciformes and Passeriformes). As a result,

the evolution of retinal pattern before this hypothesized transition will be discussed first,

and then the evolution following this transition will be discussed.

Retinal pattern evolution before the transition to a bifoveate retina with circular areas

Two avian clades retained the primitive condition of the retina with little

variation. Most anseriforms (ducks and allies) exhibited a nasal unifoveate retina with a

band-shaped area. Only the Blue Snow Goose modified this pattern; in this species, the 98 area centralis is reduced in size to a circular form concentrated around the nasally- located fovea (nasal unifoveate retina with a circular area). A similar pattern was found among charadriiforms (shorebirds): the majority of the species exhibited a nasal unifoveate retina with a band-shaped area, except the Stone Plover, Great Snipe and

Great Black-backed Gull, which all exhibited a nasal unifoveate retina with a circular area, and the Terns, which appear to have retained the nasal fovea and band-shaped area of the primitive condition, adding a second fovea on the temporal side (bifoveate with a band-shaped area). Outside of these two clades, few other species appear to have retained the primitive condition (Great Crested Grebe, Great Bustard and Dark-bodied

Shearwater). The primitive condition may have also been retained in the American Coot; however, the ancestral character state was unresolved for Fulica, therefore, the reduction

to a nasal unifoveate retina with a circular area may have occurred first followed by a

reversal back to the primitive condition.

The most common way that birds appear to have shifted away from the ancestral

retinal pattern is by reducing the area centralis in size, effectively changing it from band-

shaped to circular, while retaining a nasal fovea – in other words, a nasal unifoveate

retina with a circular area. Some clades seem to have done this extensively. Paleognath

birds universally had this pattern. Most ciconiiforms (storks and allies) exhibit nasal

unifoveate retinae with circular areas, except the American Bittern, which appears to

have added a temporal fovea to its retina (bifoveate with circular areas). Additionally,

most Pelecaniformes exhibit a nasal unifoveate retina with a circular area, except the

Brandt Cormorant, which appears to have reverted back to the primitive condition.

Evolutionary changes in the retina in galliforms (gamefowl) appear to be more 99 complicated: although most species exhibit a nasal unifoveate retina with a circular area, both the Brush Turkey and the Chicken seem to have lost the fovea while retaining a nasal area centralis whereas the Harlequin and California Valley Quails appear to have eliminated all retinal specializations.

Understanding the evolution of the other retinal morphologies in the remaining

neognaths is complicated by the unresolved character states of their ancestors or the

polyphyletic nature of their clade (based largely on the topology developed by Hackett et

al., 2008). The discussion below will address some of these clades individually.

The clade consisting of the Ralliformes, Cuculiformes and Great Bustard is

represented by four different retinal morphologies. The retina of their common ancestor

appears to be a nasal unifoveate retina with a band-shaped area and this seems to have

been retained in the Great Bustard. The common ancestor of the Ralliformes and

Cuculiformes was unresolved and exhibited a nasal unifoveate retina with either a band-

shaped area or circular area. Members of the modern orders may have either retained

these states or modified the retina to less specialized configuration: afoveate retina with a

circular nasal area (European Cuckoo) and a retina without an area centralis (Crested

Coot).

As stated before, the Hackett et al. (2008) typology for the gruiforms is

polyphyletic and the Kagu is displaced to the base of a clade that includes the

Caprimulgiformes and Apodiformes. The retinal morphology of the common ancestor of

this grouping was unresolved with four possible character states: nasal unifoveate retina

with a band-shaped area (avian primitive condition), nasal unifoveate retina with a

circular area (Kagu), bifoveate retina with circular areas (Anna Hummingbird) or 100 afoveate retina with a band-shaped area (Oilbird). Therefore, there is little clarity regarding the evolution of retinal morphology among the members of this clade.

From the preceding discussion and the scattering of afoveate retinae across phylogeny (indicated by grey, yellow, orange and pink bars on Fig. 5-4), it is apparent that there is no clade specificity for the evolution of afoveate retinal patterns. Galliforms seem to have evolved it as many as four independent times, including the afoveate retina with circular nasal areas in the distantly related Brush Turkey and Chicken, as well as once or twice among Quails. Among galliforms, afoveation may have occurred either directly from the avian ancestral condition or from the nasal unifoveate, circular area condition. The same is true for amacular retina in the galliforms; there is no evidence that an intermediate retinal condition of an afoveate area is needed for the evolution of a retina without an area centralis. The pattern observed in ralliforms and cuculiforms echo that found in galliforms: both the afoveate retina of the European Cuckoo and the retina without an area centralis of the Crested Coot may have evolved directly from foveated ancestors. Similarly, the afoveate retinae of sphenisciforms (Jackass Penguin) also may have evolved directly from the avian ancestral condition. Finally, the path by which the other afoveate retinae evolved is less clear: the afoveate retina of the Oilbird may have evolved either from unifoveate or bifoveate ancestor, and the afoveate retinae of the

Victoria Crowned and British Wood Pigeons seem to be derived from an ancestor with a unifoveate retina or a retina without an area centralis, as did the retina without an area centralis of the Wonga Wonga Dove.

The least common foveation patterns observed among the birds before the ancestral transition to a bifoveate retina were those that include a temporal fovea (purple, 101 green, red and brown on Fig. 5-4). The Common Flamingo and the Terns exhibit a bifoveated retina with a band-shaped area, which appears to have evolved by simply adding a temporal fovea to the ancestral avian condition. Similarly, the American Bittern seems to have evolved its bifoveate retina with circular areas by adding a temporal fovea and area to the ancestral condition of the ciconiforms (nasal unifoveate retina with circular area). The evolution of temporal foveae among the apodiform species is unclear: both the Common Swift and the Anna Hummingbirds exhibit temporal foveae, although the former species also retains a nasal fovea, and the evolution of temporal foveation or bifoveation in this clade is unclear because of the unresolved character state.

Retinal pattern evolution after the transition to a bifoveate retina with circular areas

The evolution of a bifoveate retina with circular areas from the primitive

condition occurred, as discussed above, in the common ancestor of the Falconiformes,

Strigiformes, Coraciiformes, Piciformes, Cariamidae, Psittaciformes and Passeriformes,

and may have occurred in the Anna Hummingbird (also discussed above). Most of the

species within the Falconiformes seem to have retained the bifoveate retina with circular

areas, except the Red-tailed Hawk, which exhibit a bifoveate retina with a band-shaped

area. Interestingly, the two species within Falconidae, which was split from the other

Falconiformes in the Hackett et al. (2008) phylogeny, also retained a bifoveate retina

with circular areas. Also, the Brazilian Seriema (Gruiformes) seems to have retained a

bifoveate retina with circular areas.

The common ancestor to the group including the Strigiformes, Coraciiformes and

Piciformes was unresolved and may have been either a bifoveate retina with circular 102 areas or nasal unifoveate retina with a circular area. Thus, the shift to the unique temporal unifoveate retina with a circular area in the Strigiformes may either have been from the bifoveate retina with circular areas or from a nasal unifoveate retina with a circular area. It seems intuitive that the former is more likely than the latter. Also, within this group the common ancestor to the Coraciiformes and Piciformes was unresolved.

All of the Piciformes and two Coraciiformes (Common Hoopoe and Black Hornbill) may have evolved a nasal unifoveate retina with a circular area, but the Kingfishers may have either retained the bifoveate retina with circular areas or underwent a reversal from the nasal unifoveate retina with a circular area back to the bifoveate retina with circular areas.

The common ancestor to the Piciformes and Passeriformes was unresolved and either retained a bifoveate retina with circular areas or evolve a nasal unifoveate retina with a circular area. The common ancestor to the Piciformes was also unresolved and either had a nasal unifoveate retina with a circular area or a retina without an area centralis. Two piciform species exhibit a retina without an area centralis, one had a nasal unifoveate retina with a circular area and one appears to have evolved an afoveate retina with a temporal circular area.

Additionally, the common ancestor to the Passeriformes was unresolved and either retained a bifoveate retina with circular areas or evolve a nasal unifoveate retina with a circular area. The common ancestor to the Corvidae and Laniidae was similarly unresolved. One species within Laniidae exhibits a bifoveate retina with circular areas, whereas the majority of the corvids exhibit a nasal unifoveate retina with a circular area.

The common ancestor to the remaining passeriforms may have exhibited a nasal 103 unifoveate retina with a circular area. A reversal to a bifoveate retina with circular areas occurred once in the ancestor to the Hirundininiae and one species within this family seem to have evolved a bifoveate retina with a band-shaped area. Finally, a few species of Passeriformes (American Crow, Whinchat, Wheater, Corn Bunting, Common

Linnet, Yellow Wagtail and White Wagtail) seem to have reverted to the ancestral character state of birds, a nasal unifoveate retina with a band-shaped area, from a nasal unifoveate retina with a circular area.

Band-shaped areas were associated with a variety of foveation patterns. The band-shaped area was associated mainly with shorebirds (Charadriiformes; nasal unifoveate in the majority of species and bifoveate in the Terns) and waterbirds

(Anseriformes; nasal unifoveate) that live in habitats that include large bodies of water

(sea, lakes), but was also associated with birds the forage in flight, such as the Common

Swift (temporal unifoveate, Apodiformes; del Hoyo et al., 1999), Red-tailed Hawk

(bifoveate, Falconiformes; del Hoyo et al., 1994), and European Chimney Swallow

(bifoveate, Passeriformes; del Hoyo et al., 2004). An afoveate retina with a band-shaped area was found in the Oilbird (Caprimulgiformes), which is a nocturnal forager and also has the highest density of rods (1,000,000 per mm 2) recorded in any vertebrate (Martin et

al., 2004). There are four hypothesized functions of the band-shaped area in birds: 1.

horizontal orientation of the head in flight (Pennycuick, 1960), 2. detection of motion

(Pumphrey, 1961), 3. a panoramic view without saccadic eye movements (Walls, 1942)

and 4. spatial orientation (Duijm, 1958). One could argue that these functions would be

beneficial to most bird species, yet the majority of birds did not have a band-shaped area.

Also, the majority of the species within the two other clades of waterbirds, 104 Pelecaniformes and Ciconiformes did not have a band-shaped area. Similarly, the majority of birds that forage in flight did not have band-shaped areas, although, some did have bifoveate retinae with circular areas (Falconiformes, Hirundininae and

Apodiformes). Additionally, the two Crocodylian species used as the outgroup are aquatic animals that forage at the air-water interface (Kröger and Katzir, 2008). In fishes, the band-shaped area is associated with species that are predominantly found at the air- water or sand-water interface (Munk, 1970). An afoveate retina with a band-shaped area has also been found in 4 amphibian species (Slonaker, 1897; American Toad [ Bufo americanus ], Leopard Frog [ Rana virescens ], Edible Frog [ Pelophylax esculentus ] and

American Bullfrog [ Rana catesbeiana ]), which have an aquatic larval stage and, as adults, breed in aquatic habitats. In Lepidosauria, the semiaquatic Nile Monitor ( Varanus niloticus ; Ross, 2004) has a foveate (the location of the fovea was not reported) retina

with a band-shaped area. However, an afoveate retina with a band-shaped area was also

found in terrestrial mammals (Slonaker, 1897; Ungulata: Cattle [ Bos primigenius ],

Bactrian Camel [Camelus bactrianus ], Pig [ Sus scrofa domestica ] and Horse [ Equus

ferus caballus ]; Rodentia: Eastern Cottontail [Sylvilagus floridanus ], European Hare

[Lepus europaeus ] and European Rabbit [ Oryctolagus cuniculus ]; Carnivora: Red Fox

[Vulpes vulpes ]), and a foveate (the location of the fovea was not reported) retina with a band-shaped area was found in terrestrial lepidosaurs (Ross, 2004; Spiny-tailed Agama

[Uromastyx acanthinurus], Shingleback (or Bobtail) Skink [Tiliqua rugosa ], Tupinambis sp. , Desert Monitor [ Varanus griseus ]. Thus, band-shaped areas are found in aquatic, flying and terrestrial vertebrates. 105 The afoveate pattern may be surprising, given that foveation is considered to be a mechanism for enhancing visual acuity. Yet, there are afoveate species across vertebrates. In fact, the only mammal species that have fovea are found within Primates

(Slonaker, 1897; Walls, 1942; Ross, 2004). Also, the only fish species recorded to have fovea are Teleosts (Slonaker, 1897; Walls, 1942; Ross, 2004). In Lepidosauria, afoveate retinae are found within Pygopodidae, Gekkonidae, Helodermatidae, Boidae, Colubridae,

Viperidae and Pythonidae. Also, afoveate retinae are found throughout Testudines, and the only know turtle to have a fovea is within Amyda (Walls, 1942). Although Walls

(1942) includes the area centralis and fovea as adaptations to diurnal activity, many nocturnal species, as well as some deep-sea fishes, exhibit both specializations (Ross,

2004). Also, the lack of a fovea cannot be strictly linked to noctural species. Indeed, most of the afoveate bird species are not nocturnal.

The bifoveate retina has been reported only in birds and Anolis lizards (Ross,

2004). Temporal fovea visualize the binocular field and nasal fovea visualize the monocular fields (Pettigrew, 1986). The bifoveate retina in birds is commonly seen in birds that forage in flight, such as the Falconiiformes, Apodiformes and Hirundininae (del

Hoyo et al., 1994, 1999 and 2004), and also in birds that forage by plunge diving, such as the Terns and Kingfishers (del Hoyo et al., 1996 and 2001). However, these functions are not used by Anolis lizards.

In conclusion, the evolution of avian retinal patterns from the hypothesized ancestral nasal unifoveate retina with a band-shaped area appears complex. Many branches within the character reconstruction remain unresolved as the number of species represented in the analysis was less than 2% of all bird species [also, the Hackett et al. 106 (2008) molecular phylogeny does differ from other phylogenies constructed using morphological characters]. Future directions include: 1. additional histological analysis of bird eyes to incorporate more species, 2. the comparison of retinal patterns to other aspects of avian ecology and behavior, such as foraging method, diel periodicity, habitat complexity and predator avoidance, 3. the potential incorporation of an avian phylogeny constructed using both molecular and morphological characters and 4. an extension of this analysis to all vertebrates.

107 Table 5-1: The Retinal Patterns in Birds. Order

Family

Genus species (Common name) Retinal pattern Location (fovea and/or area centralis) Source

Anseriformes Anatidae Chen caerulescens (Blue Snow Goose) U N 1 Chen hyperboreus (Lesser Snow Goose) UB LF N 1 Anas bochas domesticus or Anas platyrhynchos UB N 2; 3 (Mallard) Oidemia deglandi or Melanitta fusca UB N 3 (White-winged Scoter) Fuligula glacialis or Clangula hyemalis UB N 2 (Long-tailed Duck) Apodiformes Apodidae Cypselus apus or Apus apus (Common Swift) UB T 2 Trochilidae Calypte annae (Anna Hummingbird) B N, T 1 Caprimulgiformes Steatornithidae Steatornis caripensis (Oilbird) A band area 4 Casuariiformes Casuariidae Casuarius occipitalis (Westerman Cassowary) U N 1 Charadriiformes Recurvirostridae Recurvirostra avocetta (Pied Avocet) UB N 2 Laridae Sterna hirundo (Common Tern) BB N, T (above band) 1; 3 Sterna macrura or Sterna paradisaea (Artic Tern) BB N, T 2 Sterna minuta or Sterna antillarum (Least Tern) BB N, T 2 Sterna Cantiaca or Sterna sandviciensis BB N, T 2 (Sandwich Tern) Larus argentatus (Herring Gull) UB LF 1 Larus marinus (Great black-backed Gull) U N 1 Larus canus (Common Gull) UB N 2 108 Larus ridibundus (Black-headed Gull) UB N 2

Table 5-1: Continued. Burhinidae Oedicnemus scolopax or Burhinus oedicnemus U N 1 (Stone Plover or Stone Curlew) Charadriidae Squatarola squatarola or Pluvialis squatarola UB N 1; 2 (Black-bellied Plover) Charadrius hiaticula (Common Ringed Plover) UB N 2 Charadrius pluvialis or Pluvialis dominica UB N 2 (American Golden Plover) Aegialitis semipalmata or Charadrius semipalmatus UB N 3 (Semipalmated Plover) Vanellus cristatus or Vanellus vanellus UB N 2 (Northern Lapwing) Scolopacidae Totanus melanoleucus or Tringa melanoleuca UB N 1 (Greater Yellowlegs) Strepsilas interpres or Arenaria interpres UB N 2 (Ruddy Turnstone) Numenius hudsonicus or Numenius phaeopus UB N 1 (Hudsonian Curlew or Whimbrel) Gallinago media (Great Snipe) U N 2 Tringa Islandica or Calidris canutus (Red Knot) UB N 2 Tringa alpina or Calidris alpina (Dunlin) UB N 2 Ereunetes pusillus or Calidris pusilla UB N 3 (Semipalmated Sandpiper) Limosa Lapponica (Bar-tailed Godwit) UB N 2 Totanus glareola or Tringa glareola UB N 2 (Wood Sandpiper) Totanus hypoleucus or Tringa hypoleucus UB N 2 (Common Sandpiper) Numenius arquata (Eurasian Curlew) UB N 2 Alcidae Alca torda (Razorbill) UB N 2 Uria troile or Uria aalge (Foolish Guillemot) UB N 2 Cepphus columba (Pigeon Guillemot) UB N 1 Fratercula mormon or Fratercula artica (Puffin) UB N 2 Haematopodidae Haematopus ostralegus (Eurasian Oystercatcher) UB N 2 Ciconiiformes Ardeidae 109 Ardea cinerea (Grey Heron) U N 2

Table 5-1: Continued. Nycticorax nycticorax (European Night Heron) U N 1; 3 Botaurus lentiginosus (American Bittern) B N, T 1 Botaurus stellaris (European Bittern) U N 1 Cancroma cochlearia (Boat-billed Night Heron) U N 1 Threskiornithidae Plegadis falcinellus (Glossy Ibis) U N 1 Platalea leucorodia (Spoonbill) U N 1 Ciconidae Mycteria americana (Wood Stork) U N 1 Columbiformes Columbidae Leucosarcia picata (Wonga Wonga Dove) None - 1 Columba palumbus (British Wood Pigeon) A N 1 Columba livia domesticus (Pigeon) U N 3 Goura victoria (Victoria Crowned Pigeon) A N 1 Coraciiformes Alcedinidae Alcedo ispida (British Kingfisher) B N, T 1 Dacelo gigas (Laughing Kingfisher) B N, T 1 Ceryle alcyon (Belted Kingfisher) B N, T 3 Upupidae Upupa epops (Common hoopoe) U N 1 Bucerotidae Spagolobus adratus (Black Hornbill) U N 1 Cuculiformes Cuculidae Coccyzus americanus (Yellow-billed Cuckoo) U N 1 Cuculus canorus (European Cuckoo) A N 1 Dinornithiformes Apterygidae Apteryx mantelli (Mantell Apteryx) U N 1 Falconiformes Accipitridae Buteo vulgaris or Buteo buteo (Common Buzzard) B N, T 2 Buteo latissimus (Broad-winged Hawk) B N, T 1 Buteo borealis or Buteo jamaicensis BB N, T 1 (Red-tailed Hawk) 110 Gypaetus barbatus (Bearded Vulture) B N, T 1 Haliaetus leucocephalus (Bald Eagle) B N, T 1

Table 5-1: Continued. Haliaetus leucogaster (White-bellied Sea Eagle) B N, T 1 Sagittariidae Gypogeranus serpentarius or Sagittarius serpentarius B N, T 1 (Secretary Bird) Falconidae Falco sparverius (Sparrow Hawk or American Kestrel) B N, T 1; 3 Tinnunculus alaudarius or Falco tinnunculus B N, T 1 (European Kestrel) Galliformes Odontophoridae Lophortyx californicus vallicola or None - 1 Callipepla californica (California Valley Quail) Colinus virginianus (Northern Bobwhite) U N 3 Phasianidae Gallus domesticus (Chicken) A N 2; 3 Meleagris gallopavo (Turkey) U N 2; 3 Cortunix histrionica (Harlequin Quail) None - 1 Perdix cinerea or Perdix perdix (Grey Partridge) U N 2 Phasianus colchicus (Common Pheasant) U N 2 Tetraonidae Bonasa umbellus (Ruffed Grouse) U N 3 Numididae Numida pucherani or Guttera pucherani U N 3 (Crested Guineafowl) Cracidae Crax globosa (Yarrell Curassow) U N 1 Megapodidae Catheturus lathami or Alectura lathami A N 1 (Brush Turkey) Gruiformes Otididae Otis tarda (Great Bustard) UB N 1 Rhynocetidae Rhinochetus jubatus (Kagu) U N 1 Cariamidae Cariama cristata (Brazilian Seriema) B N, T 1 Passeriformes 111 Alaudidae Alauda arvensis (Sky Lark) U N 2 Anthus pratensis (European Titlark) U N 1

Table 5-1: Continued. Certhiidae Certhia familiaris (Eurasian Tree-creeper) U N 3 Corvidae Garrulus glandarius (Eurasian Jay) U N 2 Corvus frugilegus (Rook) U N 2 Corvus corax (Raven) U N 1 Corvus americanus (American Crow) UB N 1; 3 Cyanocitta cristata (Blue Jay) U N 1; 3 Cyanocitta stelleri (Stellar Jay) U N 1 Emberizidae Emberiza citrinella (Yellowhammer) U N 2 Emberiza miliaria or Miliaria calandra UB N 2 (Corn Bunting) Cyanospiza versicolor (Varied Bunting) U N 1 Poocaetes gramineus (Vesper Sparrow) U N 3 Spizella pusilla (Field Sparrow) U N 3 Junco hyemalis (Dark-eyed Junco) U N 3 Melospiza fasciata or Melospiza melodia U N 3 (Song Sparrow) Passerella iliaca (Fox Sparrow) U N 3 Passerina cyanea (Indigo Bunting) U N 3 Fringillidae Fringilla coelebs (Chaffinch) U N 2 Fringilla canaria or Serinus canaria (Island Canary) U N 2 Fringilla cannabina or Carduelis cannabina UB N 2 (Common Linnet) Spinus tristis or Carduelis tristis U N 3 (American Goldfinch) Hirundininae Tachycineta bicolor (White-bellied Swallow) B N, T 1; 3 Hirundo rustica (European Chimney Swallow) BB N, T 2 Hirundo urbica or Delichon urbica B N, T 2 (Northern House-martin) Icteridae Agelaius phoeniceus (Red-winged Blackbird) U N 3 Laniidae 112 Lanius ludovicianus (Loggerhead Shrike) B N, T 1 Mimidae Mimus polyglottus leucopterus (Western Mockingbird) B N, T 1

Table 5-1: Continued. Motacillidae Motacilla alba (White Wagtail) UB N 2 Motacilla flava (Yellow Wagtail) UB N 2 Paridae Parus caeruleus or Cyanites caeruleus (Blue Tit) U N 2 Parus major (Great Tit) U N 2 Parus atricapillus (Black-capped Chickadee) U N 1; 3 Parulidae Dendroica palmarum (Palm Warbler) U N 3 Seiurus aurocapillus (Ovenbird) U N 3 Geothlypis philadelphia or Oporomis philadelphia U N 3 (Mourning Warbler) Sylvania mitrata or Wilsonia citrina U N 3 (Hooded Warbler) Passeridae Passer domesticus (House Sparrow) U N 1; 3 Fringilla montanan or Passer montanus U N 2 (Eurasian Tree Sparrow) Prunellidae Accentor modularis (Dunnock) U N 2 Sylviidae Sylvia cinerea or Sylvia communis U N 2 (Common Whitethroat) Sylvia hortensis (Orphean Warbler) U N 2 Regulus satrapa (Golden-crowned Kinglet) U N 3 Sturnidae Sturnus vulgaris (Common Starling) U N 2 Turdidae Merula migratoria or Turdus migratorius U N 3 (American Robin) Saxicola rubethra (Whinchat) UB N 2 Aedon megaryncha (English Nightingale) U N 1 Sialia sialis or Pluvialis squatarola B; U N, T; N 1; 3 (Eastern Bluebird) Saxicola oenanthe or Oenanthe oenanthe (Wheatear) UB N 2 113 Tyrannidae Pitangus derbianus (Derby Tyrant) U N 1 Tyrannus tyrannus (Eastern Kingbird) U N 3

Table 5-1: Continued. Pelecaniformes Phalacrocoracidae Phalacrocorax carbo (Great Cormorant) U N 1 Phalacrocorax penicillatus (Brandt Cormorant) UB N 1 Sulidae Sula bassana or Morus bassana (Gannet) U N 1 Pelecanidae Pelecanus conspicillatus (Australian Pelican) U N 1 Phoenicopteriformes Phoenicopteridae Phoenicopterus roseus (Common Flamingo) BB LF N, LF T 1 Piciformes Picidae Colaptus mexicanus (Red-shafted Flicker) U N 1 Colaptus auratus (Northern Flicker) U N 3 Melanerpes erythrocephalus U N 1 (Red-headed Woodpecker) Centurus uropygialis (Gila Woodpecker) U N 1 Dryobates vel Dendrocopus major or U N 1; 2 Dendrocops major (Great Spotted Woodpecker) Podicipediformes Podicipedidae Podiceps cristatus (Great Crested Grebe) UB N 1 Procellariiformes Procellariidae Puffinus griseus (Dark-bodied Shearwater) UB LF 1 Psittaciformes Cacatuidae Cacatua galerita (Great Sulphur-crested Cockatoo) none - 1 Psittacidae Chrysotis amazona or A mazona amazonica U N 1 (Orange-winged Parrot) Conurus ridgway or Aratinga euops None - 1 (White-fronted Cuban Conure) Stringops habroptilus (Kakapo or Owl-parrot) A T 1 114 Ralliformes Rallidae Fulica americana (American Coot) UB N 1 Fulica cristata (Crested Coot) None - 1 Aramides ypecaha (Ypecha Rail) U N 1

Table 5-1: Continued. Rheiformes Rheidae Rhea americana (Rhea) U N 1 Sphenisciformes Spheniscidae Spheniscus demersus (Jackass Penguin) A T 1 Strigiformes Strigidae Strix vel Megascops asio or Otus asio U T 1; 3 (American Screech Owl) Bubo virginianus (Great Horned Owl) U T 1 Syrnium nebulosum or Strix varia (Barred Owl) U T 1; 3 Strix flammea or Tyto alba (European Barn Owl) U T 1 Strix otus or Asio otus (Long-eared Owl) U T 2 Syrnium aluco or Strix aluco (Tawny Owl) U T 1 Speotyto cunicularia hypogaea (Burrowing Owl) U T 1 Struthioniformes Struthionidae Struthio camelus (Ostrich) U N 1 Tinamiformes Tinamidae Rhyncotus rufescens (Rufous Tinamou) U N 1 Calodroma elegans or Eudromia elegans U N 1 (Martineta Tinamou or Elegant-crested Tinamou)

Retinal pattern: None, no area centralis; A, afoveate; U, unifoveate; B, bifoveate; UB, unifoveate with band-shaped area; BB, bifoveate with band-shaped area.

Location: N, nasal; T, temporal; LF, linear fovea.

References: 1, Wood, 1917; 2, Chievitz, 1891; 3, Slonaker, 1897; 4, Pettigrew and Konishi, 1984. 115

Figure 5-1. Schematics of the fundus oculi of the right eye of birds showing twelve discrete retinal patterns known: black oval – pecten; grey shading – area centralis (circular or band-shaped); black dot – circular fovea; black line – linear fovea. The boxes indicate retinal configurations that were merged for this analysis, resulting in ten different retinal patterns. 116

Photoreceptor layer

Retina

Vitreous humor

Light rays

Figure 5-2. The foveal depression within the retina creates a curved surface (black line). As light rays pass through the retina they are refracted radially due to the higher refractive index of the retina relative to the vitreous humor. As a result, images are magnified on the photoreceptor layer. 117

0.7% 0.7% 4.1% 4.7% 2.7% 11.5% 1.4%

4.7%

44.6% 25.0%

No area centralis Afoveate, circular nasal area Afoveate, circular temporal area Nasal unifoveate, circular area Nasal unifoveate, band-shaped area Temporal unifoveate, circular area Bifoveate, circular areas Bifoveate, band-shaped area Temporal unifoveate, band-shaped area Afoveate, band-shaped area

Figure 5-3. Pie graph of the retinal patterns found within birds. 118

Figure 5-4 . Character reconstruction of retinal patterns. 119

Figure 5-4. Continued. 120

Figure 5-5. Schematics of the fundus oculi of birds showing the evolution of retinal pattern: black oval – pecten; grey shading – area centralis; black dot – fovea. Solid arrows represent resolved events and dashed arrows represent unresolved events. Numbers represent the frequency of an event. 121 CHAPTER 6: HEAD-BOBBING, RETINAL MORPHOLOGY

AND FORAGING IN BIRDS

Summary

Head-bobbing is a behavior in some birds that enhances vision. The hold phase of head-bobbing is thought to aid in the vision of moving objects by stabilizing the head relative to the environment while the body continues to move forward. The thrust phase may allow for the detection of stationary objects and relative depth perception via motion parallax as the head is accelerated in front of the body. As a result, head-bobbing may aid active visual foragers. It has also been hypothesized that birds with a unifoveate retina will head-bob, whereas those without fovea or two fovea will not head-bob. This study collected data on head-bobbing, retinal morphology and visual versus tactile foraging in birds. The characters were mapped onto a phylogeny and the evolution of each was reconstructed. Relationships between the characters were also analyzed using pairwise comparisons. The ancestral character states for birds appear to be non-head- bobbing, a nasal unifoveate retina with a band-shaped area and visual foraging. None of the pairwise comparisons showed significant relationships among characters, however, some trends were apparent both within clades and across the phylogeny. One of the most surprising results was that head-bobbing was used by visual and tactile foragers alike, suggesting that head-bobbing not only aids foraging but also enhances the detection of other important visual stimuli, such as predators. Also, head-bobbing behaviors were not limited to birds with a specific retinal morphology, as it occurred in afoveate, unifoveate and bifoveate birds. 122

Introduction

Foraging behaviors represent a considerable effort in birds (up to 83% of daily activity patterns in the Rufous-sided Towhee; Greenlaw, 1969), and numerous feeding guilds have been defined based on diet and foraging mode. Whereas the senses of olfaction and hearing may be employed to identify and obtain food items (Ntiamoa-Baidu et al., 1998), vision and touch are the primary sensory mechanisms used by birds

(Schneider, 1983; McNeil et al., 1992). Visual foragers rely on direct observation of food items or identification of their presence on the substrate surface whereas tactile foragers probe into the substrate with their beaks in order capture food items or sense the vibrations of prey items through a substrate (Barbosa and Moreno, 1999; Rojas et al.,

1999).

Interspecific differences of visual field topography in birds have been related to foraging mode (Martin and Katzir, 1999), and it has been suggested that visual foragers may have evolved specializations to enhance detection of food items. Photoreceptor distribution across the fundus of the eye may offer greater insights into visual acuity (the resolution of a visual image). A key feature of the retina is the localized region of increased cell density known as the “area centralis”. The area centralis in birds can be circular in shape and located on the nasal aspect of the fundus (area nasalis or area centralis) or on the temporal aspect (area temporalis; Wood, 1917; Walls, 1942; Martin,

1985). This region of high photoreceptor density may also be in the shape of a narrow band that runs horizontally from the nasal aspect of fundus to the temporal aspect (area centralis horizontalis). In some birds, the rods are eliminated entirely within the area and 123 the region is completely composed of cones. The fovea is a depression in the retina typically found within the area centralis that is believed to function to locally magnify the image onto the photoreceptor layer (Fig. 6-1; Walls, 1942; Martin, 1985). This is because the retina has a higher refractive index than the vitreous humor, so that any light rays that do not pass perpendicular to the retinal surface are refracted radially outward, thereby stimulating a greater number of photoreceptors and, hence, improving visual acuity. The fovea may also be more sensitive to motion, which would enhance the fixation of the fovea on a moving object (Pumphrey, 1948; Martin, 1985). Foveal depressions range in number across birds from none to two per eye. In this paper,

“retinal pattern” refers to the configuration of both the area centralis and fovea.

A related and possibly confounding feature of the avian visual apparatus is head- bobbing, the fore-aft movement of the head during terrestrial locomotion found in some birds. Contrary to previous suggestions (Daanje, 1951; Dagg, 1977), head bobbing is not mechanically link to limb movements during locomotion (see chapters 3 and 4). Rather, it appears to be an optomotor response to the drift of images across the retina (Friedman,

1975; Frost, 1978). The movement has two distinct phases: a thrust phase and a hold phase (Dunlap and Mowrer, 1930). During the thrust phase, the velocity of the head is greater than the velocity of the body so that the head is translated to a point in front of the body. During the hold phase, the head is immobile (remains fixed in space) as the body travels forward, creating an illusive backward movement of the head. The thrust phase is thought to use motion parallax, the reality that objects closer to the viewer appear to move faster than objects that are farther away, to generate depth perception (Frost, 1978) and also allows for the differentiation of stationary items on a background (Davies and 124 Green, 1988), whereas the hold phase functions to stabilize a moving image on the retina (Dunlap and Mowrer, 1930).

Head-bobbing is not ubiquitous among birds. Its presence in some birds and not in others has led some researchers to seek correlated features that may explain its distribution across birds. One anatomically-based hypothesis links head-bobbing with retinal pattern and suggests that birds exhibiting a single fovea will head-bob in order to keep the image stable on the fovea whereas those lacking a fovea or having two fovea will not head-bob (Whiteside, 1967). By this hypothesis, birds with eyes that face more laterally should head-bob to stabilize the visual image in the lateral visual field with their nasal foveae, however, bifoveate birds can potentially stabilize an image in the binocular field with their temporal foveae and should not need to head-bob (Whiteside, 1967).

Alternatively, an ecological hypothesis proposes that head-bobbing is found in visually- guided ground foragers because the hold phase allows better visual differentiation of food items on the ground (Davies and Green, 1988).

The goals of this research are (1) assess the evolution of retinal morphology, foraging behavior and head movement using the modern avian phylogeny and (2) to evaluate the inter-relationship between retinal morphology, foraging behavior and head movement within birds.

Materials and Methods

Characters for head movement, retinal morphology and foraging behavior were obtained from literature sources. Characters for the foraging category included visual foraging, tactile foraging, both visual and tactile foraging (hereafter signified as 125 “visual/tactile foraging”), nocturnal foraging, filter-feeding and scavenging.

Characters for the head movement category were head-bobbing, non-head-bobbing, rare head-bobbing, head-bobbing while swimming, head-bobbing while foraging, other head movements and body-bobbing (hopping bird; Jimenez-Ortega, 2005). Characters for retinal morphology included no area centralis, afoveate with a circular nasal area,

afoveate with a circular temporal area, afoveate with a band area, unifoveate nasal with a

circular area, unifoveate nasal with a band-shaped area (circular and linear fovea),

unifoveate temporal with a circular area, unifoveate temporal with a band-shaped area,

bifoveate with circular areas and bifoveate with a band area (circular and linear fovea;

Fig. 6-2).

Foraging characters were obtained for all of the species in which retinal pattern

was available in the literature, however, head movement data was obtainable for only a

subset of the species in which retinal pattern data was available; for others head

movements of a closely related species was used as a proxy. Head movement characters

were then classified as either data for the individual species or another species within the

same genus, family or order (Table 6-1).

Retinal pattern data were obtained from literature sources (Table 1). Birds

display twelve discrete retinal patterns, varying in the presence, shape and location of the

area and fovea: no area centralis, afoveate with a circular nasal area, afoveate with a

circular temporal area, afoveate with a band-shaped area, unifoveate nasal with a circular

area, unifoveate nasal with a band-shaped area (circular and linear fovea), unifoveate

temporal with a circular area, unifoveate temporal with a band-shaped area, bifoveate

with circular areas and bifoveate with a band-shaped area (circular and linear fovea; Fig. 126 5-1). In this study, retinae with linear foveae were lumped with the same retinal pattern displaying circular fovea, assuming that linear and circular fovea are functionally similar. This resulted in ten different character states.

The avian phylogeny used in this study is a supertree constructed from multiple sources in the literature. The higher-level organization of the phylogeny was based on a class-wide analysis (Hackett et al., 2008) that used 19 loci from 169 species that represented the major extant clades of birds. Multiple analyses within avian orders, families and genera were then added to the higher-level phylogeny. If there was more than one analysis of a clade, then analyses were chosen based on the inclusive of taxa for which retinal data was known. In some cases the use of more than one analysis of a clade was necessary in order to incorporate all of the taxa. If an analysis included more than one tree, than the tree that incorporated more data was chosen. If the analysis used multiple methods to generate trees, then the trees based on either Bayesian or maximum likelihood methods were chosen. The majority of analyses used to construst the phylogeny were based on molecular data and used either Bayesian or maximum likelihood analyses (Alström et al., 2006 [Sylvioidea]; Barker et al., 2004

[Passeriformes]; Benz et al., 2006 [Picidae]; de Kloet and de Kloet, 2005

[Psittaciformes]; Ericson et al., 2005 [Corvidae]; Ericson and Johansson, 2003

[Passerida]; Feldman and Omland, 2005 [ Corvus ]; Fuchs et al., 2006 [Passerida];

Grapputo et al., 2001 [Emberizidae]; Johnson and Clayton, 2000 [Columbiformes];

Klicka et al., 2005 [Turdinae]; Moyle, 2006 [Alcedinidae]; Riesing et al., 2003 [Buteo];

Sheldon et al., 2000 [Ardeidae]; Sheldon et al., 2005 [Hirundinidae]; Spicer and

Dunipace, 2004 [Passeriformes]; Yuri and Mindell, 2002 [Fringillidae]). Two analyses 127 were based on molecular data and used parsimony methods (Oates and Principato,

1994 [Anatidae]; Wink et al., 1996 [Falconiformes]). One analysis was based on molecular, morphological and behavioral characters and used parsimony methods (Patten and Fugate, 1998 [Emberizidae]). Two analyses were based on morphological data and used parsimony methods (Dyke et al., 2003 [Galliformes]; Livezey, 1998 [Gruiformes]).

Finally, one analysis was a supertree (Thomas et al., 2004 [Charadriiformes]).

It is important to note that in the Hackett et al. (2008) phylogeny some orders were not monophyletic. First, the gruiforms were not monophyletic and, as a result, the three gruiform species (Kagu, Great Bustard and Brazilian Seriema) within this analysis were not grouped together. The Kagu instead was placed at the base of a group that included the Caprimulgiformes and Apodiformes, the Great Bustard was placed at the base of a group that included the Cuculiformes and Ralliformes and the Brazilian

Seriema was at the bas of the group that include the Passeriformes, Psittaciformes and

Falconidae. Second, the Falconiformes were split into two clades, the first, as stated above, included the Falconidae, and the second included the Accipitridae and

Sagittariidae. Third, Pelecanus conspicillatus was group with the Ciconiiformes instead

of with the other Pelecaniformes. Finally, the Piciformes were nested within the

Coraciiformes.

The out-group for this study was Crocodylia, represented by a single retinal

pattern (nasal unifoveate with a band-shaped area centralis) as reported for both

Crocodylus intermedius and Alligator mississippiensis (Chievitz, 1889). Retinal pattern

characters were mapped onto the phylogeny using Mesquite software (version 2.72; 128 Maddison and Maddison, 2009). Then, the evolution of the characters was analyzed using an unordered parsimony method in Mesquite.

The characters for each category were mapped onto a phylogeny and the maximum number of comparisons was performed using Mesquite software (version 2.72,

Maddison and Maddison, 2009). Then, the evolution of the characters for each category was analyzed using an unordered parsimony method in Mesquite. The retinal character evolution followed the analysis in Chapter 5, where Crocodylia was used as an out-group and it was represented by a single retinal pattern (nasal unifoveate with a band-shaped area centralis) as reported for both Crocodylus intermedius and Alligator mississippiensis

(Chievitz, 1889). The correlation of the retinal, head-bobbing and foraging characters was analyzed by pairwise comparison on the phylogeny of the species used in this study.

Results

Retinal Morphology

A more detailed description of retinal pattern evolution in birds is found in

Chapter 5. Briefly, retinal patterns of 165 species of birds were recorded from the literature (Table 6-1; Fig. 6-2 and 6-3A). Nearly 70% of reported retinal patterns in birds are nasal unifoveate, in either a circular area (44.6%) or band-shaped area (25%). The second most common patterns were bifoveate (16.2%), with 11.5% in a circular area and

4.7% in a band-shaped area. The least common retinal patterns include: temporal unifoveate with a circular area (4.7%), no area centralis (4.1%), afoveate with a circular nasal area (2.7%) and afoveate with a circular temporal area (1.4%). Although two 129 additional retinal patterns were reported (temporal unifoveate with a band-shaped area in the Common Swift and afoveate with a band-shaped area in the Oilbird), these were not included in the analyses because there were no data for the head movements of the two species or their genera, families or orders.

The Mequite-based character reconstruction of retinal patterns suggests that the ancestral character state of the Class Aves is a nasal unifoveate retina with a band-shaped area (Fig. 6-4). Nasal unifoveate retinae with band-shaped areas were primary found among the Anseriformes and Charadriiformes. However, all paleognathes had a nasal unifoveate retina with a circular area. Also, the nasal unifoveate retina with a circular area was found widely among neognathic birds, including multiple species within

Galliformes, Ciconiformes, Pelecaniformes, Piciformes and Passeriformes. Species with a bifoveate retina with circular areas were found among the Falconiformes (including the

Falconidae), Coraciiformes (specifically the kingfishers), Cariamidae, Hirundinidae

(Passeriformes) and Trochilidae (Apodiformes). Bifoveate retinae with a band-shaped area were found in the Common Flamingo (Phoenicopteriformes), Red-tailed Hawk

(Falconiformes), European Chimney Swallow (Hirundinidae) and all of the terns

(Charadriiformes). Only species within Strigiformes had a temporal unifoveate retina with a circular area. Retinae without an area centralis were found in two species within

Psittaciformes, two within Galliformes, one within Columbiformes and one within

Ralliformes. Two species within Galliformes, two within Columbiformes and one within

Cuculiformes had an afoveate retina with a circular nasal area. Only two species, the

Owl-parrot (Psittaciformes) and Jackass Penguin (Sphenisciformes), had an afoveate retina with a circular temporal area. 130

Head Movements

The head movement characteristics of 52 species of birds were obtained from the literature (Table 1; Fig. 6-3B). Head movement characteristics were obtained for 40 species within the same genus, 57 species within the same family and 13 species within the same order as the species within the study. No head movement data were found for three species (Common Swift, Anna Hummingbird and Oilbird). The most common head movement character was head-bobbing (36.6%), followed by non-head-bobbing

(22.8%) and body-bobbing (22%). Also, there were 10.3% of species who performed other head movements and 8.3% of species who were rare head-bobbers, including species that head-bob while foraging and while swimming.

The character state reconstruction of head movement suggests that the ancestral character state of Class Aves was non-head-bobbing (Fig. 6-5). Two of the paleognath species, the Rhea (Rheiformes) and North Island Brown Kiwi (Apterygiformes), were non-head-bobbers. All Anseriformes and Pelecaniformes were non-head-bobbers, as were all Alcidae and most Laridae species (Charadriiformes, except the Black-headed

Gull) and most Falconiformes (except the Secretary Bird). Additionally, the Jackass

Penguin (Sphenisciformes), Common Flamingo (Phoenicopteriformes) and Kagu

(Rhynochetidae) were non-head-bobbers. The greatest concentration of body-bobbing species was found within two orders of birds: the Passeriformes and Piciformes; the remaining birds that body-bob also head-bob (see below). Head-bobbing was found in three paleognathes, the two Tinamou species (Tinamiformes) and the Westerman

Cassowary (Casuariiformes). All of the species within Galliformes, Ciconiiformes and 131 Columbiformes were found to be head-bobbers. Within the Charadriiformes, all species of Scolopacidae displayed head-bobbing, as did the Pied Avocet, Black-bellied

Plover and Stone Plover. A small number of species within Passeriformes head-bob

(species within Motacillidae and two species within Fringillidae) or were both head- bobbers and body-bobbers (all species within Corvidae, Sturnidae and Turdidae, and two species of Fringillidae). Within Ralliformes one species was a head-bobber and two were rare head-bobbers.

The other birds that were reported to head-bob either did so rarely or only situationally. Species that rarely head-bob were scattered across the phylogeny, including species within Struthioniformes (Ostriches), Podicipediformes (Great-crest

Grebes), Ralliformes (Crested and American Coots), Otididae (Great Bustards),

Charadriiformes (Charadriidae and Haematopodidae) and Cariamidae (Brazilian

Seriemas). Situational head-bobbing while foraging was only observed within the Black- headed Gull (Charadriiformes) and head-bobbing while swimming was only observed within the Great-crested Grebe (Podicipediformes). Finally, other head movements occurred in the Strigiformes, Psittaciformes, Alcedinidae (Coraciiformes), and some

Charadriiformes (Burhinidae and Charadriidae), as well as one procellariiform species

(Dark-bodied Shearwaters).

A pairwise comparison of retinal pattern and head movement did not show any significant correlation of the characters. However, the proportions of retinal patterns within the different head movement categories were significantly different from the expected (total) proportion and from the each other (p < 0.001; Fig. 6-6). Similarly, the proportions of head movement characters within the different retinal pattern categories 132 were significantly different from the expected (total) proportion and from the each other (p < 0.001; Fig. 6-7). When the character matrix is coded onto the phylogeny (Fig.

6-8) it is evident that groupings of characters occur within clades. For example, 84% of body-bobbers have a nasal unifoveate retina with circular areas (Fig. 6-6), yet these species were members of only two orders: Passeriformes and Piciformes (Fig. 6-8). Also,

100% of species with temporal unifoveate retinae with circular areas displayed other head movements (Fig. 6-7), however, all of the species are within the Order Strigiformes (Fig.

6-8). Therefore, some correlations between retinal pattern and head movement are evident, however, these are principally within clades.

Foraging Type

Foraging type was identified for 165 species of birds (Table 1; Fig. 6-3C). The majority of birds were visual foragers (77%), 10% were visual/tactile foragers, 7% were tactile foragers and 5% were nocturnal foragers. Additionally, there was also one filter- feeder (Common Flamingo; Phoenicopteriformes) and one scavenger (Bearded Vulture;

Falconiformes).

The ancestral character state reconstruction of foraging characters suggests that the ancestral character state of Class Aves was visual foraging (Fig. 6-9). Most of the paleognaths were visual foragers (except the noctural North Island Brown Kiwi;

Apterygiformes). All of the Galliformes, Columbiformes, Cuculiformes, Ralliformes,

Pelecaniformes and Piciformes were visual foragers. The majority of the passeriform species were also visual foragers, except the Varied Bunting and American Robin, both which were visual/tactile foragers. Also, the majority of the falconiform species were 133 visual foragers (except the scavenging Bearded Vulture), as were the majority of the species within Coraciiformes (except the visual/tactile foraging Common Hoopoe).

Within the Anseriformes there were two visual foraging species, two tactile foraging species and one species that was a visual/tactile forager. Within the Ciconiiformes the

Ardeidae contained all visual foragers, whereas the Threskiornithidae and Ciconidae contained all tactile foragers. All three foraging types were found in the Charadriiformes: the Laridae, Alcidae, Haematopodidae and Charadriidae contained all visual foragers; the

Scolopacidae included two visual foragers, three tactile foragers and six visual/tactile foragers; and the Burhinidae and Recurvirostridae included species that were visual/tactile foragers. Finally, most of the strigiform species were nocturnal foragers, except the Burrowing Owl which was a visual forager.

Pairwise comparisons of retinal pattern and foraging type did not show any significant correlations. However, the proportions of retinal patterns within the different foraging categories were significantly different from the expected (total) proportion and from the each other (p < 0.001; Fig. 6-10), and the proportions of foraging type within the different retinal patterns were significantly different from the expected (total) proportion and from the each other (p < 0.001; Fig. 6-11). Again, this is most likely due to the relationships of some characters within clades. For example, the temporal unifoveate retina with a circular area was found within the Strigiformes, which mainly consists of nocturnal foragers (the exception being the Burrowing Owl, a visual forager). Yet, not all nocturnal birds have this retinal configuration: the North Island Brown Kiwi (a paleognath) has a nasal unifoveate retina, a pattern shared with all other paleognaths, and the Oilbird has an afoveate retina with a band-shaped area (unfortunately this is the only 134 species within the Caprimulgiformes in which the retina pattern is known). Also, birds that were only tactile feeders had a nasal unifoveate retina with either a circular or band-shaped area and these birds are found within Scolopacidae, Anatidae and

Ciconiiformes. Birds that were visual/tactile foragers exhibited either a nasal unifoveate retina with a circular area (Varied Bunting and Amerian Robin [Passeriformes], Common

Hoopoe [Coraciiformes] and Stone Plover [Charadriiformes]) or band-shaped area (seven species within Charadriiformes, Great Bustard [Otididae] and Mallard [Anatidae]), or a bifoveate retina with circular areas (Brazilian Seriema [Cariamidae]). Additionally, a pattern that was observed across the phylogeny was that species with no area centralis, an afoveate retina (except the Oilbird, which is nocturnal) or a bifoveate retina with a band- shaped area (except the Flamingo, which is a filter-feeder) were visual foragers.

Pairwise comparisons of head movement characters and foraging type also did not

show any significant correlations. However, the proportions of foraging characters

within the different head movement categories were significantly different from the

expected (total) proportion and from the each other (p < 0.001; Fig. 6-12), and the

proportions of head movement characters within the different foraging catergories were

significantly different from the expected (total) proportion and from the each other (p <

0.001; Fig. 6-13). Once more, this is most likely due to the relationships of some

characters within clades. One of the most surprising results is that head-bobbing birds

were not only visual foragers but also tactile foragers (Figs. 6-12 and 6-13). The visually

foraging head-bobbing birds (Fig. 6-8) included all of the species within Ardeidae

(Ciconiiformes), Galliformes and Columbiformes, three paleognath species (Westerman

Cassowary, Elegant-crested Tinamou and Rufous Tinamou), five species within 135 Passeriformes, three species within Charadriiformes, the Secretary Bird

(Falconiformes) and Ypecha Rail (Ralliformes). The visually foraging rare head-bobbing birds included two species within Charadriiformes, two species within Ralliformes and the Ostrich (paleognath). The head-bobbing birds that were visual/tactile foragers included seven species within Charadriiformes and the Common Hoopoe

(Coraciiformes). The rare head-bobbing birds that were visual/tactile foragers included the Brazilian Seriema (Cariamidae) and Great Bustard (Otididae). The head-bobbing birds that were tactile foragers included three species within Charadriiformes and three species within Ciconiiformes.

The non-head-bobbing birds included visual, visual/tactile and tactile foragers, one nocturnal species (North Island Brown Kiwi, paleognath), one scavenger (Bearded

Vulture, Falconiformes) and one filter-feeder (Common Flamingo, Phoenicopteriformes).

Non-head-bobbing visually foraging birds included all of the species within

Pelecaniformes, seven species within Falconiformes, ten species within Charadriiformes, two species within Anatidae, the Black Hornbill (Coraciiformes), Jackass Penguin

(Sphenisciformes), Kagu (Rhynocetidae) and Rhea (paleognath). Only the Mallard

(Anatidae) was a non-head-bobbing visual/tactile forager. Non-head-bobbing birds that were tactile foragers included the Lesser Snow Goose and Blue Snow Goose (both

Anatidae).

Birds that displayed other head movements included visual foragers (Fig 6-8; all of the species within Psittaciformes, all of the kingfishers [Coraciiformes], the Burrowing

Owl [Strigiformes] and Dark-bodied Shearwater [Procellariiformes]) and nocturnal foragers (all of the Strigiformes, except the Burrowing Owl). 136 Body-bobbing visually foraging birds included all of the species within

Piciformes and twenty-two species within Passeriformes. The Varied Bunting

(Passeriformes) was the only body-bobber and visual/tactile forager. The twelve species within Passeriformes that were both body- and head-bobbers included eleven visual foragers and one visual/tactile forager (American Robin).

Discussion

Head-bobbing is predominantly considered to be a behavior that facilitates vision in birds. It has been hypothesized that head-bobbing may be beneficial to birds with eyes that face laterally (Whiteside, 1967), since these birds experience a streaming visual image during terrestrial locomotion. In these birds, head-bobbing is believed to provide three visual benefits: (1) the thrust phase allows better visual differentiation of food items on the ground (Davies and Green, 1988), (2) the thrust phase provides information for relative depth perception via motion parallax (Frost, 1978) and (3) the hold phase can be used to detect moving objects (Dunlap and Mowrer, 1930). Given these hypothesized visual functions, it was surprising to find that all visual foragers were not head-bobbers nor was head-bobbing limited to visual foragers.

Why don’t all visual-foraging birds head-bob? One of the most common “head movements” noted for visual foragers in this study occurred in synchrony with hopping gaits, i.e., body-bobbing in passeriform birds. For body-bobbing birds, the hopping gait may be similar to head-bobbing (Davies and Green, 1988): the brief stance phase of hopping may stabilize a visual image similar to the hold phase of head-bobbing birds, whereas the aerial phase of the hop may serve to gather information for depth perception 137 and the detection of stationary items akin to the thrust phase of head bobs. Similarly, non-head-bobbing birds that forage using intermittent locomotion, e.g., charadriiform birds that wait for prey and then rapidly dash to capture it, may be obtaining the same visual benefits of the hold and thrust phases of head-bobbing birds. The other head movements of birds, such as the vertical movements of the head of perched kingfishers, may also be used to gather depth information (Casperson, 1999). Finally, many non- head-bobbing visual foragers do not forage while walking, instead they forage in flight or while diving (see below).

Head-bobbing birds consist of both visual and tactile foragers. Of the visually foraging species only 33% are head-bobbers, yet 69% of the visual/tactile foragers and

67% of the tactile foragers are head-bobbers. Additionally, 15% of the visual/tactile foragers are rare head-bobbers, whereas only 8% of visual foragers are rare head-bobbers.

This suggests that head-bobbing is not only beneficial to visual foraging birds but also to visual/tactile and tactile foraging birds. Frost (1978: 194) stated that “detection of motion and depth have obvious survival value to predator or prey alike.” In other words, tactile foragers may be using head-bobbing for the detection of predators instead of the detection of prey, whereas visual foraging birds may head-bob for both predator and

prey/food detection. Furthermore, all head-bobbing birds may capitalize on the visual

signals occurring during the thrust and hold phases as a means for identifying features of

their environment. An interesting example of this is the head-bobbing of Pigeons during

landing from flight (Davies and Green, 1988): the alternating protraction and retraction of

the head (with no stabilization periods) is believed to allow the Pigeon to determine the 138 distance to the landing surface by gathering depth perception information via motion parallax.

Despite the wide range of foraging behaviors, birds that bob their heads during terrestrial locomotion (either actuated by the neck or by whole body movements) typically have nasal foveae. Some retained the primitive condition of a nasal unifoveate retina with a band-shaped area or evolved a nasal unifoveate retina with a circular area.

Nasal foveae, paired with the orientation of the orbits, receive visual information from the lateral or posterolateral visual fields with no binocular overlap. When walking terrestrially, these visual fields will appear to stream laterally by the bird since most birds have limited eye movements within the orbit. Head-bobbing would provide an opportunity to obtain detailed information (hold phase) and to enhance perception of motion (hold phase) and depth (thrust phase). Without character mapping, it would be tempting to causally link head-bobbing with foveation. Yet, a foveated retina is the primitive condition of birds whereas head-bobbing is not. Furthermore, there are birds without an area centralis and afoveate birds that head-bob. Most of these appear to represent evolutionary losses of retinal features within clades that otherwise head-bob during terrestrial locomotion (Wonga Wonga Dove within Columbiformes; California

Valley Quail, Harlequin Quail, Brush Turkey and Chicken within Galliformes).

What about birds with anteriorly-facing eyes? These birds include the

Strigiformes (Owls; Martin, 1984) and Falconiformes (Hawks, Eagles; Martin and Katzir,

1999), and they typically have temporal foveae (either in a unifoveate or bifoveate configuration). In theory, these birds should not need to head-bob because the visual signal to the temporal foveae is primarily in front of the head rather than streaming 139 laterally, and the depth perception function of the thrust phase is replaced by some degree by binocularity. The majority of strigiform birds are nocturnal foragers that largely rely on auditory cues when foraging in flight or from a perch (Table 1; Payne and

Drury, 1971). Yet the high degree of binocularity (50°; Martin and Katzir, 1999) and the unusual combination of rods and cones in the fovea (Fite, 1973) underscore that visual signals are undoubtedly important for activity during low light conditions typically encountered by Owls. Indeed, although Owls do not head bob, they display side-to-side head movements that are believed to improve depth perception (van der Willigen et al.,

2002). The falconiforms also have anteriorly-facing eyes and most do not head-bob but they differ from strigiform birds in having a bifoveate retina (both nasal and temporal foveae). The most common foraging method of falconiforms is visual foraging from a perch or in flight (Table 1), with the scavenging Bearded Vulture being a variant of a flight-based forager. Not all birds that have a bifoveate retina are known to have anteriorly facing eyes, but most share with the falconiforms visual detection of prey while aerial and their predisposition against head-bobbing. These include birds that forage in flight (Swallows, Shrikes) and those that plunge dive (Terns).

Yet it is premature to assume a causal link between bifoveation, non-head- bobbing and aerial prey detection as several exceptions exist. Among falconiforms, the

Secretary Bird is a head-bobber with a bifoveate retina that forages while striding overground. Other ground-foraging birds that are bifoveate include the American Bittern

(Coconiformes), that head-bobs while visually foraging, and the Brazilian Seriema

(Cariamidae), that occasionally head-bobs when it forages using visual and tactile cues.

Two independent factors may be driving head-bobbing in these three species. First, if 140 their eyes are less frontally oriented, then the birds may have little or no binocularity and may use head-bobbing to obtain depth perception. The capacity to see binocularly has not been evaluated in these species, but wading birds closely related to the American

Bittern (Ardeidae) have binocular overlap to a degree similar to falconiforms (although their binocular field extends 100° below the bill, possibly functioning in the detection of aquatic prey; Martin and Katzir, 1994, 1999). Second, temporal and nasal foveae obtain visual stimuli from different regions of the visual field. It is possible that ground- foraging bifoveate birds may opt to bob their heads when focusing on near objects using the nasal fovea in a manner similar to head-bobbing birds that have a nasal unifoveate retina. Clearly, head-bobbing in these birds that are bifoveate as well as ground-foragers deserves further study.

The final bifoveate species in the sample is the Common Flamingo. Unlike the previous three species discussed, the Flamingo does not head-bob during terrestrial locomotion nor does it forage visually as it is a filter feeder. Bifoveation (within a band- shaped area) in the Flamingo results in a much larger binocular field than is typical of tactile foragers (a feature that may be useful when feeding crop-milk to the chicks or in the construction of nests; Martin et al., 2005).

It is intriguing to note that the most derived (yet paradoxically unspecialized) retinal patterns are mainly found in visual foragers that head-bob. Of the sixteen species with an afoveate retina or a retina lacking an area centralis, nine are head-bobbers (within

Columbiformes and Galliformes), one is a rare head-bobber (Crest Coot), and the remaining three display other head movements (Psittaciformes). This leads to an intriguing possibility – did the evolution of head-bobbing and other head movements 141 allow for the loss of retinal features in birds? If the one of the main functions of avian foveae is to improve the detection of motion (Pumphrey, 1948; Martin, 1985), then perhaps the stabilization provided by the hold phase of a head-bob or between head movements adequately compensates for the lack of a fovea. Similarly, the thrust phase of head-bobbing and other head movements may help differentiate stationary items from the background and provide depth perception to foveated and afoveated birds alike. Indeed, foveation is an exception, not the rule, among non-avian animals, and many of these utilize head movements to improve visual function (e.g., depth perception in rabbits and rodents; Kral, 2003).

Finally, all tactile foragers have a nasal unifoveate retina, as do the vast majority of visual-tactile foragers (96%), and most of these species were head-bobbers. When actively probing and digging for food items most tactile foragers are not striding across the landscape and may not use visual cues to forage, however, they may be using visual cues to detect predator motion. Furthermore, head-bobbing during terrestrial locomotion in these species may function to identify microhabitats likely to house hidden prey items

(below leaf litter or in the shallows of the sea) or to detect predators.

In conclusion, it is evident from the discussion above that there appear to be associations among retinal pattern, head movement and foraging even though statistical analysis failed to find significance. In the present study, ten discrete retinal morphologies were identified but these could be modified (e.g. to a smaller number of characters with multiple character states, such as nasal fovea [absent/present], temporal fovea

[absent/present], and area centralis [absent/band/circular]). Similarly, because body- bobbing may represent head-bobbing actuated by the limbs, head movement characters 142 may be condensable as well. Finally, retinal morphology is known in only 165 species of birds – a small fraction of the ~10,000 species of birds – therefore the actual diversity and distribution of retinal morphologies is poorly known. There is a comparable deficiency in the data on head movement characteristics of birds. The addition of species to this analysis should help to discriminate whether there are any significant correlations between characters across Aves or if the characters are only loosely associated within clades.

143

Table 6-1: The retinal patterns, head movements and foraging types in birds.

Order

Family source Foraging and Genus species (Common name) sourceRetinal Head-bobbing food type source Retinal patternRetinal Location (fovea and/or macula) Head-bobbing Foraging type Foraging method(s)

Anseriformes Anatidae Chen caerulescens (Blue Snow Goose) U N NB T digging, walking 1 10 11 Chen hyperboreus (Lesser Snow Goose) UB LF N NBF T digging, walking 1 7 11 Anas bochas domesticus or UB N NB T surface dipping, diving 2; 3 7 11 Anas platyrhynchos (Mallard) Oidemia deglandi or Melanitta fusca UB N NBF V diving 3 7 11 (White-winged Scoter) Fuligula glacialis or Clangula hyemalis UB N NB V diving 2 7 11 (Long-tailed Duck) Apodiformes Apodidae Cypselus apus or Apus apus (Common Swift) UB T ? V in flight 2 - 11 Trochilidae Calypte annae (Anna Hummingbird) B N, T ? V in flight 1 11 Caprimulgiformes Steatornithidae band Steatornis caripensis (Oilbird) A ? NO in flight 4 11 area

144

Table 6-1: Continued. Casuariiformes Casuariidae Casuarius occipitalis (Westerman Cassowary) U N HBG V walking 1 7 11 Charadriiformes Recurvirostridae Recurvirostra avocetta (Pied Avocet) UB N HB V, T walking, probing 2 7 11 Laridae N, T in flight, plunge diving, Sterna hirundo (Common Tern) BB (above NBF V 1; 3 7 11 surface dipping band) Sterna macrura or Sterna paradisaea in flight, plunge diving, BB N, T NBF V 2 7 11 (Artic Tern) perched Sterna minuta or Sterna antillarum (Least Tern) BB N, T NBF V in flight, plunge diving 2 7 11

Sterna Cantiaca or Sterna sandviciensis BB N, T NBF V plunge diving 2 7 11 (Sandwich Tern) Larus argentatus (Herring Gull) UB LF NB V walking, plunge diving 1 10 11 NBG/ Larus marinus (Great black-backed Gull) U N V walking 1 7; 10 11 HB(f)G NBG/ in flight, walking, Larus canus (Common Gull) UB N V 2 7; 10 11 HB(f)G surface dipping surface dipping, Larus ridibundus (Black-headed Gull) UB N HB(f) V 2 9 11 walking Burhinidae Oedicnemus scolopax or Burhinus oedicnemus U N HB/ OHM V, T walking, probing 1 7 11 (Stone Plover or Stone Curlew)

145

Table 6-1: Continued. Charadriidae Squatarola squatarola or Pluvialis squatarola UB N HB V waiting and running 1; 2 7; 10 11 (Black-bellied Plover) Charadrius hiaticula (Common Ringed Plover) UB N HB*; OHM V waiting and running 2 7; 10 11 Charadrius pluvialis or Pluvialis dominica UB N HB*; OHM V waiting and running 2 7; 10 11 (American Golden Plover) Aegialitis semipalmata or Charadrius UB N NB V waiting and running 3 10 11 semipalmatus (Semipalmated Plover) Vanellus cristatus or Vanellus vanellus UB N HB* V waiting and running 2 7 11 (Northern Lapwing) Scolopacidae Totanus melanoleucus or Tringa melanoleuca UB N HB V walking 1 10 11 (Greater Yellowlegs) Strepsilas interpres or Arenaria interpres UB N HB V, T probing, walking 2 10 11 (Ruddy Turnstone) Numenius hudsonicus or Numenius phaeopus UB N HB V, T walking, probing 1 10 11 (Hudsonian Curlew or Whimbrel) Gallinago media (Great Snipe) U N HB; HB*G T probing 2 7; 10 11 Tringa Islandica or Calidris canutus (Red Knot) UB N HBG T probing 2 7 11 Tringa alpina or Calidris alpina (Dunlin) UB N HBG T probing 2 7 11 Ereunetes pusillus or Calidris pusilla UB N HB V, T walking, probing 3 10 11 (Semipalmated Sandpiper) Limosa Lapponica (Bar-tailed Godwit) UB N HBF V, T walking, probing 2 7; 10 11 Totanus glareola or Tringa glareola UB N HBG V, T walking, probing 2 7 11 (Wood Sandpiper) Totanus hypoleucus or Tringa hypoleucus UB N HBG V walking 2 7 11 (Common Sandpiper) Numenius arquata (Eurasian Curlew) UB N HBF V, T walking, probing 2 7; 10 11 146

Table 6-1: Continued. Alcidae Alca torda (Razorbill) UB N NBF V diving 2 7 11 Uria troile or Uria aalge (Foolish Guillemot) UB N NBF V diving 2 7 11 Cepphus columba (Pigeon Guillemot) UB N NBF V diving 1 7 11 Fratercula mormon or Fratercula artica (Puffin) UB N NB V diving 2 7 11 Haematopodidae Haematopus ostralegus (Eurasian Oystercatcher) UB N HB* V walking 2 7 11 Ciconiiformes Ardeidae Ardea cinerea (Grey Heron) U N HB V walking 2 7 11 Nycticorax nycticorax (European Night Heron) U N HBF V walking 1; 3 7; 8 11 Botaurus lentiginosus (American Bittern) B N, T HBF V walking 1 7; 8 11 Botaurus stellaris (European Bittern) U N HBF V walking 1 7; 8 11 Cancroma cochlearia (Boat-billed Night Heron) U N HB V walking 1 7 11 Threskiornithidae Plegadis falcinellus (Glossy Ibis) U N HB T probing 1 7 11 sweeping bill through Platalea leucorodia (Spoonbill) U N HB T 1 7 11 water Ciconidae sweeping bill through Mycteria americana (Wood Stork) U N HBG T 1 7 11 water Columbiformes Columbidae Leucosarcia picata (Wonga Wonga Dove) None - HBF V walking 1 5; 6; 7 11 Columba palumbus (British Wood Pigeon) A N HBG V walking 1 6; 7 11 Columba livia domesticus (Pigeon) U N HB V walking 3 6 11 147

Table 6-1: Continued. Goura victoria (Victoria Crowned Pigeon) A N HB V walking 1 5; 7 11 Coraciiformes Alcedinidae perched, plunge diving, Alcedo ispida (British Kingfisher) B N, T OHMG V 1 7 11 in flight Dacelo gigas (Laughing Kingfisher) B N, T OHMF V perched, in flight 1 7 11 perched, plunge diving, Ceryle alcyon (Belted Kingfisher) B N, T OHMF V 3 7 11 in flight Upupidae Upupa epops (Common hoopoe) U N HB V, T probing, digging 1 7 11 Bucerotidae Spagolobus adratus (Black Hornbill) U N NBF V walking, in flight 1 7 11 Cuculiformes Cuculidae HBF/ NBF/ Coccyzus americanus (Yellow-billed Cuckoo) U N V perched 1 7 11 BBF HBF/ NBF/ Cuculus canorus (European Cuckoo) A N V perched 1 7 11 BBF Dinornithiformes Apterygidae Apteryx mantelli (Mantell Apteryx) U N NBG NO walking 1 10 11 Falconiformes Accipitridae Buteo vulgaris or Buteo buteo in flight, perched, B N, T NBG V 2 7 11 (Common Buzzard) walking Buteo latissimus (Broad-winged Hawk) B N, T NBG V in flight, perched 1 7 11 148

Table 6-1: Continued. Buteo borealis or Buteo jamaicensis BB N, T NBG V in flight, perched 1 7 11 (Red-tailed Hawk) Gypaetus barbatus (Bearded Vulture) B N, T NB non-V scavenger 1 7 11 Haliaetus leucocephalus (Bald Eagle) B N, T NBG V in flight, perched 1 7 11 Haliaetus leucogaster (White-bellied Sea Eagle) B N, T NBG V in flight, perched 1 7 11 Sagittariidae Gypogeranus serpentarius or B N, T HB V walking 1 7 11 Sagittarius serpentarius (Secretary Bird) Falconidae Falco sparverius (Sparrow Hawk or B N, T NBG V in flight, perched 1; 3 7 11 American Kestrel) Tinnunculus alaudarius or Falco tinnunculus B N, T NBF V in flight, perched 1 7 11 (European Kestrel) Galliformes Odontophoridae Lophortyx californicus vallicola or None - HB V walking 1 7 11 Callipepla californica (California Valley Quail) Colinus virginianus (Northern Bobwhite) U N HB V walking 3 10 11 Phasianidae Gallus domesticus (Chicken) A N HB V walking 2; 3 6 11 Meleagris gallopavo (Turkey) U N HB V walking 2; 3 7 11 Cortunix histrionica (Harlequin Quail) None - HBG V walking 1 7 11 Perdix cinerea or Perdix perdix (Grey Partridge) U N HBG V walking 2 5; 7 11 Phasianus colchicus (Common Pheasant) U N HBF V walking 2 5; 6; 7 11 Tetraonidae Bonasa umbellus (Ruffed Grouse) U N HBG V walking 3 7 11

149

Table 6-1: Continued. Numididae Numida pucherani or Guttera pucherani U N HBG V walking 3 7 11 (Crested Guineafowl) Cracidae Crax globosa (Yarrell Curassow) U N HBG V walking 1 7 11 Megapodidae Catheturus lathami or Alectura lathami walking, in flight, A N HBF V 1 7 11 (Brush Turkey) digging Gruiformes Otididae Otis tarda (Great Bustard) UB N HB*F V, T walking, probing 1 7 11 Rhynocetidae Rhinochetus jubatus (Kagu) U N NB V waiting and running 1 7 11 Cariamidae Cariama cristata (Brazilian Seriema) B N, T HB* V, T walking, probing 1 7 11 Passeriformes Alaudidae Alauda arvensis (Sky Lark) U N HBO/ BBO V walking 2 7 11 Anthus pratensis (European Titlark) U N HBO/ BBO V walking 1 7 11 Certhiidae Certhia familiaris (Eurasian Tree-creeper) U N HBO/ BBO V, T hopping, probing 3 7 11 Corvidae Garrulus glandarius (Eurasian Jay) U N BB V arboreal, hopping 2 5; 7 30 Corvus frugilegus (Rook) U N HB/BB V walking 2 7 12; 13 Corvus corax (Raven) U N HBG/ BBG V walking, in flight 1 7 14 150

Table 6-1: Continued. walking, hopping, Corvus americanus (American Crow) UB N HBG/ BBG V 1; 3 7 15 arboreal, in flight hopping, arboreal, in Cyanocitta cristata (Blue Jay) U N HBF/ BBF V 1; 3 5; 7 16 flight, walking hopping, arboreal, in Cyanocitta stelleri (Stellar Jay) U N HBF/ BBF V 1 5; 7 17 flight Emberizidae Emberiza citrinella (Yellowhammer) U N BB V hopping, arboreal 2 7 30 Emberiza miliaria or Miliaria calandra UB N HBF V hopping, arboreal 2 7 30 (Corn Bunting) arboreal, probing, Cyanospiza versicolor (Varied Bunting) U N BBF V, T hopping, pecking, in 1 7 18 flight Poocaetes gramineus (Vesper Sparrow) U N BBF V hopping, arboreal 3 7 19 hopping, arboreal, Spizella pusilla (Field Sparrow) U N BBF V 3 7 20 perched Junco hyemalis (Dark-eyed Junco) U N BBF V hopping, arboreal 3 7 21 Melospiza fasciata or Melospiza melodia walking, arboreal, U N BBF V 3 7 22 (Song Sparrow) digging, in flight hopping, digging, in Passerella iliaca (Fox Sparrow) U N BBF V 3 7 23 flight Passerina cyanea (Indigo Bunting) U N BBF V hopping, arboreal 3 7 24 Fringillidae hopping, walking, Fringilla coelebs (Chaffinch) U N HB V 2 7 30 arboreal Fringilla canaria or Serinus canaria U N HBG V hopping, arboreal 2 7 30 (Island Canary) Fringilla cannabina or Carduelis cannabina UB N HBF/ BBF V hopping, arboreal 2 7 30 (Common Linnet) 151

Table 6-1: Continued. Spinus tristis or Carduelis tristis U N HBF/ BBF V arboreal 3 7 25 (American Goldfinch) Hirundininae Tachycineta bicolor (White-bellied Swallow) B N, T BBF V in flight 1; 3 7 11 Hirundo rustica (European Chimney Swallow) BB N, T BBG V in flight 2 7 11 Hirundo urbica or Delichon urbica B N, T BBG V in flight 2 7 11 (Northern House-martin) Icteridae Agelaius phoeniceus (Red-winged Blackbird) U N BBF V walking, gaping 3 7 26 Laniidae Lanius ludovicianus (Loggerhead Shrike) B N, T BBF V perched, in flight 1 7 11 Mimidae walking , hopping, Mimus polyglottus leucopterus B N, T HBO/ BBO V perched, in flight, 1 7 11 (Western Mockingbird) arboreal Motacillidae Motacilla alba (White Wagtail) UB N HB V walking, short flights 2 7 11 Motacilla flava (Yellow Wagtail) UB N HBG V walking, short flights 2 7 11 Paridae Parus caeruleus or Cyanites caeruleus (Blue Tit) U N BBG V arboreal, in flight 2 7 11 Parus major (Great Tit) U N BBG V arboreal 2 7 11

Parus atricapillus (Black-capped Chickadee) U N BB V arboreal, in flight 1; 3 10 11

Parulidae walking, arboreal, in Dendroica palmarum (Palm Warbler) U N HBO/ BBO V 3 7 26 flight 152

Table 6-1: Continued. walking, arboreal, in Seiurus aurocapillus (Ovenbird) U N HBO/ BBO V, T 3 7 27 flight, probing Geothlypis philadelphia or U N HBO/ BBO V arboreal 3 7 28 Oporomis philadelphia (Mourning Warbler) Sylvania mitrata or Wilsonia citrina arboreal, hopping, in U N HBO/ BBO V 3 7 29 (Hooded Warbler) flight Passeridae Passer domesticus (House Sparrow) U N BB V hopping, arboreal 1; 3 7 30 Fringilla montanan or Passer montanus U N BBG V hopping, arboreal 2 7 30 (Eurasian Tree Sparrow) Prunellidae Accentor modularis (Dunnock) U N HBO/ BBO V, T walking, probing 2 7 11 Sylviidae Sylvia cinerea or Sylvia communis U N BBF V hopping, in flight 2 7 11 (Common Whitethroat) Sylvia hortensis (Orphean Warbler) U N BBF V hopping, in flight 2 7 11 arboreal, in flight, Regulus satrapa (Golden-crowned Kinglet) U N BBF V 3 7 11 hopping Sturnidae walking, hopping, in Sturnus vulgaris (Common Starling) U N HB*/ BB V 2 7 30 flight, arboreal Turdidae Merula migratoria or Turdus migratorius walking, in flight, U N HBG/ BBG V, T 3 7 11 (American Robin) probing Saxicola rubethra (Whinchat) UB N HBF/ BBF V arboreal, in flight 2 7 11 walking, in flight, Aedon megaryncha (English Nightingale) U N HBF/ BBF V 1 7 11 arboreal 153

Table 6-1: Continued. Sialia sialis or Pluvialis squatarola hopping, in flight, B; U N, T; N HBF/ BBF V 1; 3 7 11 (Eastern Bluebird) arboreal Saxicola oenanthe or Oenanthe oenanthe walking, hopping, in UB N HBF/ BBF V 2 7 11 (Wheatear) flight, arboreal, digging Tyrannidae perched, in flight, Pitangus derbianus (Derby Tyrant) U N HBO/ BBO V 1 7 11 plunge diving perched, in flight, Tyrannus tyrannus (Eastern Kingbird) U N HBO/ BBO V 3 7 11 plunge diving Pelecaniformes Phalacrocoracidae Phalacrocorax carbo (Great Cormorant) U N NB V diving 1 7 11 Phalacrocorax penicillatus (Brandt Cormorant) UB N NBG V diving 1 7 11 Sulidae Sula bassana or Morus bassana (Gannet) U N NB V plunge diving 1 7 11 Pelecanidae Pelecanus conspicillatus (Australian Pelican) U N NBG V in flight, plunge diving 1 7 11 Phoenicopteriformes Phoenicopteridae LF N, Phoenicopterus roseus (Common Flamingo) BB NB filter filter 1 7 11 LF T Piciformes Picidae arboreal, hopping, in Colaptus mexicanus (Red-shafted Flicker) U N BBG V 1 7 11 flight arboreal, hopping, in Colaptus auratus (Northern Flicker) U N BB V 3 7 11 flight 154

Table 6-1: Continued. Melanerpes erythrocephalus arboreal, hopping, in U N BBG V 1 7 11 (Red-headed Woodpecker) flight arboreal, hopping, in Centurus uropygialis (Gila Woodpecker) U N BBF V 1 7 11 flight Dryobates vel Dendrocopus major or arboreal, hopping, in U N BBF V 1; 2 7 11 Dendrocops major (Great Spotted Woodpecker) flight Podicipediformes Podicipedidae HB*/ Podiceps cristatus (Great Crested Grebe) UB N V diving 1 5 11 HB(ws) Procellariiformes Procellariidae Puffinus griseus (Dark-bodied Shearwater) UB LF OHMO V plunge diving, diving 1 7 11 Psittaciformes Cacatuidae Cacatua galerita none - OHMO V walking, perched 1 7 11 (Great Sulphur-crested Cockatoo) Psittacidae Chrysotis amazona or A mazona amazonica U N OHMF V walking, perched 1 7 11 (Orange-winged Parrot) Conurus ridgway or Aratinga euops None - OHMF V walking, perched 1 7 11 (White-fronted Cuban Conure) Stringops habroptilus (Kakapo or Owl-parrot) A T OHMF V walking, perched 1 7 11 Ralliformes Rallidae

Fulica americana (American Coot) UB N HB*G V diving, walking 1 7 11 155

Table 6-1: Continued. Fulica cristata (Crested Coot) None - HB*G V diving, walking 1 7 11 Aramides ypecaha (Ypecha Rail) U N HB V walking 1 7 11 Rheiformes Rheidae Rhea americana (Rhea) U N NB V walking 1 7 11 Sphenisciformes Spheniscidae Spheniscus demersus (Jackass Penguin) A T NBG V diving 1 7 11 Strigiformes Strigidae Strix vel Megascops asio or Otus asio U T OHMF NO perched, walking 1; 3 7 11 (American Screech Owl) Bubo virginianus (Great Horned Owl) U T OHMF NO perched, walking 1 7 11 Syrnium nebulosum or Strix varia (Barred Owl) U T OHMG NO perched, walking 1; 3 7 11 Strix flammea or Tyto alba (European Barn Owl) U T OHMF NO in flight, perched 1 7 11 Strix otus or Asio otus (Long-eared Owl) U T OHMF NO in flight, perched 2 7 11 in flight, perched, Syrnium aluco or Strix aluco (Tawny Owl) U T OHM NO 1 7 11 walking perched, in flight, Speotyto cunicularia hypogaea (Burrowing Owl) U T OHMF V 1 7 11 walking Struthioniformes Struthionidae Struthio camelus (Ostrich) U N HB* V walking 1 7 11 Tinamiformes Tinamidae Rhyncotus rufescens (Rufous Tinamou) U N HBF V walking 1 10 11 156

Table 6-1: Continued. Calodroma elegans or Eudromia elegans (Martineta Tinamou or U N HBF V walking 1 10 11 Elegant-crested Tinamou)

Retinal features: None, no area centralis; A, afoveate; U, unifoveate; B, bifoveate; UB, unifoveate with band-shaped area; BB, bifoveate with band-shaped area.

Location: N, nasal; T, temporal; LF, linear fovea.

Headbobbing: NB, non-head-bobber; NBG, non-head-bobber within same genus; NBF, non-head-bobber within same family; HB, head-bobber; HBG, head- bobber within same genus; HBF, head-bobber within the same family; HBO, Head-bobber within the same order; HB(f), head-bobs while foraging; HB(f)G, bird the head-bobs while foraging within same genus; HB(ws), head-bobs while swimming; HB*, occasionally head-bobs; HB*G, occasional head-bobber within same genus; HB*F, occasional head-bobber within same family; OHM, other head movements; OHM, other head movements within same genus; OHMF, other head movements within same family; OHMO, other head movements within same order; BB, body-bobbing; BBG, body-bobbing within the same genus; BBF, body-bobbing within the same family; BBO, body-bobbing within the same order.

Foraging type: V, visual; T, tactile; NO, nocturnal.

Food type: H, herbivorous; C, carnivorous; O, omnivorous.

Foveation references: 1, Wood, 1917; 2, Chievitz, 1891; 3, Slonaker, 1897; 4, Pettigrew and Konishi, 1984.

Head-bobbing references: 5, Dagg, 1977; 6, Dunlap and Mowrer, 1930; 7, Jimenez-Ortega, 2005; 8, Fujita, 2003; 9, Fujita, 2006; 10, Hancock, personal observation.

Foraging references: 11, del Hoyo, 1992-2008; 12, Kasprzykowski, 2003; 13, Mason and MacDonald, 2004; 14, Boarman and Heinrich, 1999; 15, Verbeek and Caffrey, 2002; 16, Tarvin and Woolfenden, 1999; 17, Greene et al., 1998; 18, Groschupf and Thompson, 1998; 19, Jones and Cornely, 2002; 20, Carey et al., 2008; 21, Nolan et al., 2002; 22, Arcese et al., 2002; 23, Weckstein et al., 2002; 24, Payne, 2006; 25, McGraw and Middleton, 2009; 26, Yasukawa and Searcy, 1995; 26, Wilson, 1996; 27, Van Horn and Donovan, 1994; 28, Pitocchelli, 1993; 29., Ogden and Stutchbury, 1994; 30, Witherby, 1943. 157

Photoreceptor layer

Retina

Vitreous humor

Light rays

Figure 6-1. Magnifying effect of the avian fovea. Arrows represent light which becomes refracted radially when passing through the fovea due to the higher refractive index of the retina relative to the vitreous humor. This results in image enlargement on the photoreceptor layer.

158

Figure 6-2: Schematics of the fundus oculi of birds showing the different retinal configurations known: black oval – pecten; grey shading – area centralis; black dot – circular fovea; black line – linear fovea (Wood, 1917; Chievitz, 1891; Slonaker, 1897; Pettigrew and Konishi, 1984). The boxes indicate retinal configurations that were merged for this analysis, resulting in ten different retinal patterns.

159

1% No area centralis 1% 4% Afoveate circular nasal area 5% 3% 11% 1% Afoveate circular temporal area Nasal unifoveate with a circular area 5% Nasal unifoveate with a band-shaped area Temporal unifoveate with a circular area 44% Bifoveate with circular areas 44% 25% Bifoveate with a band-shaped area Temporal unifoveate with a band-shaped area A Amacular band area

1% 1% 10% 5% 22% 9% Visual Body-bobbers 7% Tactile Head-bobbers 23% Visual and Tactile Rare Head-bobbers Nocturnal Non-head-bobbers Scavenger Other head movements 8% Filter-feeder 37% C 77% B Total

Figure 6-3. Pie graphs of the A) retinal patterns, B) head movements, C) foraging type, D) foraging method and E) food type within the species studied.

160

Figure 6-4. Character reconstruction of retinal morphology. 161

Figure 6-4. Continued. 162

Figure 6-5. Character reconstruction head movements within Class Aves. 163

Figure 6-5. Continued 164

Total

Other Head No area centralis Movements Afoveate circular nasal area

Non-head- Afoveate circular temporal area bobbers Nasal unifoveate with a circular area

Rare Head- Nasal unifoveate with a band-shaped area bobbers

Headmovement Temporal unifoveate with a circular area

Bifoveate with circular areas Head-bobbers Bifoveate with a band-shaped area

Body-bobbers

0% 20% 40% 60% 80% 100% Retinal pattern

Figure 6-6. The proportion of retinal patterns within the total study population and within each head movement category.

Total

Bifoveate with a band-shaped area

Bifoveate with circular areas

Temporal unifoveate with a circular area Body-bobbers Head-bobbers Nasal unifoveate with a band-shaped Rare Head-bobbers area Non-head-bobbers

Retinal pattern Nasal unifoveate with a circular area Other Head Movements

Afoveate circular temporal area

Afoveate circular nasal area

No area centralis

0% 20% 40% 60% 80% 100% Head movement

Figure 6-7. The proportion of head movement characters within the total study population and within each retinal pattern category. 165 ABC

Figure 6-8. Characters for A) retinal pattern, B) head movement and C) foraging type. The colors of A, B and C correspond with Figs. 6-4, 6-5 and 6-9, respectively. 166 ABC

Figure 6-8. Continued. 167

Figure 6-9. Character reconstruction of foraging type.

168

Figure 6-9. Continued. 169

Total

No area centralis Nocturnal Afoveate circular nasal area Afoveate circular temporal area

Visual and Nasal unifoveate with a circular area Tactile Nasal unifoveate with a band-shaped area

Foraging type Foraging Temporal unifoveate with a circular area Bifoveate with circular areas Tactile Bifoveate with a band-shaped area

Visual

0% 20% 40% 60% 80% 100% Retinal pattern

Figure 6-10. The proportion of retinal patterns within the total study population and within each foraging type category.

Total

Bifoveate with a band-shaped area

Bifoveate with circular areas

Temporal unifoveate with a circular area Visual Tactile Nasal unifoveate with a band-shaped area Visual and Tactile Nocturnal

Retinal pattern pattern Retinal Nasal unifoveate with a circular area

Afoveate circular temporal area

Afoveate circular nasal area

No area centralis

0% 20% 40% 60% 80% 100% Foraging type

Figure 6-11. The proportion of foraging type within the total study population and within each retinal pattern category. 170

Total

Other Head Movements

Non-head-bobbers Visual Tactile Visual and Tactile Rare Head-bobbers Nocturnal Head movement

Head-bobbers

Body-bobbers

0% 20% 40% 60% 80% 100% Foraging type

Figure 6-12. The proportion of foraging type within the total study population and within each head movement category.

Total

Nocturnal

Body-bobbers Head-bobbers Visual and Tactile Rare Head-bobbers Non-head-bobbers

Foraging type Other Head Movements

Tactile

Visual

0% 20% 40% 60% 80% 100% Head movement

Figure 6-13. The proportion of head movement characters within the total study population and within each foraging type category. 171 CHAPTER 7: CONCLUSIONS AND FUTURE DIRECTIONS

Ecomorphology specifies a causal relationship between morphology, performance and fitness, with attention on realized niche (Bock and von Wahlert, 1965; Arnold, 1983).

Since its inception, numerous researchers have structured their studies in the ecomorphological paradigm and have found it to be powerful tool for understanding ecological and evolutionary processes (e.g., Wainright and Reilly, 1994). This dissertation touches on multiple levels of the ecomorphological framework through its exploration of the visual system, terrestrial locomotion and foraging ecology in birds, with vision serving as the overarching theme for the dissertation.

The avian visual system was explored directly through the evolution of retinal morphology (Chapter 5). This study amassed the largest catalogue of avian retinal features (foveation, area centralis). The primitive condition for Aves is hypothesized to be a nasal unifoveate retina with a band-shaped area centralis; from this configuration, at least nine other retinal patterns likely evolved.

The study of the co-evolution of retinal patterns and foraging methods (Chapter 6) failed to find significant correlations although several have been postulated in the literature (Walls, 1942; Whiteside, 1967; Munk, 1970; Fite, 1973; Kröger and Katzir,

2008). With few exceptions, a similar suite of retinal morphologies occur in terrestrial foragers that use visual and/or tactile means to detect food items. This suggests that key functions of the visual system are shared among foraging groups, including navigation through the habitat, location of microhabitats likely to yield foraging success (regardless of whether the actual identification of the food is visual or tactile), avoidance of predators 172 and finding mates. Some associations between retinal patterns and foraging methods were noted (e.g., nocturnal foragers and a temporal unifoveate retina with a circular area).

But these associations were not “rules” as exceptions were common (e.g., the nocturnal

Mantell Apteryx with its nasal unifoveate retina).

The dissertation (Chapter 6) failed to support the predominant hypothesis that head-bobbing should be linked with visually-guided ground foragers (Davies and Green,

1988). Head-bobbing is used by visual, visual/tactile and tactile foragers alike, in part because of the shared visual needs of all birds (noted previously). Not surprisingly, given the diversity of retinal morphologies related to visual and tactile foragers, head- bobbing birds display a range of retinal patterns, from unspecialized (no area centralis) to bifoveate.

By relating terrestrial locomotion to head-bobbing, insight was obtained into the consequences of limb kinematics and center of mass (COM) mechanics on the visual system. The predominance of grounded running in birds, with its limited vertical excursion of the COM, may be interpreted as strategy for limiting vertical oscillations of the visual apparatus (Chapter 2 and 3). Similarly, the potential jarring effect of hindlimb touchdown on the visual apparatus (McMahon et al., 1987) is avoided by initiating the hold phase after the liftoff of a hind limb and ending it before touchdown of the limb

(Chapter 4). Variations in the relationship between head-bobbing and locomotor kinematics (e.g., synchronous elongation of hind limb support duration and head-bobbing hold phase) were interpreted as being driven by the visual needs of the bird. This interpretation contrasts that offered by Necker (2007); he designated stride length as the actuator for head-bobbing in Black-Headed Gulls. 173 Finally, differences in the degree of synchronization of head-bobbing and locomotor kinematics between the field-based study (charadriiforms; Chapter 4) and the laboratory trackway study (Tinamous; Chpater 3) led to several intriguing possibilities.

First, coordination of head-bobbing and locomotion may be facultative and may not be similarly beneficial to a bird in a lab versus in the field (when actively foraging). Second, because the ancestral condition of birds appears to be non-head-bobbing and because numerous clades appear to have evolved head-bobbing independently (Chapter 6), the low degree of coordination between head-bobbing and locomotion in Tinamou may actually reflect the primitive condition for birds, for paleognaths or for the Tinamidae.

Future Research

Mechanics of locomotion

Although grounded running is a contrived gait in humans, it is a normal gait in

many other vertebrates (Biknevicius and Reilly, 2006), including birds. Future research

in the mechanics of locomotion is necessary to determine how common grounded

running is across terrestrial vertebrates. Furthermore, the timing of gait transitions from

vaulting mechanics to grounded running and from grounded running to aerial running has

been poorly explored – principally because the focus of gait transition studies has been on

larger cursorial mammals that tend to shift from walking to aerial running with little or no

grounded running. Additionally, intermediate mechanics during symmetrical gaits are

most commonly found in small terrestrial mammals (Biknevicius et al. 2007) and now in

birds (Hancock et al., 2007). Future research should explore whether intermediate

mechanics is limited to small-bodied vertebrates that are supported by limbs that have 174 highly flexed joint(s). Large comparative studies such as these must utilize size- adjusted parameters, such as relative speed or Froude number rather absolute speed.

Head-bobbing and locomotion

There are few studies that explored both head-bobbing kinematics and locomotor kinematics. Differences in data reporting make comparison among these studies difficult.

For example, Fujita (2002, 2003) suggests that head-bobbing and locomotion movements are synchronized, yet the author only reported the ranges and means of the durations from a head-bobbing event to a locomotor event and did not analyze or illustrate how well the events are synchronized. This type of data reporting hinders the field of bird locomotion because it is impossible to specify whether events are precisely synchronized, coordinated but not synchronized, or neither. An expanded sampling of head-bobbing birds is sorely needed. Data on other paleognaths would help resolve the question of whether poor coordination of head-bobbing and locomotor events may be phylogenetically based. Also, revisiting Tinamou locomotion in the field is warranted since only lab data were obtainedin this study; it is possible that foraging Tinamous in the wild will coordinate head and leg movements differently (suggesting that the laboratory setting influences the coordination).

Head-bobbing, retinal morphology and ecology

Data relating head movement, retinal morphology and feeding ecology is available for only ~1% of all bird species and there is great variability in the data reported for each of these variables. For example, many studies discuss foveation but do not 175 include information on the shape and location of the area centralis. There are also numerous ways to define feeding behavior. Although this dissertation focused on foraging type (the means by which birds senses the food item, e.g., visual, tactile), foraging behavior is far more complex and includes movement patterns and other activities by the individual during foraging (e.g., walking, surface dipping; Fig. 7-1).

Analyzing this added layer of complexity may provide greater insight into the relationship between retinal morphology and head movements. Extending this analysis to other vertebrates (e.g., lizards) that also employ head movements may provide insight into the co-evolution of the visual apparatus, head movements and feeding ecology.

176

Figure 7-1. Characters for A) retinal pattern, B) head movement, C) foraging type and D) foraging method. 177

Figure 7-1. Continued. 178

Legend for column A Legend for column B

Legend for column C Legend for column D

Figure 7-1. Continued.

179 REFERENCES

Abourachid, A. 2000. Bipedal locomotion in birds: The importance of functional

parameters in terrestrial adaptation in Anatidae. Can. J. Zool . 78: 1994-1998.

Abourachid, A. 2001. Kinematic parameters of terrestrial locomotion in cursorial

(ratites), swimming (ducks), and striding birds (quail and guinea fowl). Comp.

Biochem. Physiol. Part A. 131: 113-119.

Abourachid, A. & Renous, S. 2000. Bipedal locomotion in ratites (Paleognatiform):

examples of cursorial birds. Ibis. 142: 538-549.

Ahn, A.N., Furrow, E., & Biewener, A.A. 2004. Walking and running in the red-legged

running frog, Kassina maculate . J. Exp. Biol. 207: 399-410.

Alexander, R.McN. 1976. Mechanics of bipedal locomotion. In Perspectives in

Experimental Biology . Davies, P. S. (ed.). Vol. 1, Zoology: pp.493-504. Oxford:

Pergamon.

Alexander, R.McN. 1977. Mechanics and scaling of terrestrial locomotion. In Scale

Effects in Animal Locomotion. Pedley, T.J. (ed). pp. 93-110. New York:

Academic.

Alexander, R.McN. & Jayes, A.S. 1983. A dynamic similarity hypothesis for the gaits of

quadrupedal mammals. J. Zool., Lond. 201: 135-152.

Alström, P., Ericson, P.G.P., Olsson, U. & Sundberg, P. 2006. Phylogeny and

classification of the avian superfamily Sylvioidea. Mol. Phylogenet. Evol. 38:

381-397.

Arcese, P., Sogge, M.K., Marr, A.B. & Patten, M.A. 2002. Song Sparrow ( Melospiza

melodia ). In The Birds of North America Online . Poole, A. (ed.). Ithaca: Cornell 180 Lab of Ornithology; Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/704.

Arnold, S.J. 1983. Morphology, performance and fitness. Amer. Zool . 23: 653-667.

Barker, F.K., Cibois, A., Schikler, P., Feinstein, J. & Cracraft, J. 2004. Phylogeny and

diversification of the largest avian radiation. P. Natl. Acad. Sci. Biol. 101: 11040-

11045.

Benz, B.W., Robbins, M.B. & Peterson, A.T. 2006. Evolutionary history of woodpeckers

and allies (Aves: Picidae): Placing key taxa on the phylogenetic tree. Mol.

Phylogenet. Evol. 40: 389-399.

Biewener, A.A. & Blickhan, R. 1988. Kangaroo Rat locomotion: design for elastic

energy storage or acceleration? J. Exp. Biol. 140: 243-255.

Biknevicius, A.R. & Reilly, S.M. 2006. Correlation of symmetrical gaits and whole body

mechanics: Debunking myths in locomotor biodynamics. J. Exp. Zool. 305A:

923-934.

Biknevicius, A.R., Reilly, S.M., White, T.D., Hancock, J.A. & McElroy, E.J. 2007.

Interpreting the significance of whole body mechanics in primitive mammals. J.

Morphol. 268:1049.

Blickhan, R. & Full, R.J. 1992. Mechanical work in terrestrial locomotion. In

Biomechanics: Structures and Systems. Biewener, A. A. (ed). pp. 75-96. New

York: Oxford University Press.

Boarman, W.I. & Heinrich, B. 1999. Common Raven ( Corvus corax ). In The Birds of

North America Online. Poole, A. (ed.). Ithaca: Cornell Lab of Ornithology; 181 Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/476.

Bock, W.J. & von Wahlert, G. 1965. Adaptation and the form-function complex.

Evolution . 19: 269-299.

Bullimore, S.R. & Burn, J.F. 2005. Scaling of elastic energy storage in mammalian limb

tendons: do small mammals really lose out? Biol. Lett. 1: 57-59.

Carey, M., Burhans, D.E. & Nelson D.A. 2008. Field Sparrow ( Spizella pusilla ). In The

Birds of North America Online. Poole, A. (ed.). Ithaca: Cornell Lab of

Ornithology; Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/103.

Casperson, L.W. 1999. Head movement and vision in underwater-feeding birds of

stream, lake, and seashore. Bird Behav. 13: 31-46.

Cavagna, G.A., Heglund, N.C. & Taylor, C.R. 1977. Mechanical work in terrestrial

locomotion: two basic mechanisms for minimizing energy expenditure. Am. J.

Physiol. 233: R243-R261.

Chievitz, J. H. 1891. Über das Vorkommen der Area centralis retinae in den vier höheren

Wirbelthierklassen. Arch. Anat. Entw. 1891:11–334.

Clark, J. & Alexander, R. McN. 1975. Mechanics of running by quail ( Coturnix ). J.

Zool., Lond. 176: 87-113.

Connolly, M. & Van Essen, D. 1984. The representation of the visual field in

parvicellular and magnocellular layers of the lateral geniculate nucleus in the

macaque monkey. J. Comp. Neurol. 226: 544-564. 182 Cronin, T.W., Kinloch, M.R. & Olsen, G.H. 2005. Head-bobbing behavior in

foraging whooping cranes favors visual fixation. Curr. Biol. 15:R243-R244.

Daanje, A. 1951. On locomotory movements in birds and the intention movements

derived from them. Behaviour. 3:48-98.

Dagg, A.I. 1977. The walk of the silver gull ( Larus novaehollandiae ) and of other birds.

J. Zool., Lond. 182:529-540.

Davies, M.N.O. & Green, P.R. 1988. Head-bobbing during walking, running and flying:

relative motion perception in the pigeon. J. Exp. Biol. 138: 71-91. de Kloet, R.S. & de Kloet, S.R. 2005. The evolution of the spindlin gene in birds:

Sequence analysis of an intron of the spindlin W and Z gene reveals four major

divisions of the Psittaciformes. Mol. Phylogenet. Evol. 36: 706-721 del Hoyo, J., Elliott, A. & Sargatal, J. eds. 1992 – 2008. Handbook of the birds of the

world. Volumes 1 – 13. Barcelona: Lynx Edicions.

Duijm, M. 1958. On the position of a ribbon-like central area in the eyes of some birds.

Arch. Neerl. Zool. 13: 128-145.

Dunlap, K. & Mowrer, O.H. 1930. Head movements and eye functions of birds. J. Comp.

Psychol. 11:99-113.

Dyke, G.J., Gulasand, B.E. & Crowe, T.M. 2003. Suprageneric relationships of galliform

birds (Aves, Galliformes): a cladistic analysis of morphological characters. Zool.

J. Linn. Soc. Lond. 137: 227–244.

Ericson, P.G.P., Janse´n, A.L., Johansson, U.S. & Ekman, J. 2005. Inter-generic

relationships of the crows, jays, magpies and allied groups (Aves: Corvidae)

based on nucleotide sequence data. J. Avian Biol. 36: 222-234. 183 Ericson, P.G.P. & Johansson, U.S. 2003. Phylogeny of Passerida (Aves:

Passeriformes) based on nuclear and mitochondrial sequence data. Mol.

Phylogenet. Evol. 29: 126-138.

Farley, C.T. & Taylor, C.R. 1991. A mechanical trigger for the trot-gallop transition in

horses. Science. 253: 306-308.

Feldman, C. R. & Omland, K. E. 2005. Phylogenetics of the common raven complex

(Corvus: Corvidae) and the utility of ND4, COI and intron 7 of the β-fibrinogen

gene in avian molecular systematics. Zool. Scr. 34: 145–156.

Fewster, J.B. & Smith, G.A. 1996. The influence of midstance vertical ground reaction

force on the human walk-run transition. Med. Sci. Sport Exer . 28: (Suppl. 5),

S87.

Fite, K.V. 1973. Anatomical and behavioral correlates of visual acuity in the Great

Horned Owl. Vision Res. 13: 219-230.

Friedman, M.B. 1975. Visual control of head movements during avian locomotion.

Nature. 255: 67-69.

Frost, B.J. 1978. The optokinetic basis of head-bobbing in the pigeon. J. Exp. Biol.

74:187-195.

Fuchs, J., Fjeldsa, J., Bowie, R.C.K., Voelker, G. & Pasquet, E. 2006. The African

warbler genus Hyliota as a lost lineage in the Oscine songbird tree: molecular

support for an African origin of Passerida. Mol. Phylogenet. Evol. 39: 186-197.

Fujita, M. 2002. Head-bobbing and the movement of the center of gravity in walking

pigeons ( Columba livia ). J. Zool., Lond. 257: 373-379. 184 Fujita, M. 2003. Head-bobbing and the body movement of Little Egrets ( Ergetta

garzetta ) during walking. J. Comp. Physiol. A. 189: 53-58.

Fujita, M. 2004. Kinematic parameters of the walking of herons, ground-feeders and

waterfowl. Comp. Biochem. Physiol. Part A. 139: 117-124.

Fujita, M. 2006. Head-bobbing and non-bobbing walking of black-headed gulls ( Larus

ridibundus ). J. Comp. Physiol. A 192:481-488.

Fujita, M. & Kawakami, K. 2003. Head-bobbing patterns, while walking, of Black-

winged Stilts Himantopus himantopus and various herons. Ornithol. Sci. 2:59-63

Gatesy, S.M. 1999. Guineafowl hind limb function. I: Cineradiographic analysis and

speed effects. J. Morphol . 240: 115-125.

Gatesy, S.M. & Biewener, A.A. 1991. Bipedal locomotion: effects of speed, size and

limb posture in birds and humans. J. Zool., Lond. 224: 127-147.

Gatesy, S.M. & Dial, K.P. 1996. Locomotor modules and the evolution of avian flight.

Evolution. 50: 331-340.

Grapputo, A., Pilastro, A., Baker A.J. & Marin, G. 2001. Molecular evidence for

phylogenetic relationships among buntings and American sparrows

(Emberizidae). J. Avian Biol. 32: 95-101.

Greenlaw, J.S. 1969. The importance of food in the breeding system of the rufous-sided

towhee, Pipilo erythrophthalmus (L.). PhD dissertation, Rutgers, State Univ. of

New Jersey, New Brunswick, NJ.

Greene, E., Davison, W. & Muehter, V.R. 1998. Steller's Jay ( Cyanocitta stelleri ). In The

Birds of North America Online. Poole, A. (ed.). Ithaca: Cornell Lab of 185 Ornithology; Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/343.

Griffin, T.M. & Kram, R. 2000. Penguin waddling is not wasteful. Nature. 408: 929.

Griffin, T.M., Kram, R., Wickler, S.J. & Hoyt, D.F. 2004. Biomechanical and energetic

determinants of the walk-trot transition in horses. J. Exp. Biol. 207: 4215-4223.

Groschupf, K.D. & Thompson, C.W. 1998. Varied Bunting ( Passerina versicolor ). In

The Birds of North America Online. Poole, A. (ed.). Ithaca: Cornell Lab of

Ornithology; Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/351.

Hackett, S.J., Kimball, R.T., Reddy, S., Bowie, R.C.K., Braun, E.L., Braun, M.J.,

Chojnowski, J.L., Cox, W.A., Han, K., Harshman, J., Huddleston, C.J., Marks,

B.D., Miglia, K.J., Moore, W.S., Sheldon, F.H., Steadman, D.W., Witt, C.C. &

Yuri, T. 2008. A phylogenomic study of birds reveals their evolutionary history.

Science. 320: 1763-1768.

Hancock, J.A, Biknevicius, A.R. & Stevens, N.J. 2007. Whole-body mechanics and

kinematics of terrestrial locomotion in the Elegant-crested Tinamou ( Eudromia

elegans ). Ibis . 149: 605–614.

Hayes, G. & Alexander, R.McN. 1983. The hopping gaits of crows (Corvidae) and other

bipeds. J. Zool., Lond. 200: 205-213.

Heglund, N.C., Cavagna, G.A. & Taylor, C.R. 1982. Energetics and mechanics of

terrestrial locomotion. III. Energy changes of the center of mass as a function of

speed and body size in birds and mammals. J. Exp. Biol. 97: 41-56.

Hildebrand, M. 1965. Symmetrical gaits of horses. Science. 150: 701-708. 186 Hildebrand, M. 1976. Analysis of tetrapod gaits: general considerations and

symmetrical gaits. In Neural control of locomotion . Herman, R. M., Grillner, S.,

Stein, P. S. G. & Stuart, D.G. (eds.). pp.203-236. New York: Plenum Press.

Hoyt, D.F. & Taylor C.R. 1981. Gait and the energetic of locomotion in horses. Nature.

292: 239-240.

Jimenez-Ortega, L. 2005. Avian visual perception: Interocular and intraocular transfer

and head-bobbing behaviour in birds. Dissertation at Ruhr-University Bochum,

Germany. Available at: http://deposit.ddb.de/cgi-bin/dokserv?idn=977965430

&dok_var=d1&dok_ext=pdf&filename=977965430.pdf.

Johnson, K.P. & Clayton, D.H. 2000. Nuclear and Mitochondrial Genes Contain Similar

Phylogenetic Signal for Pigeons and Doves (Aves: Columbiformes). Mol.

Phylogenet. Evol. 14: 141-151.

Jones, S.L. & Cornely, J.E. 2002. Vesper Sparrow ( Pooecetes gramineus ). In The Birds

of North America Online . Poole, A. (ed.). Ithaca: Cornell Lab of Ornithology;

Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/624.

Kasprzykowski, Z. 2003. Habitat preferences of foraging Rooks Corvus frugilegus during

the breeding period in the agricultural landscape of eastern Poland. Acta Ornithol .

38: 27–31.

Klicka, J., Voelker, G. & Spellman, G.M. 2005. A molecular phylogenetic analysis of the

'true thrushes' (Aves: Turdinae). Mol. Phylogenet. Evol. 34: 486-500.

Kral, K. 2003. Behavioural-analytical studies of the role of head movements in depth

perception in insects, birds and mammals. Behav. Proc . 64: 1-12. 187 Kram, R., Domingo, A. & Ferris, D.P. 1997. Effect of reduced gravity on the

preferred walk-run transition speed. J. Exp. Biol. 200: 821-826.

Kröger, R.H.H. & Katzir, G. 2008. Comparative anatomy and physiology of vision in

aquatic tetrapods. In Sensory Evolution on the Threshold: Adaptations in

Secondarily Aquatic Vertebrates . Thewissen, J.G.M. & Nummela, S. (eds.). pgs.

121-148. Berkeley and Los Angeles: University of California Press.

Lee, D.V., Bertram, J.E. & Todhunter, R.J. 1999. Acceleration and balance in trotting

dogs. J. Exp. Biol. 202:3565-3573.

Livezey, B.C. 1998. A phylogenetic analysis of the Gruiformes (Aves) based on

morphological characters, with an emphasis on the rails (Rallidae). Philos. Trans.

R. Soc. Lond. B Biol. Sci . 353: 2077–2151.

Maddison, W. P. & D.R. Maddison. 2009. Mesquite: a modular system for

evolutionary analysis. Version 2.72 http://mesquiteproject.org

Marsh, R.L., Ellerby, D.J., Carr, J.A., Henry, H.T. & Buchanan, C.I. 2004. Partitioning

the energetics of walking and running: swinging the limbs is expensive. Science.

303: 80-83.

Martin, G.R. 1984. The visual fields of the Tawny Owl, Strix aluco L. Vision Res. 24:

1739-1751.

Martin, G.R., Jarrett, N., Tovey, P., & White, C.R. 2005. Visual fields in Flamingos:

chick-feeding versus filter-feeding. Naturwissenschaften. 92: 351-354.

Martin, G.R. & Katzir, G., 1994. Visual fields in herons (Ardeidae): Panoramic vision

beneath the bill. Naturwissenschaften , 81: 182–184. 188 Martin, G.R. & Katzir, G. 1999. Visual fields in short-toed eagles, Circaetus gallicus,

and the function of binocularity in birds. Brain Behav. Evolut . 53: 55-66.

Martin, G.R., Rojas, L.M., Ramírez, Y. & McNeil, R. 2004. The eyes of oilbirds

(Steatornis caripensis ): pushing at the limits of sensitivity. Naturwissenschaften .

91: 26-29.

Martin, G.R., Wilson, K., Wild, J.W., Parsons, S., Kubke, M.F. & Corfield, J. 2007. Kiwi

forego vision in the guidance of their nocturnal activities. PLoS One . 2: e198.

Mason, C.F. & MacDonald, S.M. 2004. Distribution of foraging rooks, Corvus

frugilegus, and rookeries in a landscape in eastern England dominated by winter

cereals Folia Zool. 53(2): 179-188.

McGraw, K.J. & Middleton, A.L. 2009. American Goldfinch ( Carduelis tristis ). In The

Birds of North America Online. Poole, A. (ed.). Ithaca: Cornell Lab of

Ornithology; Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/080.

McElroy, E.M., Biknevicius, A.R. & Reilly, S.M. 2004. Mechanics of locomotion in

lizards. J. Morphol. 260: 311.

McMahon, T.A., Valiant, G., & Frederick, E.C. 1987. Groucho running. J. Appl. Physiol .

62: 2326-2337.

Moyle, R.G. 2006. A molecular phylogeny of kingfishers (Alcedinidae) with insights into

early biogeographic history. Auk. 123:487-499.

Muir, G.D., Gosline, J.M. & Steeves, J.D. 1996. Ontogeny of bipedal locomotion:

Walking and running in the chick. J. Physiol. 493: 589-601.

Necker, R. 2007. Head-bobbing of walking birds. J. Comp. Physiol. A. 193: 1177-1183. 189 Nolan, Jr., V., Ketterson, E.D., Cristol, D.A., Rogers, C.M., Clotfelter, E.D., Titus,

R.C., Schoech, S. J. & Snajdr, E. 2002. Dark-eyed Junco ( Junco hyemalis ). In The

Birds of North America Online. Poole, A. (ed.). Ithaca: Cornell Lab of

Ornithology; Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/716.

Oates, D.W. & Principato, J.D. 1994. Genetic variation and differentiation of North

american waterfowl (Anatidae). T. Nebraska Acad. Sci . 21: 127-145

Ogden, L.J. & Stutchbury, B.J. 1994. Hooded Warbler ( Wilsonia citrina ). In The Birds of

North America Online. Poole, A. (ed.). Ithaca: Cornell Lab of Ornithology;

Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/110.

Parchman, A.J., Reilly, S.M. & Biknevicius, A.R. 2003. Whole-body mechanics and gaits

in the Gray Short-tailed Opossum Monodelphis domestica : Integrating patterns of

locomotion in a semi-erect mammal. J. Exp. Biol. 206: 1379-1388.

Patten, M.A. & Fugate, M. 1998. Systematic relationships among the emberizid

sparrows. Auk. 115: 412-424.

Payne, R.B. 2006. Indigo Bunting ( Passerina cyanea ). In The Birds of North America

Online. Poole, A. (ed.). Ithaca: Cornell Lab of Ornithology; Retrieved from the

Birds of North America Online: http://bna.birds.cornell.edu/bna/species/004.

Payne, R.S. 1971. Acoustic location of prey by barn owls ( Tyto alba ). J Exp Biol. 54:

535-573.

Pennycuick, C.J. 1960. The physical basis of astro-navigation in birds: theoretical

considerations. J. Exp. Biol. 37: 573-593. 190 Pettigrew, J.D. 1986. The evolution of binocular vision. In: Visual Neuroscience .

Pettigrew, J.D., Sanderson, K.J. & Levick, W.R. (eds.). pgs. 208-222.Cambridge:

Cambridge University Press.

Pettigrew, J. D. & Konishi, M. 1984. Some observations on the visual system of the

oilbird, Steatornis caripensis . National Geographic Society Research Reports .

16:439–450.

Pitocchelli, J. 1993. Mourning Warbler ( Oporornis philadelphia ). In The Birds of North

America Online. Poole, A. (ed.). Ithaca: Cornell Lab of Ornithology; Retrieved

from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/072.

Pumphrey, R.J. 1948. The theory of the fovea. J. Exp. Biol. 25: 299-312.

Pumphrey, R.J. 1961.Sensory organs: vision. In Biology and Comparative Physiology of

Birds. Marshall, A.J. (ed.). pgs. 55-68. New York and London: Academic Press.

Reilly, S.M. 2000. Locomotion in the Quail ( Coturnix japonica ): the kinematics of

walking and increasing speed. J. Morphol . 243: 173-185.

Riesing, M.J., Kruckenhauser, L., Gamauf, A. & Haring, E. 2003. Molecular phylogeny

of the genus Buteo (Aves: Accipitridae) based on mitochondrial marker

sequences. Mol. Phylogenet. Evol . 27: 328-342.

Roberts, T.J., Marsh, R.L., Weyand, P.G. & Taylor, C.R. 1997. Muscular force in

running Turkeys: The economy of minimizing work. Science. 275: 1113-1115.

Rochon-Duvigneaud, A. 1943. Les yeux et la vision des vertebras . Masson: Paris. 191 Ross, C.F. 2004. The tarsier fovea: functionless vestige or nocturnal adaptation? In

Anthropoid Origins: New Visions . Ross, C.F. & Kay, R.F. (eds.). pgs. 477-537.

New York: Kluwer Academic/Plenum Publishers.

Rubenson, J., Heliams, D.B., Lloyd, D.G., & Fournier, P.A. 2004. Gait selection in the

Ostrich: mechanical and metabolic characteristics of walking and running with

and without an aerial phase. Proc. R. Soc. Lond. B 271: 1091-1099.

Ruina, A., Bertram, J.E.A., & Srinivasan, M. 2005. A collisional model of the energetic

cost of support work qualitatively explains leg sequencing in walking and

galloping, pseudo-elastic leg behavior in running and the walk-to-run transition. J.

Theor. Biol. 237: 170-192.

Schmitt, D. 1999. Compliant walking in primates. J. Zool., Lond. 248: 149-160.

Sheldon, F.H., Jones, C.E. & McCracken, K.G. 2000. Relative Patterns and Rates of

Evolution in Heron Nuclear and Mitochondrial DNA. Mol. Biol. Evol . 17: 437-

450.

Sheldon, F.H., Whittingham, L.A., Moyle, R.G., Slikas, B. & Winkler, D.W. 2005.

Phylogeny of swallows (Aves: Hirundinidae) estimated from nuclear and

mitochondrial DNA sequences. Mol. Phylogenet. Evol. 35: 254-270.

Slonaker, J. R. 1897. A comparative study of the area of acute vision in vertebrates. J.

Morph. 13:445–492.

Spicer, G.S. & Dunipace, L. 2004. Molecular phylogeny of songbirds (Passeriformes)

inferred from mitochondrial 16S ribosomal RNA gene sequences. Mol.

Phylogenet. Evol. 30: 325-335. 192 Tarvin, K.A. & Woolfenden, G.E. 1999. Blue Jay ( Cyanocitta cristata ). In The Birds

of North America Online . Poole, A. (ed.). Ithaca: Cornell Lab of Ornithology;

Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/469.

Thomas, G.H. Wills, M.A. & Székely, T. 2004. A supertree approach to shorebird phylogeny.

BMC Evol. Biol. 4, 28.

Usherwood, J.R. 2005. Why not walk faster? Biol. Lett . 1: 338-341.

Van der Willigen, R.F., Frost, B.J. & Wagner, H. 2002. Depth generalization from stereo

to motion parallax in the owl. J. Comp. Physiol. A. 187: 997-1007.

Van Horn, M.A. & Donovan, T.M. 1994. Ovenbird ( Seiurus aurocapilla ). In The Birds of

North America Online . Poole, A. (ed.). Ithaca: Cornell Lab of Ornithology;

Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/088.

Verbeek, N. A. & Caffrey, C. 2002. American Crow ( Corvus brachyrhynchos ). In The

Birds of North America Online . Poole, A. (ed.). Ithaca: Cornell Lab of

Ornithology; Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/647.

Verstappen, M. & Aerts, P. 2000. Terrestrial locomotion in the Black-Billed Magpie. I.

Spatio-temporal gait characteristics. Motor control . 4: 150-164.

Wainwright, P.S. &. Reilly, S.M (eds.) 1994. Ecological Morphology. University of

Chicago Press, Chicago. 367 pp.

Walls, G.L. 1942. The vertebrate eye and its adaptive radition. New York: Hafner

Publishing Company. 193 Weckstein, J.D., Kroodsma, D.E. & Faucett, R.C. 2002. Fox Sparrow ( Passerella

iliaca ). In The Birds of North America Online . Poole, A. (ed.). Ithaca: Cornell Lab

of Ornithology; Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/715.

Whiteside, T.C.D. 1967. The head movement of walking birds. J Physiol (Lond). 188:

31P-32P

Willey, J.S., Biknevicius, A.R., Reilly, S.M., & Earls, K.D. 2004. The tale of the tail:

Limb function and locomotor mechanics in Alligator mississippiensis . J. Exp.

Biol. 207: 553-563.

Wilson, Jr., W.H. 1996. Palm Warbler ( Dendroica palmarum ). In The Birds of North

America Online . Poole, A. (ed.). Ithaca: Cornell Lab of Ornithology; Retrieved

from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/238.

Wink, M., Heidrich, P. & Fentzloff, C. 1996. A mtDNA phylogeny of sea eagles (genus

Haliaeetus) based on nucleotide sequences of the cytochrome b-gene. Biochem.

Syst. Ecol. 24: 783-791.

Witherby, H.F. 1943. The Handbook of British Birds . Jourdain, F.C.R., Ticehurst, N.F.,

& Tucker, B.W. (eds.) Volume 1. London: H.F. & G. Witherby LTD.

Wohlschläger, A., Jäger, R. & Delius, J.D. 1993. Head and eye movements in

unrestrained pigeons ( Columba livia ). J. Comp. Psychol. 107: 313–319.

Wood, C.A. 1917. The fundus oculi of birds. Lakeside Press: Chicago.

Yasukawa, K. & Searcy, W.A. 1995. Red-winged Blackbird ( Agelaius phoeniceus ). The

Birds of North America Online. Poole, A. (ed.). Ithaca: Cornell Lab of 194 Ornithology; Retrieved from the Birds of North America Online:

http://bna.birds.cornell.edu/bna/species/184.

Yuri, T. & Mindell, D.P. 2002. Molecular phylogenetic analysis of Fringillidae, “New

World nine-primaried oscines” (Aves: Passeriformes). Mol. Phylogenet. Evol. 23:

229-243.