Avian visual perception:

Interocular and intraocular transfer and head-bobbing behaviour in birds

A Dissertation submitted for the Degree of “Philosphiae doctoris” (PhD) in Neuroscience at the International Graduate School of Neuroscience (IGSN) of the RUHR-UNIVERSITY BOCHUM by

Laura Jiménez Ortega

October 2005

Printed with permission of the International Graduate School of Neuroscience of the RUHR-UNIVERSITY BOCHUM

First Referee: Prof. Dr. Nikolaus F. Troje Second Referee: Prof. Dr. Onur Güntürkün Third Referee: Prof. Dr. L. Huber (Wien)

Date of the oral examination: 30-11-2005

Table of contents

1. GENERAL INTRODUCTION...... 1

1.2 Intraocular and interocular transfer in pigeons ...... 2 1.2.1 Intraocular transfer of information ...... 3 1.2.2 Interocular transfer of information...... 5 1.2.3 Interim summary ...... 9

1.3 Visual asymmetries in birds...... 10 1.3.1 Right eye/left hemisphere dominances...... 11 1.3.2 Left eye/right hemisphere dominances ...... 12 1.3.3 Asymmetric interhemispheric transfer...... 14 1.3.4 Interim summary ...... 16

1.4 Head-bobbing in birds ...... 16 1.4.1 Biomechanical function ...... 17 1.4.2 Image stabilization...... 18 1.4.3 Motion parallax...... 19 1.4.4 Head-bobbing birds ...... 20 1.4.5 Interim summary ...... 22

1.5 Anatomical substrate...... 22 1.5.1 The avian eye...... 22 1.5.2 Visual pathways in the avian brain ...... 26 1.5.3 Interim summary ...... 30

1.6 Goals of this work...... 31

2. GENERAL METHODS ...... 32

2.1 Experimental arena ...... 32

2.2 Motion capture system...... 33

2.3 Subjects ...... 34

2.4 Training Procedure ...... 34

3. EXPERIMENT 1, 2 AND 3: INTRAOCULAR AND INTEROCULAR TRANSFER IN PIGEONS...... 38

3.1 Experiment 1: Limits of intraocular transfer in pigeons I: frontal to lateral direction...... 38 3.1.1 Methods ...... 39 3.1.2 Results...... 41 3.1.3 Discussion ...... 46

3.2 Are pigeons capable of interocular transfer between the two yellow fields? ...... 49 3.2.1 Methods ...... 50

ii 3.2.2 Results...... 51 3.2.3 Discussion ...... 53

3.3 Limits of intraocular transfer in pigeons II: lateral to frontal direction...... 56 3.3.1 Methods ...... 56 3.3.2 Results...... 57 3.3.3 Discussion ...... 59

3.4 Interim summary...... 62

4. EXPERIMENT 4: PATTERN RECOGNITION DURING HEAD-BOBBING: ARE PIGEONS CAPABLE OF PATTERN RECOGNITION DURING THE THRUST PHASE? ...... 63

4.1 Methods...... 63

4.2 Results...... 67 4.2.1 Percentage of correct responses...... 67 4.2.2 Head-bobbing motion...... 69 4.3.3 Interim summary ...... 81

5. EXPERIMENT 5: WHY DO BIRDS BOB THEIR HEADS? ...... 82

5.1 Methods...... 83 5.1.1 List of head-bobbing and non head-bobbing birds...... 83 5.1.2 Taxonomic tree of head-bobbing and non head-bobbing birds...... 85 5.1.3 Analysis of behavioural and ecological factors under head-bobbing ...... 87

5.2 Results...... 89 5.2.1 Head-bobbing and non head-bobbing birds list...... 89

5.3 Discussion...... 100 5.3.1 List of head-bobbing and non-bobbing birds: exceptions within a family...... 101 5.3.2 Rare or occasional head-bobbing behaviour ...... 103 5.3.3 Are body-bobbing birds head-bobbing birds? ...... 106 5.3.4 Are other-head-movements functions similar to head-bobbing functions?...... 106 5.3.5 Head-bobbing evolution ...... 107 5.3.6 Ecological and behavioural factors under head-bobbing, body-bobbing and non-bobbing ...... 114 5.3.7 Interim Summary ...... 115

6. GENERAL DISCUSSION ...... 117

6.1 Intraocular and interocular transfer of information...... 117 6.1.1 Intraocular transfer of information ...... 118 6.1.2 Interocular transfer of information...... 120

6.2 Pattern recognition during head-bobbing...... 122

6.3 Why do birds bob their heads? ...... 124

6.4 Summary...... 126

iii APPENDIX...... 128

A. Intraocular and interocular transfer ...... 128

B. Pattern recognition during head-bobbing ...... 132

C. Why birds bob their heads?...... 134

REFERENCES ...... 157

iv ABSTRACT

Two aspects of the avian visual perception were investigated: inter- and intraocular transfer of information in walking pigeons as well as head-bobbing behaviour in birds. The retina of the pigeon has two areas of enhanced vision: the red field pointing into the frontal binocular field and the yellow field projecting into the lateral monocular field. The entire retina projects to the tectofugal pathway, whereas the monocular area projects to the thalamofugal pathway. The first part of this study examines how information received in different retinal areas is generalised in the pigeon brain. The pigeons’ task was to discriminate between two shapes by pecking on one of the two keys located at one end of an experimental alley, while walking between two feeders. In the first study intraocular transfer between the red and the yellow field was tested, by moving the stimulus presentation from the frontal to the lateral visual field in consecutive steps. When the stimuli were located at 45º in the experimental arena, we observed a drastic decrease of performance that may be due to a switch between the tectofugal and the thalamofugal pathway. Intraocular transfer of information was also tested from lateral to frontal direction. The transfer of information was poor or inexistent. Interocular transfer of information between the yellow fields of the eyes was also tested. A lack of interocular transfer was found in eight out of nine birds. Pigeons showed more difficulties to learn the task in the monocular right visual field than in the monocular left visual field.

It is widely accepted that head-bobbing (HB) may act as an optokinetic behaviour, stabilizing the retinal image and allowing pattern recognition during the hold phase. Pattern discrimination during HB was tested by presenting two shapes during the hold phase, the thrust phase, or randomly. The pigeons discriminated the shapes during the entire HB cycle: no differences between the phases were observed. Finally, a list of 322 species of head-bobbing and non- bobbing birds is offered. The development of HB in evolution and its behavioural and ecological characteristics were also investigated. Half of the birds of the world may be head- bobbing birds; the rest may be equally divided between body-bobbing (hopping) and non- bobbing birds. HB may have appeared for the first time in a common ancestor of all living birds. However, evidence for suppressions and several independent evolutions of HB were observed. Head-bobbing may be a mechanism developed early in evolution to solve visual demands, like pattern recognition and to monitor predators, in birds with lateralised eyes.

v General Introduction 1

1. GENERAL INTRODUCTION

Birds are the most visually dependent class of vertebrates (Güntürkün, 2002), as they are highly specialised in visual perception. They live in a wide variety of habitats from desert to rain forest and feed on a wide range of food: seeds, insects, small mammals, carcass, etc. Some birds suffer a strong predator pressure, while others are predators themselves. Birds may need to afford different visual demands that lead to different visual specialisations. The flying ability is probably one of the greatest challenges for the avian visual system. For these reasons avian models have often been used to investigate different aspects of the visual system.

The optic nerves in birds are almost completely decussated (Weidner, Reperant, Miceli, Haby, & Rio, 1985), furthermore the bird’s brain has a limited amount of interhemispheric commissures (Bischof & Watanabe, 1997; Ehrlich & Saleh, 1982; Saleh & Ehrlich, 1984) and a lack of corpus callosum. Due to these unique characteristics, cerebral asymmetries, using the avian visual model, have been broadly investigated. These asymmetries represent an important neural principle in many vertebrates’ brains, including humans (Rogers & Andrew, 2002).

There is much evidence indicating that interhemispheric interactions may be an important component for understanding visual asymmetries (Güntürkün & Bohringer, 1987; Keysers, Diekamp, & Güntürkün, 2000; Parsons & Rogers, 1993). In this context, the study of interocular transfer and intraocular transfer of information in the avian brain may also contribute to a better understanding of visual asymmetries in vertebrates.

Many studies about asymmetries have been done in chickens and pigeons (Güntürkün, 1997a). As a result, we have a better understanding of the visual system of those animals (Zeigler & Bischof, 1993), although there are still many unresolved questions. One open question is the role of head-bobbing in visual perception. Pigeons and chickens are head-bobbing birds, that is, they move the head backward and forward while walking and landing (Dagg, 1977b; Dunlap & Mowrer, 1930).

General Introduction 2

Head-bobbing is present in at least 40% of the birds, furthermore it has been demonstrated that head-bobbing is controlled visually (Friedman, 1975b).The head-bobbing function is not yet well understood, however the contribution of head-bobbing in visual perception is widely accepted (Davies & Green, 1988; Friedman, 1975; Green, Davies, & Thorpe, 1992; Green, Davies, & Thorpe, 1994; Troje & Frost, 1999, 2000).

Very often in inter- and intraocular transfer studies, the possible role of optic flow and head- bobbing in visual perception was not considered. In most cases, the experimental conditions of the studies prevent the birds to show head-bobbing (examples can be found in: Mallin & Delius, 1983 ; Nye, 1973; Remy & Emmerton, 1991b; Roberts, Phelps, Macuda, Brodbeck, & Russ, 1996). In addition, there are few experiments investigating interocular transfer between the yellow visual fields. This is probably due to the difficulties in training pigeons to solve visual tasks in the lateral visual field (Remy & Watanabe, 1993).

The main aim of this work is to combine these two aspects of the visuals system in birds: on the one hand to contribute to a better understanding of information transfer and its asymmetries, and on the other hand to investigate the functional significance of head-bobbing in birds.

1.2 Intraocular and interocular transfer in pigeons

Pigeons and many other birds have lateralised eyes. The Pigeon’s eye is specialised to acute near binocular vision at short distances during pecking and to panoramic vision at long distances (Bloch & Martinoya, 1982a; Goodale, 1983). These two different visual functions are mediated by two separate visual fields that project into two retinal areas: a binocular dorso- temporal red field and a monocular yellow field. The red field is pointing into the lower visual field, while the yellow field is pointing into the upper frontal and lateral visual fields.

In a natural situation a bird might perceive stimuli with both eyes, and also with different areas of the retina. The main goal of inter- and intraocular transfer experiments is to clarify the way

General Introduction 3 in which information retrieved from both eyes and different retinal areas is integrated in birds’ brains (Remy & Watanabe, 1993).

1.2.1 Intraocular transfer of information

Intraocular transfer experiments in birds are infrequent and difficult to interpret (Remy & Watanabe, 1993). Two main limiting factors are responsible for that: first it is difficult to establish the retinal area in which the stimuli are presented in freely moving subjects; second training birds in the lateral visual field has been proved to be an arduous task.

The first attempt to train birds in the lateral visual field was done by Nye (1973). He trained six pigeons in 4 different tasks: “rectangular bar” detection task, “rotating disk of bars” detection task, colour (red and green) and brightness (high and low) discrimination tasks. Birds learned the task successfully when the stimuli were presented behind the pecking keys in the frontal visual field. In contrast, when the stimuli were presented laterally in screens located 90º to each side of the axis of the beak, the response of the birds dropped to chance level. Further attempts to train the birds with stimuli presented in the lateral screens did not succeed. The pigeons performed at chance level after weeks of training in the “bar” and “rotating disk” detection tasks. For colour and brightness discrimination the author trained the pigeons initially in the frontal visual field and moved the screens in 18º azimuth steps to the lateral sides. A progressive drop of percentage of correct responses to chance level occurred when approaching 90º. Nye concluded that “pigeons do not possess the neural capability required to learn to use information contained in laterally located stimuli to directly control pecking behaviour”.

In a previous experiment, Levine (1952) found a drop in the discrimination performance when the stimuli were shifted from a subrostral to an anterorostral position. He interpreted these data as a lack of intraocular transfer in pigeons between two separate functionally independent retinal areas.

General Introduction 4

Mallin and Delius (1983) conducted an intraocular transfer experiment in head fixed pigeons. The birds were trained to discriminate two coloured lights using jaw movements (mandibulando) as an operant response. They found a weak transfer of light colour discrimination from one locus to another within the visual field of the same eye. Interestingly, they also found a poor intraocular discrimination transfer of coloured lights when the stimuli were shifted from frontal to lateral position and vice versa. However, the transfer from lateral to frontal position was slightly better (around 10%) than the reverse performance. These observations were confirmed by Remy and Emmerton (1991b) in head-fixed pigeons, who described the existence of information transfer from the lateral to frontal visual field and observed a lack of information transfer from the frontal to the lateral visual field in a light discrimination task.

Most recently, Roberts et al. (1996) trained eight unrestrained pigeons in an experimental chamber to perform a symbolic delayed matching to sample task. The results confirm previous findings obtained in head-fixed pigeons (Mallin & Delius, 1983; Remy & Emmerton, 1991b), that is, there is intraocular transfer of information from the lateral to the frontal visual field, but not vice versa. They also demonstrated that pigeons are capable of discriminating stimuli in the lateral visual field, in contradiction to Nye’s hypothesis (1973).

Two main suggestions have been proposed to explain intraocular transfer asymmetries. First, Nye (1973) argued that pigeons were incapable of learning a discrimination task that requires a pecking response when the stimuli are presented in the lateral visual field. However, it has been demonstrated that pigeons are capable of stimulus discrimination in the lateral visual field (Bloch & Martinoya, 1982b; Goodale & Graves, 1982; Mallin & Delius, 1983; Remy & Emmerton, 1991b; Roberts et al., 1996). Second, it has been proposed that intraocular transfer of information from the lateral to the frontal visual field may be an ecological advance. On the one hand, birds should be able to switch from lateral to frontal vision when perceiving food and approaching to peck (Friedman, 1975; Remy & Emmerton, 1991b; Roberts et al., 1996). In addition, intraocular transfer from lateral to frontal visual field may be required to monitor the environment for approaching predators while the frontal vision is occupied in feeding (Remy & Emmerton, 1991b; Roberts et al., 1996). On the other hand, there is no need for information

General Introduction 5 transfer from the frontal to the lateral field. Usually objects processed by the lateral visual field are not seen in the frontal field first (Roberts et al., 1996). Field observations suggest that if an object is processed by the lateral visual field, it is often shifted to the frontal visual field but rarely vive versa.

1.2.2 Interocular transfer of information

Interocular transfer has been traditionally studied by covering one eye during the process of learning a visual task. Once the bird reaches a certain criterion of performance, the “naive” eye is tested in the same task (Goodale & Graves, 1982).

According to Levine (1945a), the earliest interocular transfer experiment was done in 1917 by Köhler. He observed that chickens showed interocular transfer in discriminating two sheets of grey paper differing in brightness, which were located horizontally at ground level. In contrast to Köhler’s findings, a lack of interocular transfer was found in pigeons trained monocularly in a go/no-go colour discrimination task. In this experiment the stimuli were displayed on a vertical screen above ground level (Beritov & Chichinadse, 1935).

Levine conducted a set of experiments using a jumping stand in which birds were placed in a rotating perch that forced them to jump onto one of two platform according to different stimuli (colours and shapes discrimination) (Levine, 1945a, 1945b, 1952). If the bird chose the incorrect platform according to the presented stimuli, it collapsed and the animals dropped into a net. If the correct platform was chosen, the birds were allowed to stay there for 15 seconds. He observed that information transfer between the two eyes in pigeons depends on the location of the stimuli in the visual field with reference to the position of the bird’s head. If the stimuli were presented horizontally in a plane below the pigeon’s head (subrostral), interocular transfer was present. In contrast, if the stimuli were presented vertically in front of the pigeon’s head (anterostral) there was an absence of transfer (Levine, 1945a, 1945b, 1952). Two hypotheses were proposed to explain these results: the “sensorimotor integration” hypothesis ( Watanabe,

General Introduction 6

1986) and the “retinal locus” hypothesis (Goodale & Graves, 1982; Levine, 1945b; Mallin & Delius, 1983).

The “sensorimotor integration” hypothesis proposes that pigeons may transfer information depending on whether the response (“manipulandum”) and the stimuli (“stimulandum”) have the same or different spatial locations. When manipulandum and stimulandum share the same spatial locations, for example the stimulus is presented on the surface of the pecking key, interocular transfer of information is expected. If they are located in different spatial location a lack of interocular transfer is predicted (Remy & Watanabe, 1993).

To test the “sensorimotor integration” hypothesis, pigeons were trained in three conditional spatial tasks employing two pecking keys arranged either vertically or horizontally (Watanabe, 1986). No matter whether the keys were arranged horizontally or vertically, if manipulandum (response of the pigeon) and stimulandum were located in the same pecking key, there was a perfect interocular transfer of information. However, if manipulandum and stimulandum were located in different keys (for example: the stimulus was presented in the lower key and the pigeons had to peck in the upper key) the pigeons were incapable of interocular transfer.

The “retinal locus” hypothesis proposes that interocular transfer occurs when the stimuli are presented in the dorso-temporal part of the retina (red field), but not when the stimuli are presented in the other parts of the retina (yellow field) (Goodale & Graves, 1982). The first author supporting this hypothesis was Levine (1952). He proposed that in the pigeon’s eyes there are at least two independent visual areas: a retinal locus corresponding to the subrostral position in the visual field (below the head) and a retinal locus pointing into the anterostral position (in front of the head) of the visual field. He argued that only the subrostral retinal locus has the required neural connections which mediate interhemispheric transfer of information. Furthermore, anterostral-subrostral transfer did not occur within a single eye.

Catania (1965) challenged this hypothesis by training pigeons to peck on a key located in front of the pigeon head, in brightness, colour and pattern discrimination tasks. The stimuli were projected either on the frontal key or on one of two lateral screens. Pigeons showed interocular

General Introduction 7 transfer of information in both conditions. Catania offered two explanations for the lack of interocular transfer in Levine’s experiments. Pigeons are laterally far-sighted and anteriorly near sighted. In the jumping stand, the pigeons have to cock their head to one side in order to observe the stimuli with the lateral visual field. The direction in which the head has to be cocked depends on the covered-eye, and therefore this change of posture may affect interocular transfer of information. Moreover, the amount of training may influence interocular transfer: In Catania’s studies the pigeons might have been over-trained in comparison to the pigeons in Levine’s experiments (Catania, 1965).

A set of experiments replicating Levine’s jumping stand were designed to investigate Catania’s postural hypothesis (Goodale & Graves, 1982). Furthermore, the authors conducted several key-pecking (FR1) discrimination tasks to test training amount and task difficulty as possible factors in interocular transfer. They claimed that the lack of interocular transfer was a genuine phenomenon which did not depend on postural habit, amount of training and task complexity. Furthermore, birds trained binocularly in the jumping stand often showed evidence of learning with only one eye when tested monocularly. They concluded that: “the lack of interocular transfer found in situations such as the jumping stand is a consequence of the discriminative stimuli falling within the monocular field”. Therefore, Goodale and Graves (1982) proposed that interocular transfer occurs when the stimuli were projected into the red field, but not when they were presented to the yellow field

Mallim and Delius (1983) conducted an experiment with head fixed pigeons using jaw movements (mandibulando) as an operant in a colour discrimination task. They presented two coloured lights in different locations of the retina. The advantage of this experimental design is that the spatial localization of the stimuli is separated from the response and the presentation of the stimuli in a certain retinal locus is controlled. Birds showed interocular transfer of information when the discrimination task was monocularly presented inside the red field and a lack of interocular transfer when the stimulus was presented within the yellow field. These results support the “retinal locus hypothesis”, although Remy and Watanabe (1993) pointed out that, in this task, pigeons did not need to direct their response spatially. Furthermore, the

General Introduction 8 pigeon’s beak was oriented towards the position of the stimuli in the frontal position, whereas during the lateral stimulation the beak is oriented in a different direction.

The “retinal locus” and “sensorimotor integration” hypotheses may not be contradictory. Retinal locus may be crucial when a task does not require sensorimotor integration. However, if a task requires sensorimotor integration, interocular transfer will not occur even when the stimuli falls into the binocular field (Remy & Watanabe, 1993).

In addition, some experimental findings suggest that other characteristics of a task, such as biological relevance, may influence interocular transfer. A lack of interocular transfer was found in avoidance of the visual cliff (Zeier, 1970). Transfer of information was absent in heat reinforcement but it was present in a similar task using food reinforcement (Gaston, 1984). Interocular transfer was also observed in taste aversion in chicks (Bell & Gibbs, 1979; Gaston, 1984), as well as in cardiac conditioning in pigeons (Mihara & Watanabe, 1982) and in conditioned withdrawal in pigeons, chickens and gulls (Stevens & Klopfer, 1977).

Furthermore, interocular transfer in pigeons’ colour discrimination but not in motor response training has been found (Stevens & Kirsch, 1980). Pigeons’ eyes were occluded during initial acquisition of the pecking response and subsequently during learning colour discrimination. When the animals were tested with the occluded eye they were unable to respond. Once the pigeons were trained to peck a blank response key, transfer of the colour discrimination task was observed. This is an interesting result, because most of the pigeons’ discrimination experiments involve training of the motor response under binocular condition prior to monocular training on the experimental task. Additional studies of interocular transfer of the motor response would be useful to explain this phenomenon (Remy & Watanabe, 1993).

In a recent publication, interhemispheric transfer of memories was tested in pigeons (Nottelmann, Wohlschlager, & Güntürkün, 2002). Pigeons’ task was to peck on one of two vertical keys according to a pattern displayed on both keys. Six pairs of patterns were used, for half of these pairs the animals had to peck the upper pattern, for the other half the lower one. Transfer of information from the left eye/right hemisphere to the right eye/left hemisphere, but

General Introduction 9 not vice versa was observed. In this experiment, stimuli are presented within the red field and sensorimotor integration occurs, however a lack of interocular transfer is found when the stimuli are presented in the right hemisphere. Therefore, asymmetries in bird brain should be considered as an important factor for interocular transfer. Most probably, asymmetries and interocular transfer of information are interrelated phenomena in birds. In fact, Skiba et al (2000) found asymmetries in interocular transfer of information. A faster shift of learned colour cues from the dominant right to the left eye than vice versa was reported.

Finally, it should be noted that none of the hypotheses is capable of explaining all experimental results in interocular transfer. The “retinal locus hypothesis” together with the current knowledge of the anatomic aspects and asymmetries will lead to a better understanding of the interocular transfer phenomenon.

1.2.3 Interim summary

The main goal of inter- and intraocular transfer experiments is to clarify the way in which information retrieved from both eyes and different retinal areas is integrated in birds’ brains. Interocular transfer experiments test the transfer of information between the two eyes (i.e., between the two hemispheres), whereas intraocular transfer experiments test the transfer of information between different retinal areas of the same eye.

Birds have two distinctive retinal areas: a lateral monocular yellow field and a frontal binocular field, often called the red field. Intraocular transfer has been investigated training the animals to solve a visual task presented in one of these two visual fields and testing in the other. Investigations in non walking pigeons have demonstrated that there is information transfer from the lateral to the frontal visual field. However, a lack of transfer was found from the frontal to the lateral visual field.

In birds, interocular transfer has been traditionally studied by covering one eye during the process of learning a visual task. Once the animal reaches a certain criterion of performance,

General Introduction 10 the “naive” eye is tested in the same task. Interocular transfer depends on the experimental conditions and the task. Two main hypotheses have been proposed to explain the presence or absence of interocular transfer: the “sensorimotor integration” hypothesis and the “retinal locus” hypothesis.

The “sensorimotor integration” hypothesis proposes that pigeons may transfer information if the response and the stimuli share the same spatial locations. In contrast, if they do not share the same spatial location, a lack of interocular transfer will be observed. The “retinal locus” hypothesis proposes that interocular transfer occurs when the stimuli are presented within the red field, but not when the stimuli are presented in the yellow field.

1.3 Visual asymmetries in birds

Birds are highly visual animals, their visual capabilities are the most specialised among vertebrates (Güntürkün, 2003). The optic nerves in birds are almost completely decussated (Weidner et al., 1985). In addition, the amount of thalamic and mesencephalic commissures in birds is very limited (Bischof & Watanabe, 1997; Ehrlich & Saleh, 1982; Saleh & Ehrlich, 1984). Consequently, hemispheric asymmetries can be easily investigated by directing the visual information to one hemisphere by the simple mechanism of temporarily covering one eye.

These characteristics make the avian visual system a very valuable model for the study of visual asymmetries. Many studies in visual asymmetries have been done by using a variety of birds species, like marsh tits (Clayton & Krebs, 1994a), domestic chicks (Andrew, 1988; Andrew & Dharmaretnam, 1993; McKenzie, Andrew, & Jones, 1998; Rogers & Andrew, 2002, Parsons, 1993 #3354), zebra finches (Alonso, 1998; Voss & Bischof, 2003), European starlings (Hart, Partridge, & Cuthill, 2000), quails (Valenti, Sovrano, Zucca, & Vallortigara, 2003) and pigeons (Güntürkün & Bohringer, 1987; Güntürkün et al., 2000; Güntürkün & Hahmann, 1994; Prior & Güntürkün, 2001; Prior et al., 2004; Skiba et al., 2000).

General Introduction 11

1.3.1 Right eye/left hemisphere dominances

Monocular occlusion studies revealed a right eye/left hemisphere dominance in discriminating two-dimensional artificial patterns in pigeons (Güntürkün, 1985) and three dimensional natural objects by means of a grain-grit task in pigeons (Güntürkün & Kesch, 1987), zebra finches (Alonso, 1998) and chickens in a pebble-floor task (Mench & Andrew, 1986). Right eye system dominance is also found in pattern discrimination (Güntürkün, 1997a). A higher degree of illusion is observed for the right eye, when the birds were stimulated with geometrical optic illusions (Güntürkün, 2003).

Some complex studies in pigeons have shown that memories of visual engrams and pattern information are stored unilaterally in the left hemisphere (Güntürkün, 1997a; Nottelmann et al., 2002; von Fersen & Güntürkün, 1990). This asymmetry, in memorizing visual stimuli, most probably results in a right eye/left hemisphere advantage in homing (Ulrich et al., 1999), although when pigeons are tested in circumstances in which they cannot rely on visual cues, like landmarks, the left hemisphere advantage vanishes (Prior, Lingenauber, Nitschke, & Güntürkün, 2002). A right eye dominance is also found in large-scale homing as a consequence of a strong lateralisation of the avian magnetic compass and optic flow processing in favour of the left hemisphere (Prior et al., 2004).

There is evidence for left hemisphere asymmetries in cognitive processes such as “learning-to- learn”. In a colour discrimination reversal learning task birds learned faster with their right eye/left hemisphere than with the left eye/right hemisphere (Diekamp, Prior, & Güntürkün, 1999)

Although, it has been demonstrated that the European starling shows an asymmetric distribution of photoreceptors in the retina, the majority of the asymmetries described are attributed to genuine central processes. They are not due to peripheral factors such as visual acuity (Güntürkün & Hahmann, 1994), wavelength discrimination (Remy & Emmerton, 1991a), or depth resolution (Martinoya, Rivaud, & Bloch, 1983).

General Introduction 12

Due to the amount of evidences in favour of the left hemisphere superiority, accumulated early in the literature, it has been postulated as the dominant hemisphere. At the moment, it is believed that none of the avian hemispheres dominates completely the visual analysis (Güntürkün, 2003). In fact, there is also evidence for right hemisphere superiority in a variety of tasks.

1.3.2 Left eye/right hemisphere dominances

Left eye/right hemisphere dominance has been observed in geometric or spatial information in marsh tits (Clayton & Krebs, 1994b) and chicks (Tommasi & Vallortigara, 2001; Vallortigara, 2000). A right hemisphere dominance is also found for social recognition, like aggressive and sexual responding, in chicks (Rogers & Andrew, 2002; Vallortigara, 1992). Wild living Kookaburras (Dacelo gigas) search for food on the ground using preferably their left eye (Rogers, 2002). In pigeons, a left eye/right hemisphere advance was found in choice reaction times to a patter discrimination task (Di Stefano, Kusmic, & Musumeci, 1987). Most recently, left eye/right hemisphere dominances was observed in the spatial distribution of attention related to food detection for pigeons and chickens (Diekamp, Regolin, Güntürkün, & Vallortigara, 2005).

The opposite left-right specialisation hypothesis for the lateral and frontal visual fields propose that the left eye/right hemisphere may be dominant in the lateral visual field which is mainly focused at long distances, whereas the right eye/left hemisphere may be dominant in binocular vision which is specialised on short distance visual processes (Evans & Evans, 1999; Evans, Evans, & Marier, 1993; Rogers, 2000; Vallortigara, Cozzutti, Tommasi, & Rogers, 2001).

In fact, in the majority of the experiments in visual asymmetries, the stimuli were shown in the frontal binocular visual field, which is mainly analysed by the tectofugal pathway (Güntürkün & Hahmann, 1999; Skiba et al., 2000). However, some evidences of lateral visual field processing have arisen recently. Chickens fixated approaching predators by turning the head

General Introduction 13 abruptly to one side preferably using the left eye (Evans et al., 1993) and showed shorter reaction times using the left eye to detect a novel moving stimuli (a model “raptor”) (Rogers, 2000). In contrast, hens responded to a food call by fixating downward with the frontal visual field. This pattern was not observed when alarm calls and contact calls were presented (Evans & Evans, 1999). Furthermore, Social recognition experiments showed that chickens use either the lateral field of its left eye, or the frontal field of the right eye, before pecking at a stranger, but not at cagemates (Vallortigara et al., 2001).

Having a lateralised brain may allow dual attention to short distance tasks like feeding (using the right eye/left hemisphere system) and long distance tasks like vigilance for predators (left eye/right hemisphere system) (Rogers, 2000).

A recent study with pigeons found a left hemisphere dominance of the thalamofugal visual pathway in a pattern discrimination task in an open arena (Budzynski & Bingman, 2004). The thalamofugal pathway receives information from the lateral visual field, and therefore it is assumed that it may be specialised in far field information processing. Therefore, the left hemisphere dominance in the open arena may contradict the opposite left-right specialisation hypothesis for the lateral and frontal visual fields in chicks. Moreover, the main asymmetry was observed in the visual wulst, a structure that belongs to the thalamofugal pathway but also contributes to the tectofugal pathway (Bagnoli, Grassi, & Magni, 1980; Engelage & Bischof, 1994; Folta, Diekamp, & Güntürkün, 2004; Miceli, Reperant, Villalobos, & Dionne, 1987). However, it should be taken into consideration that the retinal projection of the stimuli was not controlled.

Two more alternatives for explaining hemispheric specialisation have been discussed in the literature. The right eye/ left hemisphere in chicks may be involved in the analysis of novelty and in the spatial configuration of the environment (R.J. Andrew & Dharmaretnam, 1993). This proposal is not in contradiction with the opposite left-right specialisation hypothesis for the lateral and frontal visual fields, indeed a novel stimulus may be perceived first with the lateral visual field.

General Introduction 14

In addition, asymmetries in the bird’s visual system may increase the computational speed of certain processes by concentrating them into one hemisphere. A good example is that visual lateralisation improves grain-grit discrimination success in pigeons (Güntürkün et al., 2000). Furthermore in a double task, like finding food and being vigilant for predators, not-lateralised chicks perform worse than lateralised ones (Rogers, Zucca, & Vallortigara, 2004). Consequently to the ecological advantages of being capable to attend to both predators and feeding source, the computational advantages of processing information in one hemisphere should be added. Most probably the asymmetric avian brain is a consequence of the interaction between the ecological and computational advantages mediated by the anatomical substrate.

1.3.3 Asymmetric interhemispheric transfer

Various studies have investigated asymmetries of transfer between the two hemispheres. Such investigations can give valuable cues to understand interhemispheric transfer, visual asymmetries, and their interactions.

In chickens, a poorer interocular transfer of information from the left eye system to the right eye system than in the opposite direction was described. A passive avoidance task was used, in which birds learned to avoid a bead covered with a bitter substance. Binocular and right lesions in the intermediate hyperstriatum ventrale (IMHV) resulted in amnesia (in monocularly trained birds) when the chicks were tested with the left eye open. On the other hand, left IMHV lesions did not impair performance regardless of the eye used (Sandi, Patterson, & Rose, 1993). Furthermore, lesions in the left IMHV right after training induced amnesia in binocularly trained birds, whereas bilateral ablations of IMHV lesions made one and six hours post- training did not result in amnesia (Patterson, Gilbert, & Rose, 1990). Sandi et al (1993) explained those results as a consequence of a relationship between lateralisation of IMHV function and the visual asymmetries which occur at the behavioural and structural level. They proposed a model in which the memory trace is not fixed into the left IMHV, but it is transferred within one hour to the right IMHV.

General Introduction 15

Another IMHV asymmetry has been found in chicks imprinted with an artificial object. Lesion studies showed that within the first hours after imprinting there is information transfer from left IMHV to the right IMHV which may be responsible for storing the visual characteristics of the imprinted object (Horn, 1991; Nicol, Brown, & Horn, 1995).

An asymmetrical transfer of information has been found in marsh tits for memory storage. Monocular occlusion was used to investigate lateralisation and memory transfer in a food storing and in a one-trial associative learning task. In both cases it was found that both eyes are involved in short-term storage, whereas only the right eye system is responsible for long-term storage. The results indicated that memories are transferred from the right to the left eye system between 3 and 24h after the memory formation. Seven hours after the memory formation, the engram is no longer accessible to the left eye system but has not yet reached the right eye system (Clayton, 1992; Clayton & Krebs, 1994).

In pigeons, it has been demonstrated that each hemisphere shifts colour information to the contralateral side, but the efficiency of the transfer is time and side dependent. There is a faster shift of learned colour cues from the right to the left eye than vice versa within the first 50 minutes after acquisition. For intervals longer than 3 hours, no differences have been found (Skiba et al., 2000). The authors of this experiment concluded that “inter-ocular transfer from the right to the left eye should be facilitated due to a higher bilateral representation of the left- sided tectofugal pathway”.

The interhemispheric asymmetries in food storing, colour discrimination and passive avoidance are not in the same directions. Those differences may arise by the diverse types of cognitive processes required for each task (Skiba et al., 2000). For example, a storing food task demands the utilization of spatial cues which are mainly processed in the right hemisphere, whereas visual cues needed for a pattern discrimination task are processed mainly in the left hemisphere. For giving a coherent explanation of interhemispheric asymmetry patterns, three factors should be considered, functional asymmetries, time course, and physiological characteristics.

General Introduction 16

1.3.4 Interim summary

In birds, right eye/left hemisphere dominance is observed in pattern discrimination, stimulus categorization and memory of visual stimuli. In contrast, left eye/right hemisphere superiority has been found in processing geometric information, social recognition like aggressive and sexual responding. It has been recently proposed that the right eye/left hemisphere may be dominant for short distance tasks like feeding, whereas the left eye/right hemisphere may be specialised in long distance tasks like vigilance. In addition having an asymmetric brain may increase computational speed and allow dual attention to short and long distances.

Asymmetries in the information transfer between the two hemispheres have also been described. In a passive avoidance task a poorer interocular transfer of information from the left eye system to the right eye system than in the opposite direction was found. In Marsh tits, memories are transferred from the right to the left eye system between 3 and 24h after the memory formation. In pigeons, it has been demonstrated that there is a faster shift of learned colour cues from the right to the left eye than vice versa, within the first 50 minutes after acquisition.

1.4 Head-bobbing in birds

Pigeons, chickens, moorhens, partridges, storks, crows, ibises, and many other birds show a characteristic head movement while walking (Dagg, 1977b; Davies & Green, 1988; Dunlap & Mowrer, 1930; Friedman, 1975; Friedman, 1975; Frost, 1978). In pigeons, the head moves backward and forward with respect to the body with a frequency that ranges from about 2 to 10 Hz (Troje & Frost, 2000). Head-bobbing is characterized by a hold phase and a thrust phase (Fig. 2). During the hold phase the head of the bird remains stable in space (Frost, 1978; Troje & Frost, 2000), whereas during the thrust phase the head is moved forward (Fig. 1 and Fig. 2). In pigeons, heead-bobbing movement has been observed during walking, landing flight (Davies & Green, 1991), prior to pecking (Goodale, 1983), and in other behaviours. Even a stationary bird actively observing its environment, shows head-bobbing behaviour.

General Introduction 17

Head-bobbing was first described in 1930 by Dunlap and Mowrer. Since then, three main functions have been proposed, a biomechanical function and two visual functions: image stabilization and depth perception through motion parallax.

Head movement of a walking moorhen

) 1400 s l e x i

p 1200 ( nt e 1000 m e c a

l 800 p s di

l 600 a nt o

z 400 i r o

H 200 00.20.40.60.811.21.4 Ti me (s)

Figure 1: Head position of a walking moorhen (Gallinula chloropus) in freedom. The head position in space was obtained by frame by frame analysis of walking moorhen video recordings. The horizontal coordinate of the head position, given by pixels, was plotted against the time.

1.4.1 Biomechanical function

In walking birds, head-bobbing is synchronized with the motion of the feet (Dagg, 1977a; Dunlap & Mowrer, 1930). For this reason, the first attributed function to head-bobbing was a biomechanical function (Dagg, 1977b). However, later it was shown that head-bobbing is controlled visually and can be elicited independently of active locomotion (Friedman, 1975; Friedman, 1975; Frost, 1978). Although this finding points clearly to a visual function, head- bobbing may also have a biomechanical correlation. Body movements are synchronized with head movements in various behaviours and the stride length of a walking bird is correlated with the relative magnitude of head-bobbing (Fujita, 2004). Troje and Frost (1999) proposed a

General Introduction 18 central pattern generator involved in coordinating complex motion patterns. Eye stabilization may play a role in enhancing equilibrium during walking (Fujita, 2002).

1.4.2 Image stabilization

Some remarkable evidences supporting the visual function of head-bobbing were found in Ring Doves (Friedman, 1975). A single bird was trained to walk inside a cylindrical cage; at least two of the six experimental conditions demonstrated an unequivocal dissociation between walking and head-bobbing. In one condition, the bird walked on a mobile false floor while the cage and the visual surrounding were static. Head-bobbing was not observed in this case. In another experimental condition, the bird remained static on the floor but the cage was moved smoothly in a rostral-to-caudal direction, provoking optic flow. A clear head-bobbing behaviour was observed in this situation. These findings were corroborated in pigeons: head- bobbing was abolished when the animals walking on a treadmill matched the belt velocity, confirming that head-bobbing is visually elicited (Frost, 1978).

Taking in account those results, it was suggested that the hold phase could be similar to other optokinetic behaviours stabilizing the retinal image (Frost, 1978). During the hold phase the head is not completely stabilized, but it slips slightly providing the necessary error signal that drives the compensation mechanism to stabilize the head (Frost, 1978; Troje & Frost, 2000). Stabilizing the retinal image allows object recognition and allows the visual system to distinguish between self motion and outside world motion (Davies & Green, 1988; Frost, 1978; Troje & Frost, 1999). Whooping cranes, during foraging, walk at speeds that permit them to keep their heads immobilized with respect to the visual surroundings while covering large search areas (Cronin, Kinloch, & Olsen, 2005). In spite of the fact that image stabilization is widely accepted as a head-bobbing function, it has been found that in running and landing pigeons the hold phase is replaced with a flexion phase which maintains alternation between two different head velocities. However, in these cases the retraction of the head does not compensate the fast forward movement of the body and the head is not stabilized (Green, 1998;

General Introduction 19

Green et al., 1994). These observations suggest that head-bobbing may have several functions depending on the situations and environmental demands.

1.4.3 Motion parallax

During the thrust phase, pigeons and other birds may use motion parallax computation to derive distances. Pigeons and many other birds like ibises, storks, partridges, chickens, woodcocks etc. have laterally placed eyes with very small binocular fields. Stereo vision as a cue for depth perception can play a role only in a very restricted area. It is therefore assumed that birds use motion parallax to monocularly derive depth information (Green et al., 1994).

Figure 2: Stroboscopic image from a walking pigeon that illustrates the typical head-bobbing which consists of a period in which the head remains fixed in space (hold phase) and a period in which it is quickly moved forward (thrust phase). From: (Frost, 1978).

Several animal species discriminate depth through motion parallax: locust (Collett, 1978; Wallace, 1959), praying mantis (Kral, 1998, 2003; Poteser & Kral, 1995; Poteser, Pabst, & Kral, 1998), barn owl (van der Willigen, Frost, & Wagner, 2002) and gerbils (Goodale, Ellard, & Booth, 1990). Those animals perform head movements called peering in order to generate the necessary optic flow for motion parallax computation. Bees are also capable of calculating depth through motion parallax taking advantage of the ambient optic flow generated during flying (Lehrer, Srinivasan, Zhang, & Horridge, 1988).

Motion parallax computation is based on the fact that a translation of the eye induces a displacement of the retinal image of an object. The translation of the eye ∆xh and the

General Introduction 20

displacement of the retinal image of an object ∆xr are related by the ratio of the focal length f of the eye and the distance d of the object (Fig. 3). ∆x df= ------h- ∆xr

Figure 3: Motion parallax computation where “d” is the distance to the object, “f” is the focal length, “∆xh” is the translation of the eye, and “∆xr” is translation of the retinal image.

The displacement of the eye ∆xh and the corresponding shift of the retinal image ∆xr can well be replaced by the respective velocities vh and vr: v df= ----h- vr

Whereas f is an anatomical constant and vr can be derived directly from the visual input, the velocity of the eye with respect to the visual surroundings vh has to be determined independently. In a walking bird this may be achieved by propioceptive and vestibular information. In praying mantis the propioceptive cervical hair plate sensilla are involved in the measurement of the distance to a jump target with the aid of motion parallax actively produced by translatory head motion (Poteser et al., 1998).

1.4.4 Head-bobbing birds

Although head-bobbing behaviour has been very often discussed in the literature, there exists no comprehensive list about which birds do and which ones do not. Frost (1978) reported that head-bobbing occurs in at least 8 of the 27 orders of birds, such as pigeons, doves, hens, starlings, pheasants, coots, moorhens, rails, sand-pipers, phalaropes, parrots, magpies, and quails. Dagg (1977) listed 28 head-bobbing and 21 non head-bobbing species during locomotion. These lists are rather incomplete and some birds could be misclassified, but they

General Introduction 21 suggest that at least 1/3 of the birds could show head-bobbing. Furthermore, the ecological behavioural and phylogenetic information have been rarely considered in the study of head- bobbing.

To our knowledge, apart from these reports there is a lack of information in the literature about head-bobbing birds. Most of the articles focus on pigeon’s head-bobbing behaviour without retrieving data about other species that could help to clarify its functional significance.

An unanswered question is why some birds bob their heads, whereas other birds walk without bobbing their heads. Furthermore, it is not a clear answer why some species of birds walk with or without displaying head-bobbing. Some birds like magpies can walk with head-bobbing, run or hop. Head-bobbing during walking is used for low velocities, whereas running and out-of- phase hopping are alternative gaits for higher speed in magpies. Furthermore, it is not known why some birds use running and hopping as alternative gaits and why they prefer hopping over running at high speeds (Verstappen, Aerts, & Van Damme, 2000).

Dagg (1977) reported that Mynah birds and starlings alternated between walking with head- bobbing and hopping, depending on the speed of the motion. Birds that walk and bob their heads tend to be of intermediate size. Small birds like most of the Passeriformes, living in bushes and trees, with short legs tend to hop rather than walk (Friedman, 1975).

Hopping behaviour in birds may be comparable to head-bobbing and may play a similar role (Davies & Green, 1988; Friedman, 1975). A frame by frame analysis of hopping sparrows (Passer domesticus) while foraging reveals that “bird’s head is thrust forward before the legs start to push the body into the air. Likewise, the head stops and is stabilized in the visual space before the body finished landing from the hop”. This behaviour is also observed in alert sparrows but not in somnolent ones (Friedman, 1975).

General Introduction 22

1.4.5 Interim summary

Many birds show a characteristic forward and backward head movement while walking, running, and during landing flight called head-bobbing. It is characterized by a hold phase and a thrust phase. Typically, during the hold phase the head of the bird remains stable in space while during the thrust phase the head is moved forward. Three main functions for head- bobbing have been proposed: biomechanical function, image stabilization, and depth perception through motion parallax. Although head-bobbing behaviour has been very often discussed in the literature, most of the birds that bob their heads are not listed. Dagg (1977) and Frost (1978) reported that head-bobbing occurs in at least 8 of the 27 orders of birds and in 28 species such as pigeons, doves, hens, starlings, pheasants, etc. It has been proposed that hopping in birds could be comparable to head-bobbing. It is not known why head-bobbing occurs in some species of birds but not in others. Further investigations are required to investigate the functional and ecological significance of head-bobbing behaviour.

1.5 Anatomical substrate

The avian visual system is composed of two parallel visual pathways that process retinal information from different parts of the retina: the thalamofugal and the tectofugal pathway. In addition the accessory optic system is dedicated to the analysis of optic flow. A comprehensive understanding of the three pathways is important for any attempt to understand the mechanisms underlying visual asymmetries, inter- and intraocular transfer of information, and head bobbing.

1.5.1 The avian eye

Birds have large eyes relative to their body size, suggesting that vision is an important sensory modality in the class aves (Garamszegi, Moller, & Erritzoe, 2002; Martin, 1993). The small tawny owl (450g) has eyes with a greater axial length than humans. The diameter of the ostrich

General Introduction 23 eye is considered to be amongst the largest of all terrestrial vertebrates, with an axial length close to 40 mm (Martin, 1993; Martin, Ashash, & Katzir, 2001). The resolution power of the eye does not depend entirely on its size, other important factors like the structure and concentration of rods and cones on the retina should be considered. Diverse evolutionary demands, as diurnal or nocturnal activity, result into different kind of eyes that vary in the absolute size, position, and amplitude of movement (Martin 1993).

The avian retina like the mammals retina consist on by five layers: the outer nuclear and plexiform layers, the inner nuclear and plexiform layers, and the ganglion cell layer. These layers contain five kinds of cells: photoreceptors, bipolar cells, horizontal cells, amacrine cells and ganglion cells (Fig. 4). The photoreceptors, bipolar cells, and horizontal cells make synaptic contact in the outer retinal layer. The bipolar, amacrine, and ganglion cells make contact in the inner retinal layer (Husband & Shimizu, 2001).

The avian retina shows some interesting differences compared to mammals. It contains no blood vessels; the pecten, a highly vascular structure, is responsible for providing nutrients and oxygen to the cells. Furthermore, it contains double cones, more richer intraretinal connections like horizontal and amacrine cells (Hayes, 1982; Mariani, 1982, 1987), and complex ganglion cell response properties (Pearlman & Hughes, 1976). Four different types of cones have been observed in the avian retina, whereas only three types of cones have been described in the mammalian retina. The spectral sensitivity ranges from ultraviolet (320nm) to the far red (650nm) (Remy & Emmerton, 1991a). The presence of oil-droplets covering the cones add another layer of complexity to the spectral composition of the photoreceptors in the retina.

Figure 4: Avian retina. The photoreceptors, bipolar cells, and horizontal cells make synaptic contact in the outer retinal layer. The bipolar, amacrine, and ganglion cells make contact in the inner retinal layer. Figure from Husband and Shimizu (2001).

General Introduction 24

The distribution of eye droplets in the retina defines two retinal areas (Fig. 5). The red field is characterized by a high concentration of red and orange oil droplets and the yellow field with a bigger concentration of yellow droplets.

RF Figure 5: Schematic representation of the pigeon’s retina (modified from Galifret, 1968), where RF is the red field, YF is the yellow field, P is the pecten and F is the F fovea centralis. YF P

The red field located under the beak in the dorso-temporal retinal quadrant points into the frontal visual field (Bloch & Martinoya, 1982b; Martin & Young, 1983; Martinoya, Rey, & Bloch, 1981). Within the red field most birds also have an area of enhanced vision, the “area dorsalis” which increases the acuity of the frontal binocular visual field (Martin & Katzir, 1999). This area is implicated in close sighting, feeding behaviour and the control of pecking (Goodale, 1983).

The eyes of most birds are aligned laterally (Martin, 1993), which permits birds to receive and process information of the lateral visual field through the yellow field. The lateral field also contains an area of high ganglion cell density called the “fovea centralis”. This lateral visual field serves far sighting, monitoring predators and conspecifics, as well as to detect food at some distance (Fernández-Juricic, Erichsen, & Kacelnik, 2004; Green et al., 1994). Hens tend to view distant objects laterally while the preferentially observe objects less than 20-30 cm away frontally (Dawkins, 2002).

Pigeons and some ground-foraging birds have a localized myopia in the temporal region. It can be explained as an adaptation that permits pigeons to keep the ground in focus while foraging. This localized myopia does not appear in the lateral visual field. In consequence, birds are capable of maintaining in focus panoramic views and therefore monitor relevant information like predators in the lateral visual field while foraging (Hodos & Erichsen, 1990).

General Introduction 25

The “cyclopean area” is the coverage of the total visual field of an animal around the head, that is, the summation of the frontal and lateral visual fields (Martin & Katzir, 1999). Species with large eyes have developed sunshade structures (e.g. eyebrows or eye lash-type feathering) and larger blind areas to minimize sunlight glare (Martin & Katzir, 2000). Large blind areas correspond to smaller cyclopean field, which might be reasonable in large species with low predatory risk. However, species with small eyes generally have smaller blind areas and larger cyclopean fields, because they may need a wider visual field for predators’ detection and are not so strongly affected by sunlight (Fernández-Juricic et al., 2004).

Martin and colleagues classified avian visual fields into 3 basic types and an additional category (combination of two basic classes) (Martin et al., 2001; Martin & Coetzee, 2004; Martin & Katzir, 1993, 1994, 1995, 1999, 2000).

Type 1. “Visual guidance to food items taken in the bill”: the visual projection of the bill tip falls in the centre of the binocular region. The visual field is defined by an extensive cyclopean field, with a long vertical but narrow binocular field. For example, rock pigeon, starling and cattle egret.

Type 2. “Non-visual guidance to food items taken with the bill”: the projection of the bill falls in the periphery of the visual field. A big cyclopean field is also expected with a narrow binocular field, for example, Eurasian woodcock, mallard and teal.

Type 3. “Non-visual guidance to food items taken with the feet”: the projection of the bill fall outside of the visual field. The blind area is relatively large and the binocular field is wide but vertically small, for example tawny owl.

Combination of Types 1 and 3: similar to Type 1 in which individuals visually follow and take mobile prey; but prey is taken with the feet, for example short-toed eagle.

General Introduction 26

1.5.2 Visual pathways in the avian brain

Two main visual pathways process visual information in birds: The thalamofugal pathway and the tectofugal pathway (Fig. 6). The thalamofugal pathway corresponds to the geniculocortical pathway in mammals, whereas the tectofugal pathway corresponds to the extrageniculocortical pathway in mammals. These visual pathways are structurally and functionally independent, although several connections and modulations between them have been described. The accessory optic system in birds is a third independent visual pathway dedicated to optic flow, self motion signals, and optokinetic stimulation processing.

left right Wulst

E

CT+CP TOTO Gld

Rt

RF RF F F YF YF P P

Figure 6: Schematic overview of the thalamofugal (green) and tectofugal pathways (red). Abbreviations: E, entopallium; GLd, nucleus geniculatus lateralis, pars dorsalis; OT, optic tectum; Rt, nucleus rotundus.

General Introduction 27

1.5.2.1 Thalamofugal pathway

In pigeons, but not in chickens, the thalamofugal pathway receives visual input from the yellow visual field (Güntürkün, Miceli, & Watanabe, 1993; Remy & Güntürkün, 1991), which is transmitted to the contralateral thalamic nucleus geniculatus lateralis, pars dorsalis (GLd). The GLd projects bilaterally to the visual wulst, a structure located in the telencephalon, comparable to the striate cortex in mammals (Güntürkün, 2003; Jarvis et al., 2005).

In pigeons, the thalamofugal pathway mainly processes visual input from the lateral monocular fields of the laterally placed eyes (Güntürkün & Hahmann, 1999; Remy & Güntürkün, 1991; Vallortigara et al., 2001). However, in chicks, thalamofugal lesions affect frontal viewing in chicks, suggesting that the thalamofugal system processes frontal visual field information in chicks, but not in pigeons (Deng & Rogers, 2002).

In chicks, the thalamofugal system is asymmetrically structured by means of more contralateral visual projections of the left nucleus geniculatus lateralis, pars dorsalis (GLd), to the right hyperstriatum than vice versa (Deng & Rogers, 2002).

1.5.2.2 Tectofugal pathway

The tectofugal pathway processes visual information proceeding from the entire retina. The visual input ascends from the retina to the contralateral optic tectum (OT), which projects bilaterally to the entopallium (E) via the thalamic nucleus rotundus (Rt). The tectofugal pathway is equivalent to the extrageniculocortical pathway in mammals: the optic tectum corresponds to the superior colliculus, the nucleus rotundus to the lateral posterior-pulvinar, and the entopallium to the extrastriate visual areas of the mammalian brain (Güntürkün, 2003; Jarvis et al., 2005)

Morphological asymmetries have been found in the tectofugal system of the pigeon. In the tectum and the Rt of pigeons, the soma size of visual cells is larger in the left hemisphere

General Introduction 28

(Güntürkün, 1997b; Manns & Güntürkün, 1999). The bilateral projections from the tectal lamina 13 to the Rt lead to representations of both the ipsi- and the contralateral eye in the tectofugal system of each hemisphere (Güntürkün, 2003). These ipsi- to contralareral tectorotundal projections are asymmetric (Fig. 6). On the one hand, the quantity of ipsilateral tectorotundal projections is similar. On the other hand, the number of neurons projecting contralaterally from the right tectum to the left Rt are approximately twice in number than vice versa. Therefore, the Rt on the left side receives a massive ipsilateral tectal input and also a large number of afferents from the contralateral tectum (Güntürkün, Hellmann, Melsbach, & Prior, 1998). Thus, the pigeon’s tectofugal system displays significant morphological asymmetries which might be related to the behavioural lateralisation of the animals. In fact, the left Rt is involved in acuity discrimination with the right and the left eye, whereas the right Rt has minor relevance in participating in binocular acuity (Güntürkün & Hahmann, 1999).

Furthermore, there is evidence of asymmetries in the tectal and posterior commissures connecting the tecta of both hemispheres. Field evoked potential (in response to a stroboscope flash to the contralateral eye) recorded in the left and right tectum showed that the left-to-right tectotectal modulation was more pronounced than vice versa (Keysers et al., 2000).

1.5.2.3 Tectofugal-thalamofugal projections

The tectofugal and thalamofugal pathways are not isolated systems, but they are interconnected by projections from the thalamofugal system onto the tectofugal system and vice versa (Fig. 6 and Fig. 7). The visual wulst sends ipsilateral descending projections directly to the optic tectum (Bagnoli et al., 1980; Karten, Hodos, Nauta, & Revzin, 1973; Miceli et al., 1987). This projection is probably very important for the understanding of the functioning of the avian visual system (Güntürkün et al., 1993). Recently, by recording from single units of the left and right Rt of the tectofugal pathway, a modulation of the left visual wulst on both right and left tectofugal systems has been described, whereas the right visual wulst showed only an ipsilateral influence (Folta et al., 2004, Folta et al. in preparation).

General Introduction 29

Tectum opticum Nucleus rotundus Entopallium RF F YF P Lateral geniculate nucleus V. wulst

Figure 7: Schematic representation of the connections (in black) between the tectofugal (in red) and thalamofugal systems (in green).

Tectofugal-thalamofugal projections have been also described, radial neurons located in the optic tectum project to the dorsolateral thalamus (Gamlin & Cohen, 1986; Wild, 1989). Furthermore, the information processes by both ascending visual pathways converge into the entopallium, thanks to projections from the visual wulst to the entopallium (Husband & Shimizu, 1999; Karten & Hodos, 1970; Shimizu, Cox, & Karten, 1995; Watanabe, Ito, & Ikushima, 1985). In zebra finches, the visual wulst has a significant facilitatory influence on the processing of the contralateral visual information of the entopallium (Engelage & Bischof, 1994).

1.5.2.4 The accessory optic system

In addition to these two ascending visual pathways, the accessory optic system (AOS) is a distinct visual pathway dedicated to the analysis of optic flow fields and various visual signals generated by self-motion or optokinetic stimuli (Simpson, 1984). Given that head-bobbing is triggered by optic flow (Friedman, 1975; Friedman, 1975b), it is widely accepted that the AOS is involved in head-bobbing. Furthermore, AOS is considered to play a role in the stabilization of the retinal image (Simpson, 1984; Westheimer & Blair, 1974), a function attributed to the hold phase of head-bobbing.

General Introduction 30

Numerous electrophysiological studies have shown that neurons in the AOS exhibit direction selectivity in response to large visual stimuli moving in the contralateral visual field (Frost, Wylie, & Wang, 1990). Some neurons have binocular receptive fields that encode optic flow fields induced by self-translation or self-rotation (Frost et al., 1990; Wylie & Frost, 1990; Wylie, Glover, & Aitchison, 1999; Wylie, Glover, & Lau, 1998). In birds, the AOS consists of the nucleus of the basal optic root (nBOR) and the nucleus lentiformis mesencephali (nLM). The nBOR receives input from the retinal displaced ganglion cells, the visual forebrain, the contralateral nBOR and the ipsilateral nucleus lentiformis mesencephali (nLM), and projects to diverse regions including the contralateral nBOR, the ipsilateral nLM, the vestibulocerebellum and the oculomotor complex (Frost & Wylie, 2000; Wang, Gu, & Wang, 2000).

Neurons of the nBOR directly project onto the Rt of the same hemisphere (Diekamp, Hellmann, Troje, Wang, & Güntürkün, 2001). Furthermore, projections from the nBOR onto the GLd have been described (Wylie, Bischof, & Frost, 1998; Wylie, Linkenhoker, & Lau, 1997). The data suggest that the AOS is able to modulate both thalamofugal and tectofugal ascending visual pathways. It is plausible that these projections are necessary to distinguish self- and object-motion processed by the AOS and the ascending pathways, respectively (Diekamp et al., 2001).

1.5.3 Interim summary

The retina of the pigeon has two areas of enhanced vision: the red field in the dorsotemporal retina pointing into the frontal binocular field and the yellow field projecting into the lateral monocular field. The entire retina projects to the contralateral optic tectum and continues via the diencephalic nucleus rotundus to the entopallium (tectofugal pathway). The monocular area also projects to the contralateral geniculate thalamic nucleus and continues to the wulst (thalamofugal pathway). These two different visual systems possibly operate independently in the pigeon’s eye, however they are not isolated. Both pathways converge into the entopallium, the visual wulst sends ipsilateral descending projections directly to the optic tectum, and finally radial neurons located in the optic tectum project to the dorsolateral thalamus.

General Introduction 31

In addition to these two ascending visual pathways, the accessory optic system, consisting of the nucleus of the basal optic root and the nucleus lentiformis mesencephali, is a distinct visual pathway dedicated to the analysis of optic flow fields and various visual signals generated by self-motion or optokinetic stimuli.

1.6 Goals of this work

The main aim of this work is to combine two aspects of the visual system in birds: on the one hand to contribute to a better understanding of information transfer and its asymmetries, and on the other hand to investigate the functional significance of head-bobbing in birds together with its evolutionary basis.

The first part of this work investigates how information perceived in different parts of the pigeons’ retinas, which is processes by two independent visual systems in each hemisphere, is generalised in the pigeons’ brain. To achieve this aim, several experiments were designed. In the first experiment and the third experiment, we tested intraocular transfer of information between the red and the yellow fields in walking pigeons. In the second experiment, interocular transfer of information between the yellow fields of both eyes was investigated.

It is believed that head-bobbing acts as an optokinetic behaviour allowing pattern recognition during the hold phase. The aim of the fourth experiment was to clarify the role of the head- bobbing hold and thrust phases in pattern recognition. Therefore, we conducted an experimental task in which the pigeons needed to discriminate between two stimuli presented either in the thrust phase, the hold phase, or randomly. To our knowledge, although head- bobbing behaviour has been often discussed in the literature, few head-bobbing birds have been listed until now. In the fifth experiment, a comprehensive list of head-bobbing and non head-bobbing birds is offered. Furthermore, field observations, video recordings and phylogenetic information, together with an analysis of behavioural and ecological characters, were combined to clarify the adaptive value of head-bobbing and its evolutionary foundations.

General Methods 32

2. GENERAL METHODS

Many experiments reported until now have been done in static or restrained pigeons in detriment of the ecological validity. Here, we used an experimental arena, which allowed the birds to walk between two feeders. In addition, a motion capture system was mounted in the arena to track the animal’s walking behaviour and head position. In an initial training, the pigeons were trained to discriminate between two stimuli, while walking between the two feeders.

2.1 Experimental arena

An experimental arena of 125 cm length and 54 cm width was constructed (Fig.8), with a feeder and two pecking keys on either end. The two pecking keys (2.5 cm diameter) were horizontally placed at each side of the feeder (2 cm of diameter). The lower edge of each key was 5 cm above the floor, and the two keys were spaced 23 cm apart. Two 15” LCD screens were mounted on a track surrounding the arena, which permitted an easy displacement of the screens around it. On one end of the arena, a light barrier was installed to detect the pigeon walking between the two feeders.

All these devices were placed in a symmetric experimental chamber of 190 cm length, 100 cm width and 80 cm height. On one end of the experimental chamber a camera was placed to observe pigeons’ behaviour and to record the experimental sessions.

General Methods 33

Figure 8: The experimental arena. A track surrounding the experimental arena allowed situating the two LCD screens in any position of the experimental arena.

2.2 Motion capture system

A motion capture system, constructed by the technical service of the Psychology department of the Ruhr-University-Bochum was installed in the arena. Two high speed CMOS cameras (120 frames/s) tracked the movement of a 15mm retroreflective marker (provided by VICON) fixed on the pigeon’s head (Fig. 9). The cameras (MV-D752-80 from Photon Focus) with a spatial resolution of 600 x 440 pixels sent the information to two frame grabber cards (DT 3145 from Data Translation) that digitized the images. Around the lenses, 30 ultra bright red LEDs were mounted, which were flashing in synchrony with the cameras at a temporal resolution of 120 frames/s in order to provide the necessary illumination. The motion capture information was processed and stored in a computer that also controlled the stimulus presentation on the screen, delivered food as a reward or emitted a sharp sound as a punishment.

General Methods 34

Figure 9: Setup of the motion capture system (left) and placement of the marker on the pigeon’s head (right)

2.3 Subjects

Nine pigeons (Columba livia), males and females, between 3 and 7 years of age obtained from the aviary of the biopsychology department (Ruhr-Uni-Bochum) were initially trained in the pattern recognition task. They were kept in individual cages on a 12 hours light-dark cycle. They had ad lib. access to drinking water and grit and were kept at 85% of their free-feeding weight at the beginning of the experiment. Through the training, the birds learned to obtain enough food in the training sessions to keep their weight at a normal level.

2.4 Training Procedure

Pigeons were initially trained in a pattern discrimination task, in which they need to discriminate between two shapes: “stimuli A” and “stimuli B” (Fig. 10), by pecking with a fix rate of 1, on one of the two keys. The stimuli were presented on a single LCD screen located in a central position right behind the feeder with the two pecking keys. The centre of the stimulus was situated in the middle of the screen 16 cm above the floor. The subjects were divided randomly in two groups of 5 pigeons each. One group of birds was trained to peck the right key General Methods 35

when “stimulus A” was presented and the left key when “stimulus B” was presented. For the other group this pattern was reversed, i.e. the left key corresponded to “stimulus A” and the right key to “stimulus B”.

Figure 10: Shapes presented in the discrimination task. From left to right: stimulus A and stimulus B

By a shaping procedure, the pigeons were trained to progressively approximate the target behaviour. First, the pigeons were trained to walk between two near feeders that were alternatively activated during 2 seconds each. The feeders were separated, little by little, throughout the training until they were at either end of the experimental arena (125 cm of distance). When the pigeons were accustomed to walk between the distant feeders the second phase of the shaping procedure started. The birds needed to peck on the keys located on one side of the experimental arena (frontal part) to obtain food within a trial. After two seconds of food reward in the frontal feeder, the pigeons walked to the other end of the arena (back part), where they were rewarded, without pecking, during another 2 seconds. Third, the walking birds had to peck on one of the keys according the shape presented in a screen, which was located in one end of the experimental arena (Fig. 11 and 12).

Stimulus presentation

Pigeon cross the Pigeon walks back Pigeon walks towards light barrier the pecking keys

Reward in the back Figure 11: Diagram of the pigeons’ steps feeder Incorrect Response during the experimental procedure. The pigeons’ task was to discriminate between two shapes by pecking in one of two pecking keys, located at one end of the Pigeon walks towards Correct arena, while walking between two distant the back feeder feeders.

Reward in the frontal feeder

General Methods 36

To force pigeons to walk between the two feeders, the trials had two distinctive phases. The “front phase” started with the presentation of the stimulus on the screen, the pigeon walked towards the frontal pecking keys and feeder. The stimuli were presented on the screen until the pigeons responded by pecking one of the two keys. A correct response was rewarded with 2 s of food access. An incorrect response was punished with a sharp noise lasting 2 s. During the “back phase”, regardless of the response of the pigeons, the animals were forced to walk back towards the other end of the arena (Fig. 8). When they crossed a light barrier located near the feeder, two consequences could follow: if the pigeon’s response was correct, the animal was rewarded with 2 s of food access in the back feeder, afterwards the next trial started. If the response was incorrect, the feeder was not activated and the next trial started with the stimulus presentation (Fig. 11 and 12). Each animal was trained in 4 sessions a day. A session of training consisted of 20 trials presented in random order: 10 “stimulus A” trials and 10 “stimulus B” trials.

Figure 12: A pigeon performing the task in the experimental arena. The animals were trained to peck on one of the two pecking keys according to the stimuli presented in the frontal screen.

Finally, pigeons were accustomed to walking with a reflecting marker (15 mm of diameter) fixed on top of the head. To attach the marker, a piece of Velcro (1 cm long × 2 cm wide) was affixed on the pigeon’s head by means of water-soluble non toxic crafts glue, after clipping the surface of feathers. The counter piece of the Velcro was glued to the marker. General Methods 37

The initial learning took 3-4 months depending on the animal. A smaller experimental arena with mobile walls at the screen level would accelerate the learning process. For experiments 1, 2, and 3 a shorter experimental arena would facilitate the task, although the size was ideal for experiment 4 in which the pigeons' head motions were analysed. A second mobile light barrier would be useful to control the position of the pigeon in the arena at which the stimuli were presented. This would provide an easier way to estimate at which point the pigeons were observing the stimuli. In the present work, the stimuli presentation and the response key were located exclusively in one end of the experimental arena. By presenting the task in both ends of the alley, double amount of data would be collected in each session, but the complexity of the task might be unnecessarily increased. Two animals (not used in these studies) were initially trained using a go-nogo experimental procedure, the animals learned faster to discriminate between the stimuli, but it was harder to keep them walking between the feeders. The overall design of the experimental arena, might be improved for each individual experiment, but it was optimal for the set of experiment that we performed.

Intraocular and Interocular Transfer in Pigeons 38

3. EXPERIMENT 1, 2 AND 3: INTRAOCULAR AND INTEROCULAR TRANSFER IN PIGEONS.

In the first study, transfer of information within the same visual hemisphere was tested. While the screen was gradually moved from its initial frontal position to a lateral position, we measure the time required for retraining to new positions as a function of retinal stimulus position. After this experiment the pigeons were able to perform the pattern discrimination task observing the stimuli on a screen located in one of the lateral visual fields. In the second study, interocular transfer of information between the yellow fields of the two eyes was tested by presenting the stimuli on a screen located in the lateral visual field of the untrained hemisphere. Subsequently six pigeons were trained in the “naive hemisphere”. In the third study, for the animals capable of learning the task, an extension of the intraocular transfer experiment was conducted in which the screen was moved back from lateral to frontal position.

3.1 Experiment 1: Limits of intraocular transfer in pigeons I: frontal to lateral direction.

Intraocular transfer experiments investigate how the information received within different areas of the retina is being generalised in the pigeon visual system. Previous investigations in pigeons observed information transfer from the lateral to the frontal visual field. However, a lack of transfer was found from the frontal to the lateral visual field (Nye, 1973; Mallin & Delius, 1983; Remy & Emmerton, 1991; Roberts et al, 1996).

To our knowledge, intraocular transfer between the red and the yellow visual fields in both directions have been directly investigated in few experiments (Levine, 1952; Mallin & Delius, 1983; Nye, 1973; Remy & Emmerton, 1991b; Roberts et al., 1996). This is probably due to the difficulty of training pigeons to solve visual tasks in the lateral visual field (Remy & Watanabe, 1993) and the hitches to establish the retinal area in which the stimuli are presented in freely moving subjects.

Intraocular and Interocular Transfer in Pigeons 39

Levine’s (1952) jumping stand experiments do not provide direct evidence of switching from the yellow field to the red field and vice versa. To avoid this problem, a couple of investigations were done in head-fixed pigeons in detriment of the ecological validity (Mallin & Delius, 1983; Remy & Emmerton, 1991b). In a later experiment, birds were trained to discriminate stimuli inside a Skinner box, letting the animals free to move, but in practice the animals did not walk (Roberts et al., 1996). In most cases, the experimental conditions of intra- ocular transfer studies prevent the birds to walk (Mallin & Delius, 1983; Nye, 1973; Remy & Emmerton, 1991b; Roberts et al., 1996), which might deprive pigeons from important cues such as visual flow and head-bobbing.

In the current experiment, intraocular transfer of information was tested in freely walking pigeons, by using a pattern recognition task presented in different parts of the visual field.

3.1.1 Methods

Nine of the ten pigeons that completed the initial training were used for testing intraocular transfer of information. The behaviour of the remaining bird was very instable, for this reason it was excluded from the present experiment. The pigeons’ initial task was to discriminate between two shapes (Fig. 10), randomly presented in the frontal visual field, by pecking on one of the two keys located under the screen, while walking between the two feeders (see section 2.4) (Fig. 11). A correct response was rewarded with food; an incorrect response was punished with a sharp tone. The animals were tested during 4 sessions of 20 trials each a day. When a criterion of 70% correct responses in 4 consecutive sessions was reached, the screen was moved step by step to the lateral side (Table 1 and Fig. 13). The pigeons were trained at each position of the screen until achieving the criterion before moving the screen to the next position. The screen was moved according to three rulers located at the frontal and in the lateral sides of the experimental chamber.

Intraocular and Interocular Transfer in Pigeons 40

40 30 20 10 0 10 20 30 40 0 0

10 PHILIPS 10

20 20

30 30 45 degrees 40 40

PHILIPS 50 31cm 50

60 60

70 70

80 80 13 degrees 90 90

100 100

110 110

120 120

130 130

140 140

150 150

160 160

170 170

Figure 13: Scaled drawing of the experimental arena. Three rulers were located in the frontal and the lateral sides of the arena indicating the distance in cm as references to move the screen from the frontal to the lateral positions. Afterwards, the position of the stimuli in degrees was calculated assigning the centre of the circumference in the equidistant point (31 cm) between two screens located in the centre and in the lateral visual field. The bold red line indicates the position of the light barrier. The red angle corresponds to extend of the binocular field of the pigeons, according to Martin and Young (1983), when the birds crossed the light barrier.

Intraocular and Interocular Transfer in Pigeons 41

To calculate the positions of the stimuli in degrees, the centre of the circumference was allocated in the width bisection of the arena at the equidistant point between the frontal and the lateral screen (31 cm to each screen). In summary, the screen was located at 8, 13, 18, 23, 28, and 45 cm according to the frontal ruler, 18, 28, 38, and 48 cm according to the a lateral ruler, which corresponded to 14º, 23º, 29º, 36º, 41º, 45º, 48º, 61º, 76º, and 95º.

Frontal ruler Lateral ruler Stimuli location (in cm) 8 13 18 23 28 45 18 28 38 48 Degrees 14 23 29 36 41 45 48 61 76 95

Table 1: The screen was moved from the frontal side to the lateral side of the experimental arena in 10 consecutive steps according to frontal and lateral rulers located at the sides of the arena.

Nine of the ten birds trained in the initial task achieved the required criteria to move the screen to the sides. For four of the birds, the screen was moved from the centre to the right side of the arena. For three of the birds, the screen was moved from the centre to the left side of the arena (Table 2). The two remaining birds were tested in both sides, dyhue to the difficulties encountered to perform the tasks (see section 3.1.2.1),.

3.1.2 Results

Two different measures were taken in order to study intraocular transfer of information. On the one hand the percentage of correct responses for the first 20 trials at each new screen position, on the other hand the number of trials needed to achieve a criterion of 70% of correct responses in four consecutive sessions were analysed.

3.1.2.1 Percentage of correct responses

Eight of the nine birds learned the task in the lateral visual field by moving the screen in successive steps from 0º (frontal visual field) to 95º (lateral visual field). For pigeons 51, 321, Intraocular and Interocular Transfer in Pigeons 42

333 and 347, the screen was moved to the right visual field (RVF), whereas for pigeons 251, 512 and 988 the screen was moved to the left visual field (LVF).

Due to the experimental results pigeons 246 and 259 were tested on one side and afterwards in the other. For pigeon 246 the screen was moved initially to the left side. The performance of pigeon 246 dropped to chance level (around 50% of percentage of correct responses) when the screen was located farther than 45º (Fig. 14 and Appendix A: Fig. A-1). Even after 600 trials (30 sessions) of training at this position the pigeon did not achieve the criterion. For testing the animal in the other direction, the screen was moved back to 0º where the pigeon was retrained until reaching the criterion. Subsequently the screen was moved step by step to the right side. A drastic drop of behaviour was also observed at 45º and the pigeon did not achieve the criterion after 600 trials of training. In conclusion, pigeon 246 was not able to learn the task beyond 45º in both left and right sides. Pigeon 259 was also tested moving the screen towards right and left. After 600 trials of learning, the bird was incapable of achieving the criterion at 45º on the right side (Fig. 14 and Appendix A: Fig. A-1). However, after 380 trials of training on the left side at 45º, the animal achieved the criterion and the screen was moved step by step until 95º were reached (Fig. 15-16 and Appendix A: Fig. A-1).

Pigeon 259 in the right visual field and pigeon 246 in both visual fields were incapable of learning the task beyond 45º. Pigeons 51, 321, 333, and 347 successfully learned the task in the right visual field. Moreover pigeons 251, 512, 988 and 246 learned also learned successfully the task in the left visual field. Regardless of the individual differences, all birds showed a consistent and drastic decrease of performance at 45º (Fig. 15-16 and Appendix A: Fig. A-1).

For the data analysis, the results of 4 birds for each visual field, including pigeon 259 for the left visual field were considered. At 45º the average percentage of correct responses for the first 20 trials dropped to 58%. At 29º and 36º it was 67% and 68% respectively, while in the remaining positions of the screen the percentage was above 70% of correct responses (Fig. 16).

The design employed for the percentage of correct responses data analysis was a 10*2 mixed ANOVA. The first factor was the within-subject factor of screen position with 10 levels (14º, Intraocular and Interocular Transfer in Pigeons 43

23º, 29º, 36º, 41º, 45º, 48º, 61º, 76º, and 95º). The second factor was the inter-subject factor of visual field with 2 levels. The data analysis revealed a significant effect for screen position

(F(9,54) = 2.55, p=0.01). The post hoc test accounted for significant differences between 45º with positions 14º, 41º, 61º and 76º. Significant effects were also found between position 14º and positions 29º and 36º (Fig. 16 and Appendix A: Fig. 3-A), and between 48º and 76º. The statistic analysis should be interpreted with caution because different levels within the factor screen position may not be statistically independent of each other. If this is the case, the data analysis in the factor screen position could be biased in favour of non significant differences. In comparison, the factor visual field the analysis could be biased in favour of a significant effect. If this is the case we could underestimate the significances at 45º, 29º, and 36º with respect to other positions. No significant effects were found for the visual field (F(1,6)=0.49, p=0.51) an the interaction between the factors (F(9,54)=2.56, p=0.91).

100

90 ses n o 80 esp r

ect 70 259-RVF r r 246-RVF co

f 246-LVF 60 % o

50

40 0 102030405060708090100 Position of the screen (degrees)

Figure 14: Percentage of correct responses for pigeons 259 and 246, which were incapable of learning the task beyond 45º. At 0º, the performance level of the pigeons in the last 20 trials of the initial training.

Intraocular and Interocular Transfer in Pigeons 44

100 251-LVF 259-LVF 90 512-LVF s

se 988-LVF n 51-RVF 80

spo 321-RVF e

r 333-RVF t 70 347-RVF ec r cor 60 of %

50

40 0 102030405060708090100 Position of the screen (degrees)

Figure 15: Percentage of correct responses in the first 20 trials of training for each pigeon at different position of the stimuli. At 0º, the performance level of the pigeons in the last 20 trials of the initial training is plotted.

95 Total 90 LVF RVF 85 s

se 80

espon 75 r ect

r 70

cor 65 of % 60

55

50 0 102030405060708090100 Position of the screen (degrees)

Figure 16: Average percentage of correct responses in the first 20 trials of training at each position of the stimuli in the left visual field (red), right visual field (yellow) and both visual field (blue). At 0º degrees, the average percentage of correct responses of the pigeons in the last 20 trials of the initial training. The bars represent the standard error of the mean. Intraocular and Interocular Transfer in Pigeons 45

3.1.2.2 Number of trials for achieving the criterion

The number of trials needed to achieve the criterion at each position of the screen was also analysed for the eight birds that completed the training. Up to 45º eccentricity, moving the screen to a farther position required an average of 144 trials; while at 45º the animals needed an average of 302 trials to achieve the criterion. The average of trials to move the screen from one position to another was 162, taking in account the training at 45º. Increases of the required number of trials to relearn the task with respect to the average were also observed at 29º, 36º, and 95º (Fig. 17).

420 Total 380 LVF RVF 340

300 ls ia r t

f 260 o r e

b 220 m u

N 180

140

100

60 0 102030405060708090100 Position of the screen (degrees)

Figure 17: Average number of trials needed to achieve the criterion at each position of the screen. The bars represent the standard error of the mean

A data analysis was performed with the number of trials needed to achieve the criterion at each position of the screen. A screen position x visual field ANOVA (10*2) revealed a significant

effect for screen position (F(9,54)=2.13, p=0.04). The post hoc test accounted for significant differences between 45º and positions 14º, 23º, 36º, 41º, 61º and 76º. Significant effects were also found between 95º and positions 14º, 23º and 61º (Fig. 17 and Appendix A: Fig. 4-A). As in the percentage of correct responses analysis, if the screen position levels were not Intraocular and Interocular Transfer in Pigeons 46

independent of each other, the data analysis may underestimate the differences within the

factor position. No significant effects were found for the visual field (F(1,6)=0.32, p=0.59) and the interaction between screen position and visual field (F(9,54)=0.01, p=0.91)

3.1.3 Discussion

Eight of the nine birds learned the discrimination task in the lateral visual field by moving the stimuli step by step from 0º to 90º. This contradict the results of Nye (1975) who found that pigeons were not capable of colour and brightness discrimination in the lateral visual field after moving the stimuli from the frontal to the lateral side in a sequence of 18º steps. Nye concluded that pigeons lack the neural capabilities to learn a task in the lateral visual field. Other evidence showed that pigeons are capable of learning a discrimination task in the lateral visual field (Bloch & Martinoya, 1982b; Goodale & Graves, 1982; Mallin & Delius, 1983; Remy & Emmerton, 1991b; Roberts et al., 1996), however none of those experiments tested freely moving animals while moving the stimuli to the side in consecutive steps. The main differences between Nye’s experiment and the present one are the number of steps used to move the screen to the 90º position and the amount of training at each step. Firstly, in the present experiment the screen was moved at least in 10 consecutive steps while in Nye’s experiment only 5 steps were used. Secondly, Nye trained the birds at each position until they reached a plateau level, whereas in the present experiment the pigeons were trained until they reached the learning criterion (70% of correct responses in four consecutive sessions). The amount of training seems to be the determinant factor for the birds to learn a task in the lateral visual field. In the present experiment pigeons were trained in more steps and probably more intensively within the steps.

The animals showed a significant decrease of performance at 29º, 36º, and especially at 45º degrees. Remarkably, at 45º performance in all pigeons decreased to values below 65%. The average of trials needed to achieve the criterion was 302 trials (15.13 sessions) compared to an average of 144 trials across all stimulus displacement. In addition, Pigeon 246 in both visual fields and pigeons 259 in the RVF did not manage to learn the task beyond 45º. We Intraocular and Interocular Transfer in Pigeons 47

hypothesize that this dramatic decrease of performance can be explained by a switch from perceiving the stimuli with the frontal binocular red field to perceiving the stimuli with the lateral monocular yellow field. The stimulus presentation occurred when the pigeons crossed the light barrier which was located at 118 cm from the frontal screen. Therefore, this distance corresponds to the maximum observation point of a pigeon. The frontal binocular field of the pigeon, in the horizontal plane has a maximum extension of 27º (Martin & Young, 1983). The edge of the frontal binocular field, when the observation point is set at 118 cm in the centre of the arena, corresponds to a stimulus situated at 40º in the experimental arena. A 5x5 cm stimulus presented at 41º in the experimental arena falls at least partially within the binocular field, while a stimulus presented at 45º falls entirely in the lateral visual field (Fig. 13). Therefore, all evidence indicated that at 45º the pigeons did not perceive any longer the stimuli within the frontal visual field. Video recordings confirmed that birds did not twist their heads to look at the stimuli. This discards the possibility that birds solved the task with the frontal visual field by twisting their head, when the stimuli were presented at the sides.

At 45º a failure of intraocular transfer of information may occur and the pigeons required an intensive training until the task was relearned in the lateral visual field. A lack of intraocular transfer from the frontal to the lateral visual field have been previously observed in non walking pigeons (Nye, 1973; Roberts et al., 1996) and head-fixed pigeons (Mallin & Delius, 1983; Remy & Watanabe, 1993). The decrease of performance at 45º degrees could be a consequence of two different visual systems operating independently in the pigeon brain: the thalamofugal system processing information from the lateral visual field and the tectofugal system processing information from the entire retina.

Pigeons 512, 51, and 321 showed the lowest percentage of correct responses at positions 29º and 36º. As a consequence, significant decreases of performance were observed at these positions in the data analysis. Remarkably those birds scored the highest percentage (65%) of correct responses at 45º in comparison to other birds (Appendix A: Fig. A-2). Taking in consideration that the position at which the pigeons look at the screen is not very precisely predetermined, it is plausible that birds 512, 51, and 321 observed the stimuli at closer distances than the other birds. Probably those birds started to perceive the stimuli with the Intraocular and Interocular Transfer in Pigeons 48

yellow visual field around 30º, but instead of learning the task by using the yellow field, they have modified their behaviour to observe the stimuli within the red visual field from a farther position.

A preliminary data analysis of the walking trajectory (motion capture data of 30 trials) of pigeon 251 revealed that the pigeon’s trajectory changed, towards the correct response key at different positions depending on the location of the stimulus. When the screen was located in the lateral visual field the bird changed the trajectory approximately at 20 cm to the pecking keys, while when the screen was located at 45º the bird trajectory changed approximately at 36 cm to the pecking keys (Appendix B: Fig. A-6 and A-7). This might indicate that the stimuli were observed earlier when the screen was located at 45º, than when it was located at 0º.

At 95º significant effects were observed with respect to positions 14º, 23º, and 61º. At this position the screen is located 20 cm away from the pecking keys. Pigeons have a panoramic vision of 245º (Martin & Young, 1983), which corresponds to 122.5º degrees in each lateral visual field. A stimulus located at 95º according to our experimental arena, when a pigeons is situated in the centre of the arena (i.e. 31 cm away from the horizontal plane of the stimulus), falls at 130º with respect to the pigeon eye. This corresponds to the pigeon blind area. Therefore, at this position pigeons need to memorize the response before arriving at the pecking keys. This memory component could be responsible for the difficulties to achieve the criteria at 95º.

Contrary to previous experiments (Mallin & Delius, 1983; Nye, 1973), neither the amount of training nor the percentages of correct responses were affected by the distance between the stimulus and the pecking key. Certainly, significant differences between 48º, 61º, 76º, and 95º were not found in any of the two variables. The effect of the spatial contiguity could be counteracted by the intensive training at each position of the stimuli and by the amount of steps used to move the screen from frontal to lateral. Less intensive training in fewer steps may result in significant differences between the stimuli locations due to the spatial contiguity. Finally, asymmetries between the LVF and the RVF were not observed neither in the percentage of correct responses nor in the amount of trials. Intraocular and Interocular Transfer in Pigeons 49

In conclusion, the pigeons, after an intensive retraining, were capable of learning the task in the lateral visual field. The dramatic decrease of performance observed around 45º may be due to a switch from perceiving the shapes with the frontal field to perceiving the shapes with the lateral field. This lack of intraocular transfer may have originated from the existence of two independent visual systems in the pigeon: the tectofugal and the thalamofugal pathway. Non significant differences were found between the animals trained in the left and the right visual field.

3.2 Are pigeons capable of interocular transfer between the two yellow fields?

Interocular transfer has been widely investigated. A variety of tasks and experimental designs have been used to clarify this phenomenon (see introduction). Although the retinal-locus hypothesis claims that interocular transfer depends on whether the retinal projection of a stimulus falls within the red or the yellow field, few experiments test directly interocular transfer between the two yellow visual fields (Levine, 1952; Martin & Brooke, 1991; Nye, 1973; Remy & Emmerton, 1991b; Roberts et al., 1996). In most experiments, the stimuli were either presented within the binocular red field, or the retinal projection of the stimuli was not controlled. As in intraocular transfer experiments the two main difficulties reside in training pigeons to solve visual tasks in the lateral field (Remy & Watanabe, 1993) and in controlling the retinal projection of the stimuli. Often, the ecological validity has to be sacrificed in order to control the retinal projection of the stimuli.

In the current experiment, we tested interocular transfer of information between the yellow fields in an open arena, in which the pigeons were forced to move between two hoppers to produce optic flow and display head-bobbing.

Intraocular and Interocular Transfer in Pigeons 50

3.2.1 Methods

Eight of the nine pigeons (Columba livia) used in the previous experiment (section 3.1) were tested in interocular transfer of information between the yellow fields.

Similarly to the previous experiment, the pigeons’ initial task was to discriminate between 2 shapes (Fig. 10) presented in the lateral visual field at 90º, while walking between two feeders and pecking on one of the two pecking keys situated at the end of the alley. Four pigeons were trained in the left visual field (LVF), another four in the right visual field (RVF), as described in the previous section. A correct response was rewarded with food. An incorrect response was punished with a sharp tone. In the present experiment, in contrast to the final step of experiment 3.1, the animals performed the task with the stimuli located at 90º. When a criterion of 70% correct responses was reached in two consecutive sessions of 44 trials each, 4 catch trials (no punishment, no reward) presented on the trained side were randomly inserted into each session. When the animals were accustomed to the lack of punishment or reward in 4 of the 48 trials we started with the data collection. At this step, half of the catch trials were presented in the trained visual field and the other half in the naive or untrained visual field. After collecting 10 catch trials in each hemisphere (5 sessions with a total of 20 catch trials), the pigeons were retrained at 90º in the naive hemisphere using the same pair of shapes over 10 experimental days, with 80 trials a day. Six birds continued the training until reaching the criterion or for 5760 trials (72 sessions). Due to the difficulties encountered to train the animals in the RVF, four of the six pigeons were retrained in the RVF and two birds were retrained in the LVF. The initial goal was to train to pigeons to perform the task in each visual field. Pigeons 51 and 347 initially trained in the RVF where retrained in the LVF, while pigeons 988, 51, 512 and 259 initially trained in the LVF were retrained in the RVF (Table 2).

Intraocular and Interocular Transfer in Pigeons 51

Study description Section Frontal RVF LVF Both

Initial Training 2.4 all pigeons

Interocular: frontal-lateral 3.1 51, 321, 251, 259, 333, 347, 512, 988, 246, 259 246 Intraocular: catch trials 3.2.2.1 251, 259, 51, 321, 512, 988 333, 347 Opposite field training: 3.2.2.2 251, 259, 51, 321, 80 trials 512, 988 333, 347 Opposite field training: 3.2.2.3 251, 259 51, 347 5769 trials 512, 988 Interocular: lateral-frontal 3.3 251 51, 347

Frontal test 3.3 259, 512, 988 259

Pattern recognition during HB 4 251, 259, 51, 251, 259, 512, 251, 259, 512, 988, 347 988 512, 988, 51, 347 51, 347 Table 2: History of birds over the series of experiments classified according to the visual field at which the animals were trained in each study.

3.2.2 Results

Three types of data were analysed to investigate interocular transfer of information: the percentage of correct responses for the catch trials, the learning curve of the first 800 trials (10 days of training) in the untrained hemisphere, and the learning curve in six pigeons for 5760 trials (72 days of training) or until reaching the criterion in the untrained hemisphere.

3.2.2.1 Percentage of correct responses for the catch trials

All birds had an average of correct responses above 70% in the learning trials. The average for the LVF was 81.75 and for the RVF 84.5. There were no significant differences between the LVF and the RVF. Regarding the catch trials, all birds except one showed a drop of performance in the untrained hemisphere in comparison to the trained hemisphere. Pigeon 51 is the only exception; it had 80% correct responses in the trained hemisphere and 100% in the untrained hemisphere. The average percentage of correct responses for all pigeons in the trained hemisphere was 85% and 56.25% in the untrained hemisphere (Fig. 18 and Appendix A: Fig. A-5). A hemisphere x visual field mixed ANOVA (2*2) analysis revealed significant Intraocular and Interocular Transfer in Pigeons 52

differences between the trained and the untrained hemisphere (F(1,6)=11.10, p=0.02), and non significant effects for the visual field (F(1,6) =0.69, p=0.44) and the interaction (F(1,6) =0.19, p=0.68).

100 Learning trials Catch trials Learning trials Catch trials Pigeon LVF RVF Trained Naive 90 s

51 82 100 60 e s 85 85 n 80 259 85 80 40 o 82 p 988 75 70 40 70 512 85 80 60 ct res

321 78 100 70 60 347 95 90 60

% of corre 56 251 85 80 100 50 333 80 80 20 Average 81.75 84.5 85 56.25 40 LVF Learning RVF Learning Trained Naive

Figure 18: Percentage of correct responses for each pigeon in 220 learning trials, 10 catch trials in the trained visual field and 10 catch trials in the naive visual field. The average of correct responses in the learning trials (LVF and RVF) and catch trials (trained and naïve hemisphere) is represented in panel.

3.2.2.2 Learning curve for the first 800 trials

Seven of the eight birds scored below the criterion at 70% during the first 10 days of training (800 trials). The percentage of correct responses was around chance level for all birds except for pigeons 251 and 512. Pigeon 251 showed clear signs of interocular transfer and reached the criterion within the third day, with percentages of correct responses above 70%. Pigeon 512 reached levels of correct responses over 65% in the last 3 days of training (Fig. 19). However a close analysis of the video recordings revealed that the bird was occasionally turning the head to the screen which may discard a learning effect in the untrained hemisphere.

3.2.2.3 Learning curve in the untrained hemisphere

Pigeons 51 and 347, trained initially in the RVF, learned the task in the LVF after 5120 (64 days of training) and 2400 trials (30 days of training) respectively. Pigeons 988 and 259 trained initially in the RVF were incapable of learning the task in the LVF after 5760 (72 days of Intraocular and Interocular Transfer in Pigeons 53

training). Another attempt to retrain a bird in the LVF, was made with pigeon 512. Even though this bird reached the required criterion in several sessions, the behavioural results were very unstable (Fig. 20). The analysis of the video recordings of pigeon 512 during the training revealed that this pigeon reached the criterion in sessions during which it was turning the head towards the screen, whereas this behaviour was not observed in any other pigeon.

95

85 ses

n 251-RVF o

sp 75 259-RVF e r

t 512-RVF c e

r 988-RVF r 65

co 51-LVF f

o 321-LVF e 55

tag 333-LVF

en 347-LVF c r e 45 P

35 80 160 240 320 400 480 560 640 720 800 Number of trials

Figure 19: Percentage of correct responses during the first 10 days of training in the untrained hemisphere. Each point represents 80 trials (1 day) of training. Pigeons 251, 259, 512, and 988, trained initially in the LVF, were retrained in the RVF, while pigeons 51, 321, 333 and 347, trained initially in the RVF, were retrained in the LVF

3.2.3 Discussion

Interocular transfer of information was not observed in 7 of the 8 pigeons. The performance level in the untrained hemisphere was around chance level (50%), whereas in the trained hemisphere birds obtained data above the criterion level (70%). In consonance to previous experiments (Goodale & Graves, 1982; Levine, 1945b; Mallin & Delius, 1983), our results confirm a lack of information transfer when the stimuli are perceived within the yellow visual field. Pigeon 251 showed clear interocular transfer of information. We did not observe any Intraocular and Interocular Transfer in Pigeons 54 differences during the experimental procedure that could explain those results. Video recordings of the animal training excluded that pigeon 251 was turning the head to observe the stimuli with other areas of the visual field. In addition, the animal reached the learning criterion after 3 sessions of training in the naive hemisphere. Most probably the bird did not reach the criterion earlier due to an extinction phenomenon after being exposed to catch trials without punishment and reward in the naive hemisphere. Further observations confirmed that pigeon 251 scored a higher percentage of correct responses in the RVF than in the LVF. Those results could be a consequence of individual differences in the organization of the visual system.

95

85

251-RVF es 75 259-RVF 512-RVF spons e 988-RVF r t 65 c 51-LVF e r r 347-LVF o

c 55 of %

45

35 0 500 1000 1500 2000 2500 3000 3500 4000 4500 5000 5500 6000 Number of trials

Figure 20: Learning curve of the originally untrained hemisphere. Each point represents 80 trials (1 day) of training. Pigeons 251, 259, 512, and 988, trained initially in the LVF, were retrained in the RVF, while pigeons 51 and 347, trained initially in the RVF, were retrained in the LVF.

Birds initially trained in the RVF relearned the task in the LVF after a long lasting training, whereas birds initially trained in the LVF (except pigeon 251) did not relearn the task in the RVF at all. Learning to discriminate between two shapes in the lateral visual field while walking between two feeders in an open arena has been demonstrated to be very a demanding task for the pigeons. Long distances between the discriminative stimuli and the response site Intraocular and Interocular Transfer in Pigeons 55

may increase the difficulties of a discrimination task (Watanabe, 1986). In experiment 3.1 we did not find any effect of the spatial contiguity, most probably because in this study the distance between the stimuli and the response key was gradually increased, whereas in the present experiment, the animals were directly trained to discriminate stimuli located 20 cm away from the pecking keys. Pigeons are significantly better discriminating stimuli in their frontal visual field than in their lateral visual field, but experiments with a closer spatial contiguity (Mallin & Delius, 1983; Remy & Watanabe, 1993; Roberts et al., 1996) found fewer difficulties to train the birds in the lateral visual field.

The organization of the avian visual system itself also seems to be responsible for the performance of the pigeons. It has been demonstrated that pigeons detect movements in their lateral visual fields significantly better than in their frontal binocular fields (Bloch & Martinoya, 1982b). The lateral and the frontal visual fields may be specialised in solving different environmental demands. Furthermore, it has been proposed that the right eye/left hemisphere may be dominant for short distance tasks like feeding, while the left eye/right hemisphere could be dominant for long distance tasks like vigilance (Rogers, 2000). This could explain why the birds retrained in the naive RVF did not learn the task while the animals retrained in the naive LVF did. However, stimulus discrimination in pigeons may be a task mainly processed by the tectofugal pathway, which receives information from the entire retina. It may occur that the visual information perceived in the lateral visual field is mainly processed by the thalamofugal pathway, which may need the mediation of the tectofugal pathway to process the information in a pattern discrimination task. More pronounced ipsilateral projections from the right wulst into the right optic tectum than from the left wulst to the left optic tectum have been recently described (Unpublished data from Patzke, Freund, Güntürkün and Manns). If solving a discrimination task in the lateral visual field requires the contribution of the thalamofugal and the tectofugal pathways, this weaker projection between the left visual wulst to the left optic tectum may also explain why the animals retrained in the RVF-left hemisphere did not manage to learn the discrimination task.

Intraocular and Interocular Transfer in Pigeons 56

3.3 Limits of intraocular transfer in pigeons II: lateral to frontal direction.

Several investigations in pigeons have demonstrated a lack of transfer from the frontal visual field to the lateral visual field (Levine, 1945a, 1945b; Mallin & Delius, 1983; Nye, 1973; Remy & Watanabe, 1993; Roberts et al., 1996; Watanabe, 1986). However, information transfer from the lateral to the frontal visual field has been rarely tested. Mallin and Delius (1983) found a higher level of intraocular transfer from lateral to frontal position than vice versa in a colour discrimination task using jaw movements as an operant response. Intraocular transfer from the lateral to frontal visual field was observed in a stimulus detection task in head fixed pigeons (Remy & Emmerton, 1991b) and in a matched to sample task (Roberts et al., 1996).

Intraocular transfer of information was tested from the lateral to the frontal visual field by moving the stimuli to the central visual field in consecutive steps. The pigeons’ task was to discriminate between two shapes by pecking on one of the two keys located at the end of the experimental arena.

3.3.1 Methods

Pigeons 51, 251, and 347, which in the previous experiment learned the task in the initially “naive” hemisphere, were used for testing interocular transfer of information from lateral to frontal direction. The pigeons’ initial task was to discriminate between two shapes (Fig. 10) randomly presented on a screen located at 95 in the lateral visual field, by pecking on one of the two keys located at one end of the experimental alley. The animals were trained to walk between the two feeders while performing the task. Correct responses were rewarded with food; incorrect responses were punished with a sharp tone. The animals were tested during 4 sessions of 20 trials each a day. When a criterion of 70% correct responses in 4 consecutive sessions was reached, the screen was moved step by step to the frontal side. The pigeons were retrained at each position of the screen, until achieving the criterion, before moving the screen Intraocular and Interocular Transfer in Pigeons 57

to the next position. The same stimuli position used in the previous interocular transfer experiment were chosen, except that the screen was moved in the opposite direction. These positions corresponded to 76º, 61º, 48º, 45º, 41º, 36º, 29º, 23º, 14º and 0º. The performance of correct responses for1 the first twenty trials for each positions and the number of trials to achieve the criterion for each animal were analysed. The data obtained when the screen was moved gradually from the lateral to the frontal visual field were compared with the data obtained in experiment 3.1 in which the screen was moved from the frontal to the lateral visual field.

Additionally, as a measure of the learning effect during the initial training, pigeons 988, 259 and 512 were tested in the same task exclusively at 0º. These pigeons shared the same training history as pigeons 51, 251, and 347 (Table 2). Initial training at this position (see section 2.4) in the same task took place for the last time at least one year before for all pigeons. Therefore, testing these birds at 0º may provide a valuable measure of the effect of the initial training in transfer of information from the frontal to the lateral direction.

3.3.2 Results

Moving the screen from the lateral to the frontal visual field (lateral-frontal condition) took a total amount of 800, 900, and 1180 trials for pigeons 251, 51, and 347 respectively. In study 3.1 moving the screen from the frontal to the lateral visual field (frontal-lateral condition) took a total amount of 1260, 1820 and 900 trials for pigeons 251, 51, and 347 respectively (Fig. 21). A one tailed t-test analysis between the number of trials needed at each position to move the screen from frontal to lateral direction and vice versa, confirmed that pigeon 251 needed significantly less sessions of training in the frontal-lateral condition than in the lateral-frontal condition (t=2.29, df=9, p<.05). Similarly, pigeon 51 also needed significantly less sessions of training in the frontal-lateral condition (t=2.29, df=9, p<.05). Pigeon 347 did not show significant differences between the frontal-lateral and lateral-frontal condition (t= 1.02, df = 9, p >.05). Intraocular and Interocular Transfer in Pigeons 58

At position 45º, pigeons 251, 51 and 347 needed 80, 100, and 240 trials in the lateral-frontal condition whereas they needed 280, 400, and 140 trials in the frontal-lateral condition respectively.

The percentage of correct responses in the frontal-lateral condition is similar to the lateral- frontal condition (Fig. 21). No significant differences were found for any of the three pigeons. In both conditions a decrease of performance close to chance level in the first 20 trials was observed at 45º for pigeons 51 and 347, while pigeon 251 showed percentages of correct responses closed to chance level at 61º, 48º and 45º in the lateral-frontal condition and at 39º and 48º in the frontal-lateral condition.

Pigeons 988, 259 and 512 tested at 0º (1 year after the initial training) showed a percentage of correct responses around chance level in the first 400 trials. Pigeons 259 and 512 achieved the training criterion after 640 and 1120 trials. Pigeon 988 reached a percentage of correct responses superior to 70% after 840 trials of training in several sessions. However, due to the instability of the behaviour, the bird never managed to achieve the criterion (70% of correct responses in 4 consecutive sessions) after more than 3000 trials of training. In the initial frontal training at 0º (see section 2.4) birds 988, 259 and 512 required 2820, 920, 1480 trials respectively to achieve the criterion. A Wilcoxon statistic analysis was performed between the numbers of trials needed to achieve the criterion in the frontal initial training and in the one year later frontal retrained. Even assuming that pigeon 988 reached the criterion at 840 trials, no significant differences were found (z=-1.61, N-Ties= 3, p=,11) between both trainings. Bearing in mind that the number of birds tested is very small, we can conclude that the pigeons encountered similar difficulties in learning the task during the initial frontal training and during the later frontal training.

Intraocular and Interocular Transfer in Pigeons 59

100 A B 100

90 90 e e s

80 80 on p pons s s e e r r

t 70

c 70 ect e r r r o cor c 60 60 251-RVF of

of 251-LVF % % 51-RVF 51-LVF 50 50 347-RVF 347-LVF Average Average 40 40 0 102030405060708090100 100 90 80 70 60 50 40 30 20 10 0 Position of the screen (degrees) Position of the screen (degrees)

420 251-LVF 420 C D 251-RVF 380 51-RVF 51-LVF 380 340 347-RVF 347-LVF Average 340 Average ls 300 ls ia 300 r ia t tr f

260 260 f o r o r e

220 e

220 b mb m u 180 180 u N N 140 140

100 100

60 60 0 102030405060708090100 100 90 80 70 60 50 40 30 20 10 0 Position of the screen (degrees) Position of the screen (degrees)

Figure 21: Percentage of correct responses for the first 20 trials at each position for moving the stimuli from frontal to lateral (A) and from lateral to frontal (B) in 10 consecutive steps for pigeons 251, 51, and 347. Number of trials needed to achieve the criterion at each position from frontal to lateral (C) and lateral to frontal (D) for pigeons 251, 51, and 347.

3.3.3 Discussion

Moving the screen from the lateral to the frontal visual field took less training than vice versa for two of the analysed pigeons. The third animal did not show significant differences between the directions. This animal was exceptionally good in the frontal to lateral condition which may explain the lack of differences. In the video recordings, no sign of turning the head towards the screen was observed. Intraocular and Interocular Transfer in Pigeons 60

In two of the three animals, a smaller amount of trials was required to achieve the criterion at 45º when moving the stimuli from lateral to frontal direction than vice versa. Although the percentage of correct response decreases around 45º, the number of trials required to achieve the criterion is comparable to other screen positions for the three animals in the lateral to frontal direction. On the one hand, at 45º pigeons may switch from one visual mechanism to another, which may result in the observed decrease of performance. On the other hand, after this initial decrease of performance, at 45º pigeons needed an amount of trials comparable to other positions in order to achieve the criteria. This result discards the influence of other factors like, visibility problems and orientation of the screen in the performance. Nevertheless, it may not demonstrate a higher level of intraocular transfer of information in lateral to frontal direction. Two factors can explain the better performance of the pigeons in the lateral to frontal direction: pigeons have previous experience solving the same task in the frontal visual field and a closer spatial contiguity between stimuli and response keys may facilitate the task as well.

The initial learning in the frontal visual field took place at least 1 year earlier for all pigeons. To estimate the effect of the frontal initial training in the results, we tested at 0º pigeons 988, 259 and 512 which did not manage to relearn the task in the “naive hemisphere” (see experiment 3.2) at 0º. Surprisingly, during the first 400 trials of training all pigeons performed at chance level and they needed more than 600 trials to achieve the criterion, although they had no problems performing the task with the screen located at 90º. The statistical analysis suggested that most probably pigeons have forgotten the task in the frontal visual field. However, a previous experiment found intraocular transfer of information from lateral to frontal position in a matching to sample task in which the stimuli were presented in a lateral key (at 90º) and on a frontal key (Roberts et al., 1996). Pigeons 988, 259 and 512 were not capable to perform the task neither due to the previous experience in the frontal visual field nor to the presence of intraocular transfer.

These results are somehow contradictory: Three of the birds showed less difficulties solving the task by moving the screen from lateral to frontal direction than vice versa. In addition, we Intraocular and Interocular Transfer in Pigeons 61

found no influence of the previous learning when three other birds were retrained on the frontal visual field. On the one hand one could conclude that there is intraocular transfer of information when the screen is moved from lateral to frontal direction, but on the other hand no traces of intraocular transfer of information where found when retrained three of the birds for the second time at 0º in the frontal visual field

Our data do not provide clear evidence supporting intraocular transfer from the lateral visual field to the frontal visual field. In the frontal-lateral condition the distance between stimuli and response key increases, while in the lateral-frontal condition the distance decreases. A decrease of the distance between stimuli and response key when moving the stimuli from the lateral to the frontal direction could facilitate the task.

In a colour discrimination task in head-fixed pigeon’s similar results were observed (Mallin & Delius, 1983). A poor intraocular transfer when the stimuli were shifted from a frontal to a lateral position and vice versa within the field of view of the same eye. Intraocular transfer was about 10% higher from a lateral to a frontal condition than vice-versa, which could also be as a result of the spatial contiguity between stimuli and response orientation. In contrast, more clear evidence of intraocular transfer from the frontal to the lateral visual field was found in head- fixed pigeons in a stimulus detection task. It occurred when the stimulus presentation was changed from the lateral to the frontal visual field but not vice versa (Remy & Emmerton, 1991b).

The extent of intraocular transfer of information from the lateral to the frontal direction may be task dependent. Remarkably, intraocular transfer was very poor or not observed in tasks with a more manifest memory component, like colour discrimination (Mallin & Delius, 1983) and pattern discrimination. However, intraocular transfer was observed in tasks without a clear memory component, like stimulus detection (Remy & Emmerton, 1991b) and matching to sample (Roberts et al., 1996).

Intraocular and Interocular Transfer in Pigeons 62

The information received within the yellow field system may be transferred into the red visual field system in order to switch attention to interesting or important environmental stimuli, in this case information transfer of complex memories contents may not be necessary.

3.4 Interim summary

A lack of intraocular transfer of information in the same eye from the frontal to the lateral visual field was observed in walking pigeons. Intraocular transfer of information from the lateral to the frontal visual field was poor or inexistent. Transfer of information from the lateral to the frontal visual field may be task dependent. The lack of intraocular transfer can be as a result of two independent visual systems operating in the pigeons’ brain: the tectofugal and the thalamofugal pathways.

Pigeons showed more difficulties to learn a pattern discrimination task directly presented in the right than in the left monocular visual field. The left visual field might be specialised in processing visual tasks that are more likely to be observed at far distances, whereas the right visual field might be specialised in processing visual tasks which are more likely to be observed at close distances. Interocular transfer between the yellow fields of both eyes was rarely observed. It may be due to the poor interhemispheric connections described in pigeons.

Pattern Recognition During Head-bobbing 63

4. EXPERIMENT 4: PATTERN RECOGNITION DURING HEAD-BOBBING: ARE PIGEONS CAPABLE OF PATTERN RECOGNITION DURING THE THRUST PHASE?

It is commonly accepted that head-bobbing acts as an optokinetic behaviour, stabilizing the retinal image and allowing pattern recognition (Davies & Green, 1988; Frost, 1978). The hold phase of head-bobbing in which the retinal image is stabilized may be a determining factor for object recognition. It also may allow birds to distinguish between visual motion due to external movements and self-induced visual motion. The hold phase during head-bobbing is not completely stabilized, but the head moves relative to the surrounding with a velocity that serves as an error signal to stabilize the image (Frost, 1978; Troje & Frost, 2000). This error signal is assumed to be processed by the accessory optic system which plays a central role in the stabilization of the retinal image (Simpson, 1984; Westheimer & Blair, 1974).

According to this theory, pattern recognition would take place mainly during the hold phase. Although this theory is widely accepted (Davies & Green, 1988; Frost, 1978; Morgan & Frost, 1981; Troje & Frost, 2000; Wylie& Frost, 1990) there is no direct behavioural data confirming it.

To investigate the role of the hold phase and thrust phase in pattern recognition, we tested pigeons in a pattern recognition task presenting the stimuli exclusively either in the hold or in the thrust phase.

4.1 Methods

Pigeons 51, 251, 259, 347, 512, and 988 previously utilized in experiments 3.1, 3.2 and 3.3 were used in the present experiment. All of these birds were capable of discriminating between two shapes (Fig. 10) located in the lateral visual field. In each trial, one of the two stimuli was randomly presented simultaneously on the two TFT screens located at 90º in the experimental arena (Fig. 13). Pattern Recognition During Head-bobbing 64

A reflecting marker was attached to the pigeons’ head to capture the head movements of the animals. The pigeons were motion captured in real time (see section 2.2) during the experiment. The temporal resolution of the motion capture system was 120 frames/s and the spatial resolution was 1 mm. The time and x, y, z coordinates of the position of the head trajectory was stored and analysed for each experimental condition.

Each day the pigeons were trained over 96 trials, divided in 4 consecutive sessions of 18 training trials and 6 catch trials each. For the data collection 15 sessions in which pigeons achieved more than 70% of correct responses in the training trials, were considered. Opposed to the previous experiments, the stimuli were presented intermittently for periods of maximally 75ms (see bellow for more details). In two hold catch trials the stimuli were presented exclusively during the hold phase. A hold phase presentation occurred when the speed of the head was smaller than 240 mm/s (Fig. 22C). In two thrust catch trials the stimuli were presented exclusively during the thrust phase. A thrust phase presentation occurred when the speed of the head was bigger than 480 mm/s (Fig. 22D). For two random catch trials and the training trials the stimuli were randomly presented following two distributions of the hold and thrust phase duration (Appendix B: Fig. B-2). Those distributions represents real head-bobbing phases durations. They are calculated to obtain similar presentation patterns to the hold and the thrust presentation, constructed according to very restrictive criteria.

For calculating the distributions that control the random presentation of the stimuli, a total amount of 1907 head-bobbing cycles captured in 10 trials for each pigeon were analysed. A threshold of 240 mm/s was used to differentiate between hold and thrust phases. For each presentation of the stimuli, a virtual head-bobbing cycle was created. For this purpose, a hold and a thrust phase duration was randomly chosen according to the distributions. In one phase of the virtually created head-bobbing cycle, a stimulus was presented for a maximum interval of 75 ms, while during the rest of the time and during the other phase the screen remained black.

Pattern Recognition During Head-bobbing 65

400 A B 500 300 400

200 300

200

) 100 ) m m 100 m m e ( 0 e ( 0 anc anc t t s s -100

Di -100 Di -200

-200 -300

-400 -300 -500

-400 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 Time (s) Time (S)

400 C D 500 300 400

200 300

200 ) ) 100 m m 100 m m e ( e ( 0 0 anc anc t st s -100 Di Di -100 -200

-200 -300

-400 -300 -500

-400 0.8 1 1.2 1.4 1.6 1.8 2 2.2 2.4 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 Time (s) Time (s)

400 E

300 Figure 22: Samples of the head motion of the pigeons. The x coordinate of the head motion was plotted in blue, 200 whereas the y coordinate was plotted in red. The grey fills in figures A and B represent the real time detection 100 ) m of the hold and the thrust phases, respectively. The grey m 0 fills in figures C and D indicate stimulus presentation ance ( st i times during the same trials. Grey fill in figure E stands D -100 for the random presentation in a trial in which the stimulus was not locked to the motion of the head. -200

-300

-400 0.4 0.6 0.8 1 1.2 1.4 1.6 Time (s)

Pattern Recognition During Head-bobbing 66

The delay of the real time motion capture system was measured to be in a range from 8 to 17 ms, due to the refresh rate of the screen (60 Hz) and the temporal resolution of the motion capture system (120 Hz). A maximum presentation time of 75 ms was established for all trials, to avoid presenting the stimuli out of the target phase as a consequence of the possible delay. In courting pigeons the mean durations of the hold and the thrust phases were 156 ms and 132 ms respectively (Troje & Frost, 2000). Initial measures in the present experiment revealed a hold phase duration of around 95 ms and a thrust phase duration of around 100 ms in walking pigeons. Taking into account the hold and thrust phase durations and the delay of the system, a maximum presentation time of 75 ms (9 frames) ensured that the stimuli fell within the target phase (Fig. 22).

In conclusion, to investigate the role of the hold phase and thrust phase in pattern recognition, the pigeons were tested in a pattern discrimination task presenting the stimuli exclusively either during the hold phase, during the thrust phase, or randomly for similarly short periods. First, the six animals were tested until collecting 30 trials in each experimental condition in the lateral visual fields, by a simultaneous presentation of the stimuli on two screens located at 90º on either side. Second, pigeons 988, 259, and 512 were tested exclusively in the right visual field, whereas pigeons 251, 51, and 347 were tested in the left visual field, during 15 sessions (i.e. another 90 catch trials). Finally, all animals were tested in the frontal visual field for another 90 catch trials distributed in 15 sessions; the stimuli were presented in the centre of the experimental arena at 0º. If a pigeon did not achieve a percentage of correct responses in the training trials superior to 70% the session, the session was excluded in the data analysis and the session was repeated.

Pattern Recognition During Head-bobbing 67

4.2 Results

Two dependent variables were analysed. On the one hand, we evaluated the percentage of correct responses in the hold, thrust, and random catch trials. On the other hand, we analysed the head-bobbing motion by calculating the average duration, amplitude, and speed of the hold and thrust phases.

4.2.1 Percentage of correct responses

When the stimuli were presented simultaneously in both lateral visual fields at 90º, the average percentage of correct responses for the hold, thrust, and random presentations were 80%, 77%, and 84%, respectively (Fig. 23). A one way within subjects ANOVA revealed no significant differences (F(2,10)=1.41, p=0.29) between presenting the stimuli exclusively during the hold phase, thrust phase, or randomly.

When the stimuli were presented for half of the birds in the right visual field and for the other half in the left visual field, the average percentage of correct responses for the hold, thrust, and random phases was 71%, 79%, and 77% respectively. The average percentage of correct responses for the hold, thrust and random phase in the animal tested in the left visual field, was 70%, 81%, and 70% respectively, whereas for the right visual field was 74%, 78%, and 85% respectively (Fig. 23). A 3*2 (presentation phase X visual field) mixed ANOVA design was employed to analyse the data. There were no significant effects for presentation phase

(F(2,8)=1.54, p=0.27), or for the visual field (F(2,8)=2.28, p=0.21). Finally, the presentation x

visual field interaction was not significant (F(2,8)=2.07, p=0.19).

In the experimental condition in which the stimuli were presented on a screen located at 0º in the frontal visual field, data from pigeon 988 were not taken into consideration. Pigeon 988 showed a very unstable behaviour solving the task in the frontal visual field, and it was not possible to collect enough sessions in which the percentage of correct responses of the training trials was superior to 70%. Therefore, data from 5 birds were analysed. The percentage of Pattern Recognition During Head-bobbing 68

correct responses for the hold, thrust, and random presentation was 73%, 80%, and 82% respectively (Fig. 23). A one way within-subjects ANOVA showed no significant differences

between hold, thrust, and random presentation (F(2,8)=1.36, p=0.31).

A further 3*3 within-subjects ANOVA analysis was calculated for all birds except for pigeon 988 with factors presentation (levels: hold, thrust, and random presentation) and position (levels: both lateral visual fields presentation, one lateral visual field presentation, and frontal

presentation). The main effect of presentation was not significant (F(2,16)=.33, p=0.73.

Significant effects were found for the position of the stimuli (F(2,16)=5.64 p=0.03). A post hoc analysis revealed significant differences between presenting the stimuli in both lateral visual fields and a single lateral visual field presentation (p=0.02). No significant differences were observed between presenting the stimuli in the frontal visual field and other positions of the screens (p>.11) The interaction between presentation and position of the screen was not

significant (F(4,16)=.54, p=0.71).

100 A 100 B 90 90

80 s 80 s e e 70 70 pons pons 60 60 e e 2 LFs r

r LVF t t c c 50 1 LF 50 e e RVF r r r r 40 FF 40 o o c 30 c 30 of of % 20 % 20 10 10 0 0 Hold% Thrust% Random% Hold% Thrust% Random% Presentation of the stimuli Presentation of the stimuli

Figure 23: Percentage of correct responses for the hold, thrust, and random trials in both lateral visual fields (LFs), one lateral visual field (LF), and in the frontal visual field (FF) (Figure A). Percentage of correct responses for the hold, thrust, and random trials in the left visual field (LVF) and the right visual field (RVF) when the stimuli were presented exclusively on one screen (Figure B).

Pattern Recognition During Head-bobbing 69

4.2.2 Head-bobbing motion

The individual head-bobbing motion of each pigeon during hold, thrust, and random trials were analysed when the stimuli were presented in both lateral fields simultaneously and when they were presented in the frontal visual field. The condition in which the birds observed the stimuli in exclusively in one lateral visual field was excluded in order to reduce the complexity of the data analysis. In addition, during the one lateral visual field presentation the animals did not receive a double signal of the stimuli like in the other two conditions. In addition, initial analysis revealed not significant differences in the motion, between presenting the stimuli in one or two screens.

The head trajectory was divided into head-bobbing cycles. A thrust phase was defined when the speed of the head was equal or bigger than 240 mm/s in 3 or more consecutive frames (25 ms). A hold phase occurred, when the speed of the head was smaller than 240mm/s in 3 or more consecutive frames. For the data analysis only complete head-bobbing cycles were considered. That is, a hold phase is considered if it is followed by a thrust phase. A thrust phase is considered only if it is predated by a hold phase. If a head-bobbing cycle was not complete, or more than two missing frames were observed (for example due to marker occlusion by the pigeon body), it was not considered in the data analysis. If within a phase a single frame was missing or did not fulfil the speed requirements, the phase was not interrupted and was counted in the data analysis. The head-bobbing phases that were bigger than 32 frames (260ms) were not considered in the data analysis.

Finally, for the hold, the thrust and the catch trials presented in the frontal and in both lateral visual field, the duration, the amplitude, and the speed, of the hold and the thrust phase were averaged over 200 head-bobbing cycles obtained from 30 trials in each condition.

Pattern Recognition During Head-bobbing 70

4.2.2.1 Duration of the hold and thrust phases

The average duration for the hold and the thrust phase were calculated for the hold, thrust and random catch trials. When the stimuli were presented simultaneously in both lateral visual fields, the hold phase duration for the hold, thrust, and random catch trials was 93.1 ms, 92.7 ms, and 94.4 ms, respectively. The average thrust phase duration for the hold, thrust, and random catch trials was 102.7 ms, 101.6 ms, and 102.35 ms, respectively. When the stimuli were presented in the frontal visual field, the hold phase duration for the hold, thrust, and random catch trials was 77.8 ms, 79.3 ms, and 76.2 ms, respectively. The thrust phase duration for the hold, thrust, and random trials was 93.6 ms, 94.5 ms, and 93.1 ms, respectively (Fig. 24). The average duration of the hold phase in the frontal and the lateral visual fields was 78 ms and 93 ms, respectively. Whereas, the thrust phase average duration was 93 ms and 102 ms, respectively. The hold phase took place during 45% of the time in the frontal visual field and during 47% of the time in the lateral visual field. The head-bobbing frequency was 5.8 Hz and 5.1 Hz for the frontal and the lateral visual fields, respectively.

A Lateral visual fields B Frontal visual field C 110 110 110 100 100 100 )

) 90 ) 90 90 s s s m m

m Hold phase ( ( 80 ( e e 80 80 e Thrust phase m m m Ti Ti 70 Ti 70 70

60 60 60

50 50 50 hold thrust random hold thrust random lateral frontal Presentation of the stimuli Presentation of the stimuli Visual field

Figure 24: Hold and thrust phase duration for the frontal visual field (A) and the lateral visual fields (B) in the hold, thrust and random catch trials. Total average of the hold and the thrust phase duration in the frontal and in the lateral visual fields (C)

A 2*3*2 (visual field*presentation*phase) within-subject ANOVA was calculated was calculated for 5 pigeons (all birds except 988). The visual field has two levels, lateral visual fields and frontal visual field location of the stimuli. The presentation has three levels, hold phase presentation, thrust phase presentation and random presentation. The phase has two levels, hold phase and thrust phase. Significant differences were found between observing the Pattern Recognition During Head-bobbing 71

stimuli in the frontal visual field and in the lateral visual fields (F(1,4)= 9.65, p=0.03). Therefore, the head-bobbing cycle lasted significantly longer when the stimuli were presented in the lateral visual fields (171.5 ms) than in the frontal visual field (195.95 ms). No significant

differences were observed in presentation (F(2,8)=0.06, p=0.94), phase (F(1,4)=4.12, p=0.11), and in the interactions between the factors (p>0.05).

Individual 2*3*2 (visual field*presentation*phase) analyses of the motion characteristics for each pigeons were also calculated. The phase durations of 30 catch trials in each condition were analysed. Pigeons 251, 347, 51, and 259 showed significant differences (Table 3 and Appendix B: Fig. B-2) in visual field and phase. On the one hand, the head-bobbing cycle was significantly longer when the stimuli were located in the lateral visual fields than in the frontal visual field. On the other hand, the thrust phase duration was significantly longer than the hold phase duration.

Pigeons 251, 347, and 51 showed significant differences in the interaction between visual field*phase (Table 3 and Appendix B: Fig. B-2). For these three pigeons, the difference in the hold phase duration between the lateral and the frontal visual field was significantly bigger than the difference in the thrust phase duration between the lateral and the frontal visual field. That is, the hold phase duration was longer in the lateral visual field than in the frontal visual field. Pigeon 512 showed exclusively significant differences between the hold and thrust phase durations. Any other factor or interaction between the factors was not significant (Table 3, and Appendix B: Fig. B-2).

4.2.2.2 Amplitude of the hold and the thrust phases

The average amplitude for the hold and the thrust phases were calculated for the hold, thrust, and random catch trials. When the stimuli were presented simultaneously in both lateral visual fields, the hold phase amplitude for the hold, thrust, and random catch trials were 2.7 mm, 2.8 mm, and 2.6 mm, respectively. The average thrust phase distances for the hold, thrust, and random catch trials were 80.5 mm, 79.5 mm, and 81.6 mm, respectively. When the stimuli Pattern Recognition During Head-bobbing 72

were presented in the frontal visual field, the hold amplitude distances for the hold, thrust, and random catch trials were 2.2 mm, 2.3 mm, and 2.2 mm, respectively. The thrust phase durations for the hold, thrust, and random trials were 75.4 mm, 78 mm, and 76.7 mm, respectively (Fig. 25).

A Lateral visual fields B Frontal visual field C 90 90 90 80 80 80 70 70 70 ) ) m

m 60 60 60 m m mm) ( ( ( e e e 50 50 50 d d

ud Hold phase u t t i i 40 40 litu l 40 pl p p Thrust phase m m m 30 30 30 A A A 20 20 20 10 10 10 0 0 0 hold thrust random hold thrust random lateral frontal Presentation of the stimuli Presentation of the stimuli Visual field

Figure 25: Hold and thrust phase average amplitude for the frontal visual field (A) and the lateral visual fields (B) in the hold, thrust and random catch trials. Total average of the hold and the thrust phase amplitude in the frontal and in the lateral visual fields (C).

A 2*3*2 (visual field*presentation*phase) within-subject ANOVA was also calculated with the average amplitude in the hold and the thrust phases in five of the pigeons. No significant

differences were observed for factors visual field (F(1,4)=2.57, p=.18) and presentation

(F(1,4)=.18, p=.83). Significant differences were found for factor phase (F(1,4)=813.05, p<.001). Therefore, as expected the hold phase amplitude was significantly smaller than the thrust phase amplitude. No significant effects were found in the interactions between the factors (p>.05)

The individual 2*3*2 (visual field*presentation*phase) ANOVA for each pigeon revealed significant differences in all pigeons for the factor phase (Table 3 and Appendix B: Fig B-2). In

addition, pigeon 251 also showed significant differences in factors visual field (F(1,29)=42.3,

p>.001), presentation (F(1,28)=4.9, p=.01), in visual field*phase interaction (F(1,29)=34.2,

p>.001), and in presentation*phase interaction (F(2,58)=5.98, p=.005).

Pattern Recognition During Head-bobbing 73

4.2.2.3 Speed of the hold and the thrust phases

The average speed for the hold and the thrust phases was calculated for the hold, thrust, and random catch trials. When the stimuli were presented simultaneously in both lateral visual fields, the hold phase speed for the hold, thrust, and random catch trials was 29.9 mm/s, 25.8 mm/s, and 25.4 mm/s, respectively. The average thrust phase speed for the hold, thrust, and random catch trials was 810.4 mm/s, 797.5 mm/s, and 805.5 mm/s, respectively. When the stimuli were presented simultaneously in the frontal visual field, the hold phase speed for the hold, thrust and random catch trials was 30.1 mm/s, 30 mm/s, and 30.2 mm/s, respectively. The average thrust phase speed for the hold, thrust, and random catch trials was 823.9 mm, 825.75 mm/s, and 806.5 mm/s, respectively (Fig. 26).

A B C Lateral visual fields Frontal visual field 900 900 900 800 800 800 700 700 700 )

s 600 / s)

S) 600

/ 600 / m m m hold phase 500 m 500 m m 500 ( ( (

d thrust phase d d

e 400 e 400 400 e e ee

p 300 Sp Sp 300 300 S 200 200 200 100 100 100 0 0 0 hold thrust random hold thrust random lateral frontal Visual field Visual field Presentation of the stimuli

Figure 26: Hold and thrust phase average speed for the frontal visual field (A) and the lateral visual fields (B) in the hold, thrust and random catch trials. Total average of the hold and the thrust phase amplitude in the frontal and in the lateral visual fields (C).

A 2*3*2 (visual field*presentation*phase) within-subject ANOVA was calculated with the average speed for each pigeon in the hold and the thrust phases in five of the pigeons. No

significant differences were observed in factors visual field (F(1,4)= 2.01, p=.23) and

presentation (F(1,4)=.82, p=.47). Significant differences were found in factor phase

(F(1,4)=586.8, p<.001). No significant effects were found in the interactions between the factors (p>.05).

Pattern Recognition During Head-bobbing 74

The analysis of the individual motion of the pigeons revealed that the speed in the frontal field presentation was significantly bigger than the speed in the lateral visual fields presentation for pigeons 51, 259, and 347 (Fig. 27) The speed of the pigeon during the random catch trials was significantly smaller than in the thrust and hold catch trials for pigeons 347 and 512. Significant differences in the thrust phase speed between the random trials and both the hold and thrust phase trials were observed in pigeons 259, 347, and 512. No differences were observed between the hold presentation and the thrust presentation.

500 p <.05

p <.05 p >.05 p >.05 Figure 27: Head-bobbing cycle speed for each 450 individual pigeon, when the stimulus were presented

) either simultaneously in both lateral visual fields or s

/ p <.05 400 m in the frontal visual field m (

d

ee 350

sp 300

250 51 251 259 347 512 lateral visual fields Pigeon frontal visual field

In addition, the data analysis for pigeon 347 found significant differences in all interactions. Pigeon 512 also showed significant differences in the interactions screen*presentation and screen*presentation*phase. Finally, in pigeon 259, significant effects were observed in the interactions also in presentation*phase.

In conclusion, significant differences were found between the hold and thrust phases in all analysed parameters. The head-bobbing cycle duration was significantly longer in the lateral visual fields than in the frontal visual field in three birds. This effect is mainly explained by a longer hold phase duration in the lateral visual field than in the frontal visual field condition. The overall speed of a head-bobbing cycle, and therefore the speed of the motion, was significantly larger in the frontal visual field than in the lateral visual field. In some animals significant differences were found in the speed, between the random trials and both thrust and Pattern Recognition During Head-bobbing 75 hold trials. No significant differences were found in duration, distance, and speed between presenting the stimuli during the hold phase and during thrust phase.

Significant effects ( p<0.05 ) Visual Presentation Phase (Ph) Vf.*Pr. Vf.*Ph. Pr.*Ph. Vf.*Pr.*Ph field (Vf) (Pr) 51,251, 51,251, 51, 251, Duration 259,347, 259,347, 347 all 512 251 251 51,251, 251 Amplitude 259,347, 512, all 51,259, 347,512 51,251, 347,512 347 259,347, 347,512 Speed 347 259,347, 512 512, all

Table 3: Names of the birds that showed significant effects (p<0.05) observed in the ANOVA for duration, distances, and speed in the different experimental conditions.

4.3 Discussion

Pattern recognition in pigeons was studied by presenting two shapes either during the hold phase, thrust phase, and random presentation. The pigeons were tested in both lateral visual fields (90º), a single lateral visual field (90º), and the frontal visual field (0º). Both the percentages of correct responses were analysed and the motions of the pigeons’ heads were analysed. For the hold and the thrust phase, the duration, the amplitude, and the speed were calculated in the frontal and in both lateral visual fields presentation.

4.3.1 Pattern recognition during the thrust phase

During the head-bobbing hold phase, the head is stabilized in the space (Davies & Green, 1988; Friedman, 1975b; Frost, 1978; Troje & Frost, 2000) and saccadic eye movements have rarely been observed (Bloch, Lemeignan, & Martinoya, 1987; Pratt, 1982; Wohlschläger, Jäger, & Delius, 1993). Therefore, it is believed that head-bobbing may aid in pattern recognition by stabilizing the retinal image (Davies & Green, 1988; Frost, 1978). Furthermore, Pattern Recognition During Head-bobbing 76

it has been observed that the thrust phase, where the head is moving forward, is accompanied by saccades. It has been proposed that during the thrust phase there is probably no visual input and the sensitivity of the visual system is reduced by saccadic suppression (Brooks & Holden, 1973; Pratt, 1982). For this reason, we expected higher difficulties in solving the discrimination task when the stimulus presentation occurred within the thrust phase.

Surprisingly, such differences were not found between presenting the stimuli during the hold phase, thrust phase, or randomly within the percentage of correct responses. Contrary to our initial predictions, the animals were able to discriminate the stimuli when they were presented within the thrust phase as successfully as when presented within the hold phase, or randomly. Given that pigeons’ eyes are located laterally, the amount of optic flow perceived during the thrust phase when a stimulus is located in the lateral visual field must be greater than the optic flow perceived in the frontal visual field. Therefore, we expected advantages in perceiving a stimulus during the thrust phase in the frontal visual field with respect to the lateral visual field. No significant differences between observing the stimuli in the frontal visual field and the lateral visual field were found during the thrust phase. The presentation of the stimuli within the thrust phase was strictly controlled.

Significant effects were observed between the different conditions of presentation in one lateral visual field or simultaneously in both lateral visual fields. Taking into account that the pigeons were trained in both visual fields, the double presentation might have enhanced the perception of the stimuli and facilitated the task. No significant asymmetries were observed between the left and the right visual fields.

The data provide solid evidence to affirm that pigeons are capable of pattern discrimination during the trust phase. Two factors may help to explain that: the existence of saccadic eyes movements during the thrust phase and the characteristics of the stimuli.

Pattern Recognition During Head-bobbing 77

4.3.1.1 Eye movements during the thrust phase

Pratt (1982) found that saccadic eye movements in pigeons were present in around 80% of the thrust phases, and were not made during hold phases. He proposed that the saccadic thrust phases have no particular visual function, other than to move the eye to a new viewpoint. He also suggested that the remaining 20% of thrust phases without saccades might serve to calculate depth through motion parallax. Other authors observed three kinds of saccades in unrestrained pigeons: embedded saccades in the thrust phase (86%), overlapping saccades that started earlier or finished later than the thrust phase (7.2%), and saccades occurring during the hold phase period (6.8%) (Wohlschläger et al., 1993). The possibility that pattern recognition occurred during the thrust phases without saccadic eyes movements is very unlikely. Pigeons performed 10 head-bobbing cycles on average while walking form the light barrier to the feeders. Therefore, in this period of time a maximum of two thrust phases without saccades may occur. In addition the stimuli were presented exclusively during 75 ms within the thrust phase.

A pigeons’ oculomotor strategy depends on the visual field. An attention-calling stimulus presented in the lateral field of one eye elicits an ipsilateral orienting saccade towards the target which only sometimes coincides with the movement of the contralateral eye. In contrast, an attention-calling stimulus presented in the frontal binocular visual field elicits coordinated converging saccades (Bloch et al., 1987).

On the one hand, we found that pigeons are capable of pattern discrimination during the thrust phase. On the other hand saccadic eye movements have been observed in pigeons exclusively during the thrust phase during walking (Bloch, Jaeger, Lemeignan, & Martinoya, 1988; Pratt, 1982). The possibility that the saccadic eye movements may help to stabilize the retinal image during the thrust phase should be taken into consideration. In fact backward and forward saccadic movements, lasting around 22.8 ms, were observed in pigeons (Wohlschläger et al., 1993). Although most of the saccades may serve to monocularly fixate visual features (Bloch et al., 1988; Friedman, 1975; Wohlschläger et al., 1993), a fraction of the backward saccades Pattern Recognition During Head-bobbing 78 may be used to maintain fixation by compensating the retinal image shift due to the forward head thrust (Wohlschläger et al., 1993).

4.3.1.2 Characteristic of the stimuli

Another possible explanation for the capability of the pigeons to discriminate stimuli within the thrust phase could be the characteristic of the stimuli used. Although the difficulty level of the task was very high, the patterns presented were easy to discriminate. They were two 5*5 cm white shapes on a black background. It may well happen that, small objects, like seeds and grains, are more difficult to discriminate during the thrust phase. Head-bobbing birds may have difficulties to discriminate small features hidden in a low contrast complex background, like a grain hidden in the yellow grass. In humans saccades are very fast at speeds up to 900º/s, in order to reduce the image blur to a minimum (Land, 1999), whereas the thrust phase in birds is much slower, which may allow pattern recognition. In addition, the hold phase can be substituted by a flexion phase in which the retinal image is no longer stabilized. Head-bobbing should be involved in other important functions. Motion parallax is a strong candidate, but most probably head-bobbing has some other unexplored functions in different birds.

The lateral visual field serves for far sighting, monitoring predators and conspecifics, as well as to detect food at some distance (Fernández-Juricic et al., 2004; Green et al., 1994). Even if pigeons are not be capable of discriminating small and low contrast objects within the thrust phase, being capable of perceiving a predator during the entire head-bobbing cycle may constitute an ecological advantage. During foraging, pigeons mainly focus on targets with the frontal visual field (Goodale, 1982) and longer periods of image stabilization have been observed in hooded cranes (Cronin et al., 2005). Due to the lateral disposition of the pigeon’s eyes, the amount of optic flow in the frontal visual field is reduced. Focusing small targets with the frontal visual field might have the advantage of receiving a double signal, minimizing the motion blur and increasing the visual acuity.

Pattern Recognition During Head-bobbing 79

On the one hand, pigeons are capable of distinguishing medium sized easy patterns during the thrust phase. On the other hand, head-bobbing may assist to discriminate small features at short distances. Objects located at long distances are less affected by the optic flow provoked by the birds’ motion. In addition, image stabilization could be achieved in some birds by rotating the eyes, although in relation to the skull, birds have large eyes that limit eye mobility (Martin and Katzir, 1995). The eye mobility of many birds may be enough to counteract the retina slip at medium to long distances, but the blur of small features located at short distances may require the contribution of head-bobbing. That would explain why about 75% of the head-bobbing birds were foraging birds looking for small feeding sources, like seeds, grain, worms, etc (see section 5).

4.3.2 Head-bobbing motion

The motion analysis of the pigeons’ heads revealed different strategies when observing a stimulus in different parts of the visual field. First, four of the five pigeons, depending on the stimulus presentation, varied the overall speed of the motion by decreasing the duration of the head-bobbing cycle, while keeping the distance invariable. If the stimuli were located in the frontal visual field, the speed of the animals was increased and the duration of the head- bobbing cycle was reduced in comparison to the lateral visual field. Second, thrust phases were longer than hold phases. Third, it was observed that an increase of the speed resulted in a reduced hold phase. We observed significantly shorter hold phases in the frontal visual field than in the lateral visual field in three animals. This phenomenon has previously been reported in pigeons by single-frame analysis of walking pigeons. If the speed of the motion is increased, the hold phase is reduced until speeds above 810mm/s (Davies & Green, 1988), in which the hold phase is replaced by a flexion phase. It was suggested that head-bobbing is controlled by the speed and the visuomotor requirements of the birds, but not by the form of locomotion (Davies & Green, 1988).

Interestingly, one bird used a different strategy varying the head-bobbing frequency: it kept constant the motion speed across the conditions, varying the amplitude and the duration of the Pattern Recognition During Head-bobbing 80 head-bobbing cycles. The head-bobbing frequency was bigger in the frontal visual field than in the lateral visual field presentations, whereas the overall speed of the pigeon motion did not vary. Although this strategy only appears in one of the birds, it may be a common strategy of birds. Chickens confronted with new objects tend to walk slower and perform large head movements (Dawkins, 2002). An extreme example occurs when static pigeons and other birds are found doing head-bobbing while observing their environment. Variations in the head- bobbing cycle duration independent of the motion speed can be useful to increase the optic flow, and therefore enhance distance calculations. These data point out that head-bobbing behaviour in birds is a rather flexible behaviour, which varies according to the situations.

Previous head-bobbing motion analyses found longer durations of the head-bobbing cycle than in the present experiment. The mean duration of the hold phase found in courting birds and during turning behaviour was 156 ms and 139 ms, respectively, whereas the mean duration for the thrust phase was 132 ms and 96 ms, respectively (Troje & Frost, 2000). In the present experiment, the average duration of the hold phase in the frontal and the lateral visual fields was 78 ms and 93 ms, respectively, whereas the thrust phase average duration was 93 ms and 102 ms, respectively. The hold phase took place during 45% of the time in the frontal visual field, and during 47% of the time in the lateral visual fields. In comparison, Frost (1978) described that the hold phase took place during 63% of the time. Troje and Frost (2000) found that the hold phase represented around 57% of the time.

Factors like different experimental conditions and techniques for capturing the head-bobbing behaviour may explain the differences in the results, although the pigeons’ speed may be mainly responsible of these differences. The average speed during the hold phase and the thrust phase was around 27 mm/s and 818 mm/s respectively, while in previous experiments it was found to be around 3.6 mm/s and 630 mm/s (Frost, 1978; Troje & Frost, 2000). The higher speed, combined with the chosen threshold (240 mm/s) could result in the observed higher speed during the hold phase.

Pattern Recognition During Head-bobbing 81

4.3.3 Interim summary

Pigeons are capable of recognizing medium and big patterns during the thrust phase. Although further investigations are needed, head-bobbing may help recognizing small features hidden in complex backgrounds. In addition it may be involved in solving other visual challengers such as depth perception. The capability of discriminating relevant stimuli, like predators, during the entire head-bobbing cycle may increase the survival rate of birds. Pigeons can modify their head-bobbing speed and frequency depending on the stimulus location in the visual field. In addition, different visual inputs may lead into different head-bobbing styles.

Why Do Birds Bob Their Heads? 82

5. EXPERIMENT 5: WHY DO BIRDS BOB THEIR HEADS?

Although head-bobbing behaviour has been very often discussed in the literature, the birds that bob their head are not listed, only around 30 species of head-bobbing birds have been listed in the literature (Cronin et al., 2005; Daanje, 1951; Dagg, 1977b; Dunlap & Mowrer, 1930; Frost, 1978; Fujita, 2003, 2004). Frost (1978) reported that head-bobbing occurs in at least 8 of the 27 orders of birds. Dagg (1977) listed 28 head-bobbing species during locomotion such as pigeons, doves, hens, starlings, pheasants, coots, rails, sand-pipers, phalaropes, parrots, magpies and quails. These lists are very incomplete, but they suggest that at least 1/3 of the birds could show head-bobbing. Comparisons between head-bobbing or non head-bobbing birds have been rarely done and the ecological and behavioural characteristics were not often considered. Many questions remain open. It is not known if all birds of a family or order are head-bobbing birds. Furthermore, it is unknown if head-bobbing could be a determinant trait in a phylogenetic classification.

Retinal stabilization is probably the most accepted head-bobbing function in pigeons, much evidences suggest that head-bobbing might have other important contributions. Pigeons and other birds can display head-bobbing without stabilising the head. It has been observed that during landing and running the hold phase is replaced by a flexion phase, in which the velocity of the head relative to the surrounding is different from zero. The Hooded crane stabilizes the retinal image during foraging (Cronin et al., 2005). In contrast, The Kori bustard, also during foraging, shows rapid thrust phases which most probably are used in order to amplify the relative motion of an object on the ground, this facilitates object detection. Different birds may need to afford diverse ecological and visual demands. Therefore, head-bobbing may play different roles in different environments.

A list of head-bobbing birds and non head-bobbing birds may be a useful tool to estimate the importance of the head-bobbing in the class aves. It would provide a frame in which to integrate previous and future research in restricted conditions in the laboratory.

Why Do Birds Bob Their Heads? 83

The aim of this study is to analyse the ecological and behavioural factors underlying head- bobbing behaviour, such as predatory pressure, source of feeding, habitat, etc. An open question is whether head-bobbing has evolved once or several times during the evolution. Taking in account these findings together with previous laboratory investigations, the functional significance and the phylogenetic origins of head-bobbing will be discussed.

5.1 Methods

This study was divided into three parts. Firstly, birds belonging to a variety of different orders and families were observed and video recorded. The species were classified as head-bobbing, non head-bobbing, hopping and/or other head movement’s birds following a 2004 update of Clement’s classification (Clements, 2000). Second, those data were fed into Sibley & Ahlquist’s (1990) taxonomy tree. Finally, the ecological and behavioural characters under head-bobbing behaviour were analysed.

5.1.1 List of head-bobbing and non head-bobbing birds

Table C-1 (see appendix C) list 322 species of birds which were observed and/or videotaped in their natural environment or in different zoos across Germany. The main aim was to establish a list of head-bobbing and non head-bobbing birds covering all orders and most of the families of the world. Birds’ observation was done in the natural environment in Europe, Africa, and North America. Rare birds and species from other continents were observed in semi-captivity in different zoos of Germany. The birds were organized following Clements checklist of birds (Clements 2000; 2004 updated version). This classification is constantly being updated and offers an easy and intuitive way of organizing birds especially useful in field observation.

From the 204 families of the birds of the world (Clements, 2000; 2004 updated version), 100 families and around 58% of the living birds belong to the Passeriformes order. Although data about Passeriformes order were also collected, major efforts were done in collecting data in Why Do Birds Bob Their Heads? 84

non Passeriformes birds. In addition to the field observation data, bibliographic data from previous studies were also incorporated to the list.

Each bird was classified as head-bobbing (HB), hoping bird (BB), non-bobbing bird (NB) and/or showing other characteristics head movements (OHM).

Classifying a bird in one of these categories does not imply necessarily exclusion from another category. For example, some birds were classified as HB and BB.

Head-bobbing behaviour occurs when a bird shows repeated backward and forward movement of the head in the horizontal plane in any ecological or behavioural context like walking, landing, swimming, etc. If the head-bobbing behaviour was exclusively observed during courtship, the specie was not included in the HB bird category. Head-bobbing is a flexible behaviour among the species. Some species of birds display it often and in a wide variety of circumstances, while other species show head-bobbing occasionally and in very specific situations. For this reason the following notation within the category HB was used:

- HB: Head-bobbing during walking - HB(s): Head-bobbing during swimming. - HB(ws) Head-bobbing during walking and swimming - HB(wl): Head-bobbing during walking and landing - HB(*): Rare or occasional Head-bobbing

The hopping behaviour observed in many birds may be similar to head-bobbing and may play a similar role (Davies and Green, 1988; Friedman, 1975). Hopping can be described as having a hold phase, while the birds stand in the ground, and a thrust face, while the birds are jumping. For this reason hopping birds are classified as body-bobbing birds (BB).

Non-bobbing birds (NB) are those birds, which did show neither head-bobbing nor body- bobbing. In other words they do walk without a backward and forward horizontal movement of Why Do Birds Bob Their Heads? 85 head and they do not hop. A traditional example of this type of bird is the Silver Gull (Dagg, 1977b).

Under the category OHM, birds that move the head vertically, like plovers (Casperson, 1999), were included. Birds that have side to side head movements, like owls (van der Willigen et al., 2002) and parrots were also classified as OHM. Although, this category has been considered to elaborate the bird list, it is not conceptually equivalent to the others. The other categories corresponds to different head or body movements during locomotion, while birds observed doing OHM were generally static.

Classifying a bird within a category is often a difficult task. A bird was classified only when clear evidences were observed. However, we cannot exclude that a bird may be able to display additional behaviours that we missed to observe.

The number of orders, families and species that showed HB, BB, NB, and OHM were counted and analysed. In the list (Fig. 28 and Appendix: Table C-3) if two motion styles appear separated by “&”, it means that the same bird shows both behaviours. For example magpies are classified as HB&BB because they can display both kinds of behaviours. If within a family or an order, any of those motion styles appears separated by a comma, it means that two or more motion styles have been found in different species. For example, marabous are a NB birds and white storks are HB birds, both of them belong to the Ciconiidae family, therefore the Ciconiidae family is classified as “HB, NB”.

5.1.2 Taxonomic tree of head-bobbing and non head-bobbing birds

In order to investigate the evolutionary characteristics of head-bobbing, the observed species of birds were classified following a taxonomic tree. A taxonomic tree was preferred instead a phylogenetic tree, because a phylogenetic tree is always fluctuating and some taxa are better known than other. A taxonomic tree has less detailed information and it is easier to handle and less changeable (von Euler, 2001). For this reason we choose Sibley and Ahlquist taxonomy Why Do Birds Bob Their Heads? 86

(1990; 1988) to classify our list of birds, although phylogenetic would be considered in specific cases. In this taxonomy, the species were divided in 16 different taxonomic levels by using DNA-DNA hybridisation studies. Each of these levels refers to a particular range of genetic

distinctness weight (given in brackets) measured in ∆T50H. The levels proposed by Sibley et al. (1988) are the following: species (0-2.2), subtribe (2.2-4.5), tribe (4.5-7), subfamily (7-9), family (9-11), superfamily (11-13), parvorder (13-15.5), infraorder (15.5-18), suborder (18- 20), order (20-22), superorder (22-24.5), parvclas (24.5-27), infraclass (27-29), subclass (29- 31), class (31-33), superclass (33-36) (Table 4). When the classification was first developed,

one ∆T50H was equivalent to 4.5 million years. According to Sybley and Alquist (1990) “it is likely that this calibration factor is not applicable to all birds, but only to those with ages at first breeding of ca. 2-4 years”.

Category Ending Delta TsoH

Superclass 33-36 Table 4: Categories, endings of categorical names, and Class 31-33 ∆T50H ranges for each categorical rank in Sybley and Subclass ornithes 29-31 Ahlquist’s taxonomy (1990) Infraclass aves 27-29 Parvclass ae 24. 5-27 Superorder morphae 22-24.5 Order iformes 20-22 Suborder i 18-20 Infraorder ides 15.5-18 Parvorder ida 13-15.5 Superfamily oidea 11-13 Family idea 9-11 Subfamily inae 7-9 Tribe ini 4.5-7 Subtribe ina 2.2-4.5 Congeneric spp. 0-2.2

Sibley and Ahlquist´s work has been often discussed in the literature, many critics have arisen since the publication of their work {references and review can be found in O'Hara, 1991 #3329}, but many research support the overall topology of Sibley & Ahlquist (Cooper & Penny, 1997; von Euler, 2001). The authors themselves do gather the different opinions about their own work in the preface to the second printing of “phylogeny and classification of birds” (Sibley & Ahlquist, 1990). In spite of the critics it is the most complete phylogenetic and taxonomic tree found in the literature, an essential prerequisite for this work. Why Do Birds Bob Their Heads? 87

By creating this list of head-bobbing and non head-bobbing birds, it was observed that in most of the cases if a member of a family has head-bobbing, the others members do also have head- bobbing. In case of exceptions to this rule, the detailed phylogenetic relationship among the birds of this family was considered, taking in account several points of view from different authors and methods.

5.1.3 Analysis of behavioural and ecological factors under head-bobbing

For each covered family, a representative bird was selected. If a family contained different species with different bobbing styles (HB, BB, NB), a representative bird of each style was chosen. In total 73 species were analysed: 39 HB birds, 28 NB birds, and 6 BB birds. For each of those birds, the following ecological and behavioural characters were analysed: feeding source, feeding method, group size, peak of activity, habitat, flying abilities, and predator pressure (Apeendix C: Table C-3).

The feeding source was classified in 6 categories. The food size was taked in consideration for creating the categories: 0 = small molluscs, 1 = fruits, leaves, seeds, small vertebrates, invertebrates, insects, 2 = small fishes, squid, crustaceans, 3 = small vertebrates, reptiles, frogs, 4 = vertebrates: birds and small mammals, and 5 = carrion.

Feeding method was classified in seven categories: 1 = foraging or wading, 2 = collecting in trees and bushes, 3 = diving in water, 4 = stalking and spearing, 5 = hunting and searching from the air, 6 = digging and scooping, and 7 = filtering.

The group size was ranked from zero to two: 0 = solitary or pairs, 1 = small groups (from 2 to 10 birds), 2 = big groups (up to 10 birds).

The peak of activity was range according to the intensity of the light, it has three categories: 1 = nocturnal, 2 = crepuscular, and 3 = diurnal. Why Do Birds Bob Their Heads? 88

The habitat was classified from dry to aquatic habitats: 1 = savannah and other open woodlands like highlands, 2 = forests, 3 = swamps, marshes and muddy shores, 4 = lakes and rivers, and 5 = see, ocean and coastal areas.

The flying ability was classified from terrestrial birds to skilled fliers: 0 = terrestrial, 1 = fly short distances (awkward), 2 = normal flyer, and 3 = elegant and skilled.

Finally, the predator pressure was ranged from 0 to 2, where 0 = low predator pressure, 1 = medium pressure, 2 = high predator pressure.

The ecological and behavioural characteristics were evaluated by doing an exhaustive bibliography search (Baran-Marescot, 2005; Jonsson, 1999; Perrins, 1991; Peterson & Hollon, 1983; Stevenson & Fanshawe, 2001). After collecting all necessary data in an initial list, values were given for each bird in each character. Independent judgments from two experimented persons in field observation were used. To simplify the data, each species was assigned exclusively to one range in each trait. For example if the same specie lives in different habitats, the most common habitat was chosen for the data analysis. Occasionally, it was difficult to assign a bird to a range or category, in case of disagreement a third judger helped to classify the bird. In spite of the careful method used, some controversy remained in a few cases.

Selecting only one bird from each family could be a problem regarding the representativeness of this bird. In other words, the question might be if the selected species of bird share the main ecological and behavioural characteristics with all the birds belonging to this family. In most of the families this should not be a problem, given the fact that behavioural and ecological characters are often use as an effective method for classifying birds (Hughes, 1996; Hughes & Baker, 1999). The problems might arise in families in which there is not a general acceptance about the classification due to the different methodology used. In this case, a bird which is not on the centre of the controversy was chosen. Furthermore, in some cases the head-bobbing behaviour might contribute to clarify some parts of birds’ classification.

Why Do Birds Bob Their Heads? 89

5.2 Results

A total amount of 322 species of birds were observed and video recorded in their natural environment or in semi-captivity conditions. From the 32 orders of Clements classification, birds belonging to 26 orders were observed (Fig 28 and Appendix C: Table C-1). It was not possible to collect data about species belonging to Caprimulgiformes, Apodiformes, Dinornithiformes, Trochiliformes, Trogoniformes, and Opisthocomiformes orders. The Caprimugiformes and Dinornithiformes (kiwis) order contains species of shy nocturnal birds difficult to observe. Birds belonging to the Apodiformes (swifts and treeswitfs) and Trochiliformes (hummingbirds) orders have tiny legs, are mainly aerial and walk rarely on the ground. Trogons (Trogoniformes) and Hoatzin (Opisthocomiformes) are arboreal birds that also walk rarely in the ground. As a result of it, it was not possible to gather any information regarding head-bobbing in these species of birds.

Data about the Passeriformes order were collected, but the efforts were focused in observing non Passeriformes birds. As a result of it, we collected data in 68 non Passeriformes families and 25 Passeriformes families. Approximately half of the families of the birds of the world are represented in the list.

5.2.1 Head-bobbing and non head-bobbing birds list

139 birds were classified as head-bobbing birds (HB), 64 were body-bobbing birds (BB), 6 birds showed other head movements (OHM), 23 species were classified as head-bobbing and body-bobbing birds (HB&BB), 2 birds did have head-bobbing and other head movements (OHM) and 87 species did display neither head-bobbing nor body-bobbing (Fig. 29 and Appendix C: Table C-1 ).

In 85 families out of 93 families, all members observed belonged to the same category. For example all species of pigeons are head-bobbing birds. In the remaining eight families, Why Do Birds Bob Their Heads? 90 different species within a family show different motion styles. For example, some species of ovenbirds (Furnariidae) hop while others bob their heads. Six families (Musophagidae, Furnariidae, Turdidae, Sturnidae, Ploceidae, Fringillidae) alternate between HB and BB. These behaviours may be equivalent from the visual point of view (Davies & Green, 1988; Frost, 1978). Clear exceptions are found in three families Cuculidae, Charadriidae, and Ciconiidae. The observed cuculidaes species, were BB or HB unless the Greater Roadrunner (Geococcyx californianus). All charadridae birds were HB birds, unless Southern Lapwing (Vanellus chilensis). Head-bobbing was observed in all storks (ciconiidae) unless Marabou (Leeptoptilos crumeniferus). Eighteen of the 26 orders contain birds that share the same motion pattern. Birds within Passeriformes order alternate between HB, BB and HB&BB. Seven orders contain species showing different motion styles (Fig. 28 and Appendix C: Table C-1).

ORDER FAMILY SPECIES Figure 28: Number of orders, families and species that show head-bobbing (HB), hopping (BB), no HB 7 35 139 head-bobbing and no hopping (NB), and other BB 2 22 64 head movements (OHM). Within and order or a NB 5 23 87 family, if two bobbing styles are separated by a OHM 3 5 7 comma, it means that two or more motion styles HB&BB 1 1 23 have been found in different species. For example, HB&OHM 2 marabous are NB birds and white storks are HB HB, BB 2 birds, both of them belong to the Ciconiidae HB, NB 3 1 family, therefore this family is classified as “HB, BB, NB 1 NB”. When two bobbing styles appear separated HB, NB, OHM 1 by “&”, it means that the same specie shows both BB, HB&BB 3 behaviours. For example magpies are classified as HB, HB&OHM 1 “HB&BB”. HB,NB,HB&OHM 1 HB,BB, HB&BB 1 NB, BB, HB, HB&BB 1 1 Total 26 93 322 Clement’s Checklist 32 204 ~10.000

Head-bobbing was observed in 164 species (48%), 43 families and 15 orders. Body-bobbing was observed was observed in 88 species (25%), 20 families and 7 orders. No-bobbing birds were observed in 87 species (25%), 25 families and 12 orders. Other-head-movements were observed in 8 species (2%), 6 families and 5 orders (Fig. 29). Why Do Birds Bob Their Heads? 91

NB OHM 25% 2% ORDER FAMILY SPECIES

HB 15 44 165 HB BB 7 29 88 BB 48% NB 12 25 87 25% OHM 5 6 8

Figure 29: The first and the second columns represent the number of orders and families that contain species displaying HB, BB, NB, and OHM. The third column represents the number of species that show any of the bobbing styles. An order, family, or specie can be included in several categories. For example, Magpie is counted twice as HB and as BB. The chart represents the percentage of species that show HB, BB, NB, or OHM.

Twenty species of birds showed a rare or occasional head-bobbing (HB*). In 10 species it was possible to establish under which circumstances head-bobbing occurred (Table 5). Head- bobbing was found in ostriches walking through long grasses in the savannah (Fig. 32), but not while foraging or in captivity. Head-bobbing was also observed in Emu chicks in semi- captivity conditions, but not in adults. A related species, Greater Rhea (Rhea americana) did not show head-bobbing in captivity and it was classified as NB, but it cannot be excluded that under certain circumstances rheas may perform head-bobbing. Plovers and lapwings (Charadriidae family) showed head-bobbing while walking slowly in tidal mudflats, shallow flood pools, but not in open short-grass lands or bare soil. Bustards performed fast head- movements while walking very slowly in the savannah looking for food. Jackdaws, crows and starlings showed head-bobbing in slow locomotion, but not at fast speed.

Other species like moorhens, grebes, and coots also exhibit occasional head-bobbing while swimming, but it was not possible to undoubtedly establish when the head-bobbing behaviour occurs. Nevertheless, longer periods of field observation in their natural environment revealed that these aquatic birds showed head-bobbing while swimming slowly, when a novel stimulus appears, and while approaching the shore. They did not show head-bobbing while swimming fast, while escaping, or while moving towards a clear target like food.

In addition, marabous were classified as NB birds, because they normally do not show head- bobbing. However, in a very specific situation, a flock of marabous showed traces of head- Why Do Birds Bob Their Heads? 92

bobbing behaviour. The birds were occupied in searching small pieces of carrion left by the crocodiles, dragged by the water in a river, during migration time in Tanzania. Unfortunately, it was not possible to record a clear footage of the birds.

Common name Scientific name HB situation Ostrich Struthio camelus Walking through long grasses Australian Emu Dromaius novaehollandiae Chicks in captivity Little Ringed Plover Charadrius dubius Foraging in shallow flood pools Egyptian Plover Pluvianus aegyptius Foraging in shallow flood pools Northen Lapwing Vanellus vanellus Foraging in shallow flood pools Blacksmith Lapwing Vanellus armatus Foraging in shallow flood pools Kori Bustard Ardeotis kori While foraging Black Bustard Eupodotis afra While foraging Jackdaws & crows Corvus sp. Walking slow Starling Starling sp. Walking slow Oystercatcher Haematopus sp. Walking slow Table 5: In ten species of birds with rare or occasional head-bobbing, it was possible to establish the circumstances under which head-bobbing occurs. Where sp. = all species contained in a genus

5.2.2 Taxonomic tree of head-bobbing and non-bobbing birds

The field observations about head-bobbing and non-bobbing birds were feed into Sibley and Ahlquist taxonomy. Furthermore, the information given in this part regarding phylogenetic relationships is based on the Sybley and Ahlquist taxonomy (Sibley & Ahlquist, 1990) unless explicit citation. In this section, the information gathered in “Fig. 30” and in “Appendix C: Table C-2” is described together with a resume of the birds ecological and behavioural characteristics included in each taxon. In Fig. 30 when all observed birds within a taxon shared a common bobbing style, further sub-taxa were not included. In comparison in “Appendix C: Table C-2” all studied taxa and sub-taxa were included.

In Sibley and Alhquist’s classification, the modern birds Neornithes are divided into two infraclasses: Eoaves (Paleonagthae) and Neoaves (Neonagthae). All species belonging to the infraclass Eoaves (ostriches, rheas, cassowaries, tinamous and emus) except for rheas (Ratites), were head-bobbing birds. Ratites constitute a monophyletic group being the head-bobbers tinamous their nearest living relatives. Rheas were exclusively observed in captivity and it can Why Do Birds Bob Their Heads? 93

not be discarded that these birds, like the closely related ostriches (Sibley & Ahlquist, 1990; Sibley et al., 1988; van Tuinen, Sibley, & Hedges, 2000), may exhibit head-bobbing in the natural environment under specific circumstances (Fig. 30 and Appendix C: Table C-2).

Infraclass Neoaves contains parvclasses Galloanserae, Turnicae, Picae, Coraciae, Colidae, and Passerae. Parvclass Galloanserae (including ducks and fowls) are closer to paleognaths than to other neognaths. Therefore they constitute and ancient group in birds’ evolution. All fowls show very distinctive head-bobbing behaviour, while ducks do not exhibit head-bobbing. Only one exception was observed in Wood Duck: head-bobbing was video recorded in a bird before an escaping flight. Unfortunately, further observations could not confirm the data. These head movements may be an alert signal for other birds. Therefore Wood Duck has been classified as NB. Parvclass Turnicae, containing the order Turniciformes and the family Turnicidae, commonly called button-quails are also head-bobbing birds. Parvclass Picae that includes woodpeckers, honyeguides, barbets, toucans, jacamars, and puffbirds (Fig. 30 and Appendix C: Table C-2) are body-bobbing birds. Within the Parvclass Coracidae species from four families were observed: ground hornbills (NB), typical hornbills (BB), hoopoes (HB), and ground- rollers (NB).

The majority of the living birds are classified within the parvclass Passerae and a more detailed analysis is needed. Superorders Cuculimorphae, Psittacimorphae, Strigimorphae and Passerimorphae are included in parvclass Passerae. The superorder Cuculimorphae contains cuckoos and the Hoatzin. All birds in this group are HB, BB, or both, the only exception is the Great Roadrunner, this bird moves rarely at slow motion. Cuckoos are HB an ancient lineage with much genetic diversity among the living groups. Unfortunately, data about Hoatzin are lacking. The species observed within the Psittacimorphae superorder (parrots) were classified as OHM, but more detailed observations are needed. However, casual observations revealed that some species hop while others walk without head-bobbing. For example, recent observations showed that Kea (Nestor notabilis) walk without head-bobbing.

Why Do Birds Bob Their Heads? 94

Class Aves Subclass Neornithes Infraclass Eoaves , HB* (unless Rhea NB) Infraclass Neoaves Parvclass Galloanserae Superorder Gallomorphae HB Superorder Anserimorphae NB Parvclass Turnicae HB Parvclass Picae BB Parvclass Coraciae Superorder Bucerotimorphae Order Bucerotiformes: Family Bucerotidae BB Family Bucorvidae NB Order Upupiformes HB Superorder Coraciimorphae Order Coraciiformes NB Order Coliiformes BB Parvclass Colidae BB Parvclass Passerae Superorder Cuculimorphae HB, BB, HB& BB (unless roadrunners NB) Superorder Psittacimorphae OHM Superorder Strigimorphae Order Musophagiformes HB&BB Order Strigiformes OHM Superorder Passerimorphae HB Order Columbiformes HB Order Gruiformes HB (unless kagus NB) Order Ciconiiformes Suborder Charadrii HB (unless thick-knees and Laoroidea superfamily NB) Suborder Ciconii Infraorder Falconides NB (unless Secretarybird HB) Infraorder Ciconiides Parvorder Podicipedida: HB Parvorder Sulida: NB Parvorder Ciconiida Superfamily Ardeoidea HB Superfamily Scopoidea HB Superfamily Phoenicopteroidea NB Superfamily Threskiornithoidea Superfamily Pelecanoidea NB Superfamily Ciconioidea Family Ciconiidae Subfamily Cathartinae NB Subfamily Ciconiinae HB Superfamily Procellarioidea NB Order Passeriformes HB, BB, HB&BB

Figure 10: Short summary of the motion birds list following Sibley and Ahlquist’s (1990) taxonomy. When a group is classified in a bobbing category, all subgroups contained in it are included in this category.

Why Do Birds Bob Their Heads? 95

Superorder Strigimorphae includes Turacos (order Musophagiformes) and owls (order Strigiformes). Turacos were traditionally classified as being related with cuckoos (superorder Cuculimorphae), although DNA comparisons found that they are not one another’s closest living relatives (See references and discussion in Sibley & Ahlquist, 1990). Turacos, like some cuckoos, hop and bob their heads. It is well known that owls perform side-to-side head movements that serve as a depth cue for motion parallax (van der Willigen et al., 2002). It was not possible to observe the motion of other nocturnal birds like nightjars, nighthawks, oilbirds, and frogmouths that also belong to the order Strigiformes.

Superorder Passerimorphae is divided in four orders: Columbiformes, Gruiformes, Ciconiiformes, and Passeriformes. After observing 24 species within the Columbiformes order, no exceptions were observed. All pigeons and doves are head-bobbing birds. The Columbiformes family is a monophyletic taxon with no close living relatives.

The order Gruiformes (parvclass Passerae) is a highly divergent monophyletic taxon including cranes, sunbitterns, bustards, limpkins, sungrebes, trumpeters, kagu, rails, and mesites. All birds observed within this order are head-bobbing birds except kagu (family Rhynochetidae). In semi-captivity, those birds run very fast from one side to another, never walking in slow motion. Seriemas are closely related with kagus. Initially the seriemas were classified as non head-bobbing birds, but further observations revealed short periods of head-bobbing, so they were reclassified as HB*.

The order ciconiiformes is divided in suborders Charadrii and Ciconii. The Charadrii, with 366 species, is one of the largest avian suborders. It contains two infraorders: Pteroclides and Charadriides. In addition, Charadriides infraorder is divided into Scolopacida and Charadriida parvoorders. Infraorders Pteroclides (sandgrouse) and Parvoorder Scolapacida (seedsnipe, woodcocks, snipes, curlews, pharalopes, etc) exclusively contain head-bobbing birds. Within the Charadriida parvorder HB and NB have been observed. NB birds are observed in Burhinidae (thick-knees), Glareolidae (Crab Plover and pantrincoles), Laridae (gulls and allies), whereas occasional HB birds are observed in Charadriidae family (plovers, lapwings, Why Do Birds Bob Their Heads? 96 etc) (Appendix C: Table C-2). Phylogeny of suborder Charadrii has been broadly investigated by using different methods which resulted into different phylogenetic trees (Chu, 1995; Ericson et al., 2003; Sibley & Ahlquist, 1990). It constitutes an example in which head-bobbing can be used as a characteristic that may help to clarify the phylogenetic tree of suborder Charadrii.

Ciconii suborder is a very eclectic group in terms of their motion styles: NB, HB, BB, and OHM species have been observed within this group. Ciconii suborder is divided between infraorders Falconides and Ciconiides. Species within infraorder Falconides (hawks, eagles, vultures, etc), except for Secretary bird, do not show head-bobbing. Infraorders ciconiides contains Podicipedida, Phaethontida, Sulida, and Ciconiiida parvorders. The grebes linage (parvorder Podicipedida) classified as HB birds, branched earlier from the common ancestor in the ciconiides and have no close living relatives. In Phaethontida parvorder, no convincing data were collected, although a short video of a Red-tailed Tropic Bird (Phaeton rubricauda) from Internet Bird Collection (Lynx Editions, 2005) show one of those animals hopping. Bobbies and gannets belonging to the family Sulidae and cormorants belonging to Phalacrocoracoidae family (parvorder sulida), are classified as non-bobbing birds. Anhingidae, the sister family of Sulidae and Phalacrocoracoidae families, is integrated by anhingas and darters. As in Phaethontida parvorder, clear data are lacking, but a very short internet bird collection video (Lynx Editions, 2005) show one anhinga doing head-bobbing while emerging from the water. Cormorants are classified as non-bobbing birds.

Herons, hammerkop, storks, ibises, flamingos, pelicans, albatrosses, petrels and allies have been traditionally included in the Cinoniida parvorder. Pelicans (superfamily Pelecanidae), flamingos (superfamily Phoenicopteridae), albatrosses, petrels, shearwaters, diving-petrels, penguins, loons and frigatebirds (superfamily Procellarioidea) are non-bobbing birds. Hammerkop (superfamily Scopoidea), herons and egrets (superfamily Ardeoidea) are HB birds. Morphological similarities and DNA that New World vultures (family Catthartinae) and storks (family Ciconiinae) are closely related, they have been included within superfamily Ciconiidea. New world vultures are NB birds, while all storks except for the Marabou (Leptoptilos crumeniferus) are HB birds.

Why Do Birds Bob Their Heads? 97

The Passeriformes order contains around 5500 of the 10000 species classified within Clements checklist (Clements, 2000, 2004 updated version). Data about 86 species covering 25 families were collected. Further observations are required to obtain representative data. However, the preliminary observations found that suborder Tyranniy (suboscines) contains HB and BB birds. Similarly, the suborder Passeri (oscines) contains HB, BB and HB&BB birds. Most Passeriformes are small, land-dwelling birds that feed primarily on insects, seeds, fruit, or nectar. The Passeriformes diverged from the non-passeride clade composed of the columbiformes, gruiformes and ciconiiformes birds. Columbirformes and gruiformes are mainly HB birds, whereas ciconiiformes is a miscellaneous group where HB, NB and BB species have been observed.

5.2.3 Analysis of behavioural and ecological factors under head-bobbing

Around 72% of the HB birds feed on seeds, fruits, small invertebrates, insects, etc., 18% feed on fish, squids, and crustaceans, and 10% feed on small vertebrates, reptiles, and frogs. In comparison, 35% of NB birds feed on seeds, fruits, small invertebrates, insects, etc., 36% feed on fish, squids, and crustaceans, 21% feed on small vertebrates, reptiles, and frogs, 14% on carrion, and 4 on tiny molluscs (Fig. 31 and Appendix C: Table C-3). Although it is controversial to assume that the feeding source is rated according to the size, a Spearman correlation revealed a significant negative correlation between head-bobbing and food size (r =-.38, n=67, p<0.001). Head-bobbing birds prefer small sized food in comparison to non- bobbing birds that choose bigger sources of feeding.

The analysis of the feeding methods revealed that 78% of the head-bobbing birds look for food by foraging or wading, 10% stalking and spearing, and the remaining 12% were divided equally between diving, collecting in bushes and trees, digging, and hunting or searching from the air. In contrast, 33% of the non-bobbing birds look for food foraging or wading, 25% diving in water, 26% hunting or searching from air, 7% filtering, and the remaining 8% are divided equally between stalking or spearing, and digging or scooping (Fig. 31 and Appendix C: Table C-3). The group size analysis revealed that 36%, 31%, and 33% of HB birds live in Why Do Birds Bob Their Heads? 98

solitary, small groups, and in groups, respectively. While 46%, 25%, and 29% of NB birds live in solitary, small groups, and groups, respectively (Fig. 31 and Appendix C: Table C-3).

All head-bobbing birds were diurnal birds. All NB birds were diurnal except one (Stone curlew). Around 38% of the HB birds live in the savannah and other open woodlands, 38% live in swamps, muddy shores, and marshes, 21% live in forests, and only 3% live on lakes and rivers. 32% of NB birds live in the savannah and other open woodlands, 20% live in forests, 16% live on swamps, muddy shores and marshes, and 32% lives in sees, oceans, and coastal areas (Fig. 31).

It was found that 46% of the HB birds have poor flying abilities (short distances, awkward), 26% are normal flyers, 18% are elegant and skilled, and 10% do not fly (terrestrial). In comparison, 21% of the NB birds are awkward or fly short distances, 24% are normal flyers, 48% are elegant and skilled flyers, and 7% (penguins) do not fly (Fig. 31). Head-bobbing birds are less skilled in flying than non head-bobbing birds. A negative correlation, between head- bobbing and flying abilities, was found (r = -.312, n=67, p< 0.01).

HB birds: feeding NB birds: feeding 18% 14% 3% 3% 24%

10%

21% 72% 35% 1 Fruits, leaves, seeds, small vertebrates, invertebrates, insects, etc. 0 Molluscs 2 Fishes, squids, crustaceans 1 Fruits, leaves, seeds, small vertebrates, invertebrates, insects, etc. 2 Fishes, squids, crustaceans 3 Small vertebrates, reptiles, frogs 3 Small vertebrates, reptiles, frogs 4 Vertebrates 5 Carrion

HB birds: feeding method NB birds: feeding method 3% 3% 7% 10% 4% 3% 33% 3% 26%

78% 4% 26% 1 Foraging and wading 2 Collecting in trees and bushes 1 Foraging and wading 3 Diving in water 3 Diving in water 4 Stalking and spearing 4 Stalking and spearing 5 Hunting or searching fron the air 5 Hunting, searching fron the air 6 Digging or scooping 6 Digging or scooping 7 Filtering Why Do Birds Bob Their Heads? 99 HB birds: group size HB birds: group size

33% 36% 33% 36%

31% 0 S o li ta ry or pairs 0 Solitary or pairs 31% 1 Small groups (2-10) 1 Small groups (2-10) 2 Groups 2 Groups

HB birds: habitat NB birds: habitat 3% 31% 31% 38% 38%

4% 21% 15% 19% 1 Savannah and other open woodlands 2 Forest 1 Savannah and other open woodlands 3 Swamps, marshes, and muddy shores 2 Forest 4 Lakes and rivers 3 Swamps, marshes, and muddy shores 4 Lakes and rivers 5 Seas, oceans, and coastal areas NB birds: flying abilities HB birds: flying abilities 7% 18% 10% 21% 48%

26% 46% 24% 0 Terrestrial 0 Terrestrial 1 Short distances, awkward 1 Short distances, aukward 2 Normal flyer 2 Normal flyer 3 Elegant and skilled 3 Elegant and skilled

HB birds: predator pressure HB birds: predator pressure 4% 13% 9% 30%

67% 57% 0 Low 0 Low 1 Medium 1 Medium 2 High 2 High Figure 31: The following ecological and behavioural characteristics of HB and NB birds were analysed: feeding source, feeding method, group size, habitat, flying abilities, and predator pressure. Why Do Birds Bob Their Heads? 100

In HB birds, the predator pressure was low, medium, and high in 30%, 57%, and 13% of the birds, respectively. In NB birds, the predator pressure was low, medium, and high in 67%, 29%, and 4%, respectively (Fig. 31 and Appendix C: Table C-3). A positive correlation between head-bobbing and predator pressure was observed (r = .391, n=67, p< 0.001). Head- bobbing birds are more likely to suffer a higher predator pressure than non head-bobbing birds.

Only six species of BB birds were analysed, therefore the sample may not be representative, and further analysis would be needed. The analysed BB feed on fruit, leaves, invertebrate, small vertebrate, etc. (45%). They also eat fishes, squids, and crustaceous (22%), small vertebrates, and reptiles (33%). They are diurnal birds that collect food mainly in trees and bushes (66%), foraging and wading (17%), and diving in water (17%). They prefer forests and bushes to live (66%), but 34% of them are divided equally between open lands and marine habitats. They live normally in solitary (67%) or in small groups (33%). They are generally normal flyers (55%), or skilled (33%). The observed predator pressure was low in all species.

5.3 Discussion

A list of 322 species of head-bobbing, body-bobbing and non-bobbing birds was elaborated. Approximately 80% of the orders and almost half of the families on all continents of the world were represented in the list. Head-bobbing was found in 15 of the 26 orders analysed and in 163 species of birds. Almost half of the species observed were head-bobbing birds, 25% were body-bobbing birds, and the remaining birds were non-bobbing birds. In addition, a small percentage of birds (2%) were observed performing other-head-movements.

Previous reports found HB in 8 orders and in 28 of the 48 analysed species (Dagg, 1977b). Until now, the occurrence of head-bobbing in birds may have been underestimated due to the lack of data, but also because careful observations are needed to detect head-bobbing in species that display it only occasionally. Around half of the living birds may be capable of doing head-bobbing. The other half may be equally divided between NB birds and BB birds. The real Why Do Birds Bob Their Heads? 101

proportion of NB birds is probably even smaller because Passeriformes were under represented in our sample The order Passeriformes includes around half of the living birds, and all of them were either HB or BB birds.

The lack of data in Caprimulgiformes, Apodiformes, Dinornithiformes, Trochiliformes, Trogoniformes, and Opisthocomiformes orders barely affect these estimations. On the one hand, these orders contain few species of birds and on the other hand, predictions of the locomotive behaviour of these birds can be made. Oilbirds, nightjars, frogmouths, and allies (Caprimugiformes) are nocturnal birds which form a monophyletic group with owls, swifts and hummingbirds (Trochiliformes) (Sibley & Ahlquist, 1990). All these groups are probably integrated by NB nocturnal birds. In addition, Apodiformes and Trochiliformes are mainly aerial species, which may not have head-bobbing. In this study it has been observed that HB occurs more often in terrestrial species. Dinortithiformes (kiwis) are most probably head- bobbing birds that together with cassowaries and the Emu form a monophyletic group, although it may well occur that HB appears rarely in kiwis due to their nocturnal habits. Trogons and Hoatzin are arboreal species, which may prefer hopping for moving between branches and bushes, although HB behaviour cannot be discarded.

5.3.1 List of Head-bobbing and non-bobbing birds: exceptions within a family

Around 90% of families share the same bobbing style. In other words, if HB is observed within a family, most probably all other members of the family are also HB birds. Approximately 7% of the birds alternated between HB and BB, which, as discussed in section 5.3.3, may have similar roles. Clear exceptions were found in three families Cuculidae, Charadriidae, and Ciconiidae. These tree families were consisted only by HB birds except for one species. Exceptions in non-bobbing families were not found. In addition the non-bobber Stone curlew (Burhinus oedicnemus), and the head-bobber Secretary bird (Sagittarius serpentarius), were considered in our analysis as exceptions, because all closed relatives of these species are non- bobbing birds. In this section, the selection pressures that may lead to an exception in closely related birds are discussed. Why Do Birds Bob Their Heads? 102

The Greater roadrunner is the only specie observed within the Cuculidae family that shows neither HB nor BB. It is a large, long-legged, long-tailed cuckoo that spends most of the time on the ground. Roadrunners are capable of sprinting up to 25 km an hour and rarely seem to walk slowly. The high speed during locomotion may explain the lack of head-bobbing. It has been observed that at high speed head-bobbing is suppressed in pigeons (Davies & Green, 1988) and field observations in many other species have also confirmed that head-bobbing is suppressed at high speeds (Dagg, 1977b).

The Southern lapwing (Vanellus Chilensis) is the only bird within the Vanellus genus (Charadriidae family) that walks without head-bobbing. In other lapwings, head-bobbing while walking is observed in marshes and other wet lands. Video recordings of the Southern lapwing were done exclusively in grass areas. Therefore, further observations in marshes and wet lands are needed to determinate whether the Southern lapwing is a HB bird.

All ciconiidaes, except Marabou, were observed walking with head-bobbing. The Marabou is the only stork that feeds on carrion. The Marabou behaviour was observed during long periods of time in natural conditions. When the animal was feeding from big carcasses no traces of head-bobbing were found. Although in a very specific situation a flock of marabous showed traces of head-bobbing. The birds were occupied in searching small peaces of carrion left by the crocodiles, dragged by the water in a river, during migration time in Tanzania. Unfortunately, it was not possible to record clear footage of the birds. For this reason the birds were classified as NB birds. The feeding habitats and the life style of marabous may not require head-bobbing, but that might occur under certain circumstances. During head-bobbing walking, the necks of the marabous were extended, whereas during non head-bobbing walking the necks were shrunk, which may be related to the lack of head-bobbing.

The Stone Curlew (Burhinus oedicnemus), an integrant of the Burhinidae family, is a bird closely related with Charadriinae subfamily. It is active at night, and its large yellow eyes enable it to locate food when it is dark. On the one hand it cannot be dismissed that Stone Curlew may exhibit HB in certain conditions, as the animals were observed during the day and Why Do Birds Bob Their Heads? 103

they were not very active. On the other hand, all observed nocturnal birds did not display head- bobbing which leads to the conclusion that most probably head-bobbing is no longer effective during night. For nocturnal birds like owls having eyes located frontally and receiving a double signal of the same visual area, may be an advantage which would increase the possibilities of detecting a target.

The Secretary birds (Sagittarius serpentarius), belonging to the infraorder Falconides is the only bird of prey capable of performing head-bobbing. It has long legs, it feeds on snakes or other reptiles, amphibians, tortoises, rats and other small mammals as well as young game birds. Secretary bird is the only terrestrial bird of prey hunting on the ground at medium or short distances. Head-bobbing may aid these birds to distinguish preys hidden in the floor. Assuming that all falconides have a common ancestor, this can be a case of head-bobbing reinvention.

In conclusion, the study or exceptional cases, regarding the bobbing style in closely related group of birds, revealed that head-bobbing may be suppressed in birds that move generally at high speed, in non ground foraging birds, and in nocturnal birds. On the contrary, head- bobbing may appear in ground foraging birds.

5.3.2 Rare or occasional head-bobbing behaviour

Ostriches, plovers and lapwings, bustards, crows, starlings, moorhens, coots and grebes display rare or occasional head-bobbing. In these animals, the analysis of the situation in which head- bobbing occurs may contribute to understanding the functional significance of head-bobbing.

Head-bobbing was found in ostriches walking through long grasses in the savannah (Fig. 32), but not while foraging or in captivity. The predators of Ostrich are Cheetah, Lion, Leopard, Wild Dog and Spotted Hyena. Adult males have been seen to kick out at, and wound, large predators. In open lands, ostriches are able to outrun most of their enemies reaching 70 km per hour in a short sprint in open country, and 50km per hour over a distance of 30km. A Why Do Birds Bob Their Heads? 104

vulnerable situation may occur when crossing long grasses, where a predator can be hidden. Head-bobbing in ostriches may serve to seek for predators hiding in grasses. It has been proposed that head-bobbing serves for pattern recognition and for distinguishing between own and external motion (Davies & Green, 1988; Frost, 1978). By stabilizing the retinal image, the ostriches may be more efficient detecting static and moving predators.

Plovers and lapwings (Charadriidae) showed head-bobbing while walking slowly in tidal mudflats, and shallow flood pools, but not in open short-grasslands or bare soil. They feed on insects, worms, spiders, molluscs, and shrimps (Jonsson, 1999). Many members of the Charadriidae family were observed performing head-bobbing and also a type of up-and-down head movement. Those vertical head movements may aid a bird in acquiring visual information avoiding the distortion caused by the water reflection and refraction (Casperson, 1999). According to the model proposed by Casperson, the head-bobbing behaviour may not help much solving this problem, because vertical movements are necessary. Therefore, head- bobbing in shallow flood pools may have other functions most probably related with foraging.

Figure 32: A male ostrich walking in Ngorongoro National Park (Tanzania). The Ostrich (Struthio camelus) shows a clear head- bobbing while walking in its natural environment through long grasses. Head-bobbing was neither observed during foraging nor in captivity.

Foraging in mudflats and shallow food pools may require head-bobbing in order to localize small animals hiding in the mud. Plovers and lapwings walk deliberately slowly, thus rapid head movements may also help to identify objects lying in the mud at different distances. That is, by moving the head fast, the relative displacement of a worm on the retina, which is lying on the surface of the mud, will increase with respect to the relative displacement of the ground on the retina.

Why Do Birds Bob Their Heads? 105

Bustards have been observed bobbing their heads while walking slowly in grass-lands. The head-bobbing behaviour of bustards is very peculiar because they perform fast and long thrust phases, intercalated by very brief hold phases. They feed on variety of insects, small mammals, lizards, snakes, seeds, and berries. The long thrust phase exhibited by bustards may serve for distance calculations that may allow a better detection of small animals and seeds located in uniform floors.

Jackdaws, crows and starlings showed head-bobbing in slow motion. In magpies (Pica pica), it has been observed that running and hopping are alternative gaits at speeds higher than walking. It is unknown why these birds can walk doing head-bobbing, run, or hop, and why they prefer hop to run at high speeds. Kinematics studies of the three types of locomotion did not find any suitable explanation (Verstappen et al., 2000). If hopping is equivalent to head-bobbing and plays a similar role in visual perception, this may be an explanation why Corvidae birds prefer hopping over running at high speeds. In addition, hopping may be more efficient for small birds when moving through trees and bushes.

Other species like moorhens, grebes, and coots also exhibit occasional head-bobbing while swimming. Head-bobbing was observed while walking (in coots and moorhens), swimming slowly, when a novel stimulus appears, and while approaching to the shore. Head-bobbing was not observed while swimming fast, while escaping, or moving towards a clear target like food. It is not possible to clearly determinate when head-bobbing occur in these birds, but most probably head-bobbing may facilitate prey detection, distance calculation and image stabilization. Head-bobbing may be useful to keep a stabilized image of the environment while the animal is pulled by a water string.

The analyses of the situations in which head-bobbing occurs in HB* birds revealed similar results to the observation made on section 5.3.2. Head-bobbing occurs at slow motion, when birds need to distinguish stimuli hidden in a complex background at short or medium distances. Why Do Birds Bob Their Heads? 106

5.3.3 Are body-bobbing birds head-bobbing birds?

It has been suggested that HB and BB may play similar functions from the visual point of view (Davies and Green, 1988; Friedman, 1975, Davies, 1988 #2125; Frost, 1978 #1141). The data collected from HB and BB birds support this hypothesis. Both types of locomotion have been observed in turacos, blackbirds, jays, crows, magpies, weavers, and couas. Furthermore, six families (Musophagidae, Furnariidae, Turdidae, Sturnidae, Ploceidae, and Fringillidae) and the entire Passeriformes order alternate between HB and BB. Similarly to head-bobbing, body- bobbing may appear early in evolution. All observed birds within parvclass Picae (woodpeckers, honeyguides, barbets, etc), one of the oldest avian lineage with living descendants, are body-bobbing birds (Fig. 30).

Body-bobbing may facilitate the motion of small birds between branches and trees, without losing the visual advantages of HB. Hopping can be described as having a hold phase, while the birds stand on the ground, and a thrust face, while the birds are jumping. It may occur that while foraging the hopping birds have longer period of head stabilization than during normal locomotion. Frame by frame analysis of hopping sparrows (Passer domesticus) found that “bird’s head is thrust forward before the legs start to push the body into the air. Furthermore, the head stops and is stabilized in the visual space before the body finished landing from the hop”. This behaviour is observed in foraging and alert sparrows but not in somnolent ones (Friedman, 1975). Further analyses of the kinematics characteristics of body-bobbing and head-bobbing are required to confirm this suggestion.

5.3.4 Are other-head-movements functions similar to head-bobbing functions?

Parrots and owls are not closely related, hence the head-movements found in these groups may have a completely different origin. To our knowledge few data exists about head-movement in parrots. Barn owls, with frontally located eyes, perform side to side head-movements that are classified in the present study as OHM, which similarly to head-bobbing may aid head- Why Do Birds Bob Their Heads? 107

stabilization (Hyde & Knudsen, 2002) and motion parallax calculation (van der Willigen et al., 2002). Owls may have frontally located eyes as a consequence of their nocturnal activity. For retrieving parallax information head-bobbing is not longer effective and horizontal movements are required. On the contrary, the vertical head movements observed in plovers may play different role from head-bobbing. It has been proposed that these head-movements may aid in acquiring visual information avoiding the distortion caused by water reflection and refraction (Casperson, 1999).

5.3.5 Head-bobbing evolution

Two simplified evolutionary scenarios for head-bobbing evolution should be considered. On the one hand head-bobbing may have evolved at some point in the evolution in a common ancestor of all birds. On the other hand, similar selection pressures may have resulted into several evolutions of head-bobbing at different moments in evolution. Probably, a combination of these two simplified evolutionary scenarios is given in head-bobbing evolution. Head- bobbing is a highly distributed trait in the avian phylogeny. Therefore, suppressions and independent evolutions of head-bobbing may easily have occurred in 130 million years of avian evolution.

5.3.5.1 Did head-bobbing appear in a common ancestor of the modern birds?

Ratites and tinamous (Eoaves) are traditionally located at the base of the phylogenetic tree of modern birds (For discussion and references see Sibley, 1990) Morphologic and DNA sequences analyses tend to support this traditional view (Sibley & Ahlquist, 1990; van Tuinen et al., 2000), mitochondrial DNA analysis revealed that passeriforms occupy the most basal position among all modern birds (Harlid, Janke, & Arnason, 1997, 1998). The discussion rests on whether earlier aves were heavy-bodied, ground-dwelling birds, or arboreal perching birds. According to our observations, the prototypical head-bobbing bird is terrestrial, with low flying Why Do Birds Bob Their Heads? 108 abilities and ground-foraging that normally does not live in marine habitats (except for tidal areas).

Independently of these two hypotheses, head-bobbing has been found in both Passeriformes and Paleonagtae (Eoaves) birds, which are postulated to be the oldest linage of modern birds. Therefore, head-bobbing may have appeared early in the evolution and it may be a characteristic inherited from a common ancestor of modern birds. In fact, head-bobbing does not appear in more recently evolved birds like eagles, hawks, and kites, with frontally located eyes. During evolution, selective pressures for binocularity may have lead to the loss of the head-bobbing behaviour.

The relative brain and eye size of birds of prey and nocturnal birds is bigger than in ground foraging birds. Hunting species, with larger eyes, may need to increase the amount and the complexity of visual information processing (Garamszegi et al., 2002). Head-bobbing might be an early solution for solving visual demands, such as pattern recognition, distance calculation, and vigilance in birds with lateralised eyes.

The two major competing hypotheses on avian evolution are the "basal archosaur hypothesis" and the "theropod dinosaur hypothesis". The “basal archosaur hypothesis” proposes that the first birds descended directly from ancestral reptiles about 230 million years ago. The "theropod dinosaur hypothesis" proposes that the aves derive from dinosaurs some 130 million years ago (Shimizu & Karten, 1991). Most birds and reptiles have lateralised eyes (Husband & Shimizu, 2001). Interestingly head movement behaviour connected with visual processing has been found in lizards (Flanders, 1985, 1988; LeBas & Marshall, 2000). African chameleon follows horizontal, sinusoidal cricket (bait) movement with sinusoidal head movement and no apparent eye movement. The chameleons used head amplitudes that minimized the motion of the bait relative to the head (Flanders, 1985, 1988). Head-bobbing in Scincid lizard (Lampropholis guichenoti) serves to enhance visual acuity (Torr & Shine, 1994).

If head-bobbing evolved first in an ancestor of the class aves, it may have been suppressed several times during evolution. As explained above, the development of binocularity can also Why Do Birds Bob Their Heads? 109

result in head-bobbing suppression. Convergent suppression of head-bobbing can be found in palmate and totipalmate birds. Those birds constitute a polyphyletic non-bobbing group belonging to different evolutionary lineages adapted to live in aquatic environments. Head- bobbing suppression may be related to the evolution of anatomical adaptations in water birds, like palmate fingers. Interestingly, grebes, coot and moorhens are good swimmers but lack palmate fingers. Examples of head-bobbing suppression within a family can be found in Road- runner adapted to move fast, Marabou adapted to feed on carcass, and Stone curlew adapted to live during the night (see section 5.3.2).

5.3.5.1 Did head-bobbing evolved as a convergent trait due to similar selection pressure?

Although some evidences suggests the existence of a common ancestor for all head-bobbing birds, it can not be dismissed that head-bobbing may have evolved independently at different moments in evolution. Secretary bird is the only falconides with head-bobbing (see section 5.3.2), and assuming that all birds within infraorder Falconides share a non-bobber common ancestor, this can constitute an example of independent head-bobbing development. Furthermore, similar selection pressure may lead into a convergent HB behaviour in species that belong to different evolutionary lineage. It may well happen that HB has evolved at least 10 times in evolutions (see Fig. 30) probably as a consequence of similar feeding habits or predatory pressure. For example, in gallomorphaes (fowls), columbiformes (pigeons and doves), ciconiinae (storks), charadrii (lapwings and plovers) which are ground foraging birds, head-bobbing may have evolved independently as a mechanism to detect small features at medium to short distances, while monitoring the environment for predators and other relevant stimuli.

Interestingly, we observed, within a head-bobbing monophyletic family, examples of head- bobbing suppressions, but we did not observe, within a non-bobbing monophyletic family, examples of head-bobbing reinventions.

Why Do Birds Bob Their Heads? 110

5.3.5.2 Bobbing style in monophyletic and polyphyletic groups of birds

Interestingly, within the aves, generally monophyletic groups share the same bobbing style, while polyphyletic groups may not share the same motion style. With very few exceptions, birds descended from a single ancestor share the same bobbing style; while birds descended from different ancestors many not.

Monophyletic groups

Columbiformes and Gruiformes orders are monophyletic taxon containing HB birds. The exception is the Kagu (Gruiformes order), a bird that always moves at high speed. Head- bobbing may not be functional at high speeds (Davies & Green, 1988). Further observations are needed to determinate whether kagus show occasional head-bobbing like seriemas (a sister group of kagu). Gruiformes order is a highly divergent monophyletic taxon. The Kagu case can be an example of head-bobbing suppression in evolution.

All fowls (Gallomorphae) show very distinctive head-bobbing behaviour, while waterfowls (Anserimorphae) do not exhibit head-bobbing. These two closely related sister groups are both monophyletic taxa that belong to the parvclass Galloanserae (Sibley & Ahlquist, 1990). Fowls are ground foraging, whereas waterfowls seem to be adapted to feed in the water. Some ducks are broad-billed and sieve water or mud to extract the small crustaceans or vegetable particles, other ducks dive in water, geese and swans eat aquatic vegetation and grain (Jonsson, 1999; Perrins, 1991). The feeding habits of the waterfowls, together with a common ancestor, can explain the lack of head-bobbing in anserimorphes, this is an example of divergent evolution. Furthermore, the lack of head-bobbing in waterfowls may be an example of head-bobbing suppression in waterfowls or independent evolution of head-bobbing in fowls.

Polyphyletic groups

Within the Parvclass Coracidae species the following families were observed: Ground Hornbills (NB), Typical hornbills (BB), Hoopoes (HB), and Ground-Rollers (NB). Many gaps Why Do Birds Bob Their Heads? 111

exist in the phylogeny of Parvclass Coracidae (Sibley & Ahlquist, 1990) which may be a polygenetic group of birds. These may explain the different motion patterns observed within Coracidae Parvclass.

Cuculiformes family is an ancient lineage with substantial genetic diversity among the living groups, turacos may be the nearest living relative. That may explain why cuckoos and turacos alternate between HB, BB, or both. The only exception is the Greater Roadrunner, a bird which is never observed walking at slow motion. HB and BB behaviour appear often associated in closely related birds.

Traditionally, the former Ciconiiformes order was divided into two orders: Ciconiiformes and Pelecaniformes (Sibley & Ahlquist, 1990). Herons, storks, ibises, and allies which were classified as Ciconiiformes, are head-bobbing birds. The former Pelecaniformes order contained totipalmate birds like pelicans, boobies, gannets, cormorants, anhingas, frigate birds, and tropicbirds. Apart from the non conclusive observations in anhinga, all totipalmate birds (all four toes connected by a web) are NB birds. The traditional order Pelacaniformes is polyphyletic and it may present the most complex and controversial group in the avian phylogeny. It is possible that some shared characters of these birds have evolved more than once, while other may have been lost during evolution (Sibley & Ahlquist, 1990). This may be an example of a polyphyletic group that due to the selection pressure was lead into a convergent evolution. Palmate and totipalmate birds adapted to live in aquatic environments are non-bobbing birds (see section 5.3.4.1).

5.3.5.3 Systematic relationship among suborder Charadrii taking in account the bobbing style

The phylogeny of Charadriides (order Ciconiiformes) is very controversial. Morphological data (Chu, 1995), DNA-DNA hybridization methods (Sibley & Ahlquist, 1990), and DNA sequence data (Ericson et al., 2003) analyses resulted in different phylogenetic trees. The morphologic and DNA-DNA hybridization analyses define 2 main groups of birds: on the one hand Laridea, Charadriodea and Chionidoidea, on the other hand Scolopacoidea and Jacanoidea (Fig. 33). In Why Do Birds Bob Their Heads? 112

addition DNA-DNA hybridization divides the first group into two subgroups Laroidea and Charadrioidea plus Chionidoidea. Nuclear DNA sequence also defines two main divisions, but associate Laridea with Scolopacoidea plus Jacanoidea as sister groups, while Charadrioidea plus Chionidoidea form an independent group.

The observation data on head-bobbing behaviour were feed onto the phylogenetic trees (Fig. 33). Three groups are defined: HB, HB* (rare or occasional head-bobbing) and NB birds. Non- bobbing species may be more closely related to species that show rare or occasional head- bobbing than to head-bobbing species. Therefore, the head-bobbing data are more concordant with the DNA-DNA hybridization phylogeny. In addition, DNA-DNA hybridization may support the observations made on Stone Curlew (Burthinidae family), classified as a non- bobbing bird. However, it can not be excluded that future observations reveal occasional or rare head-bobbing. Given that normally monophyletic groups share the same bobbing style, knowledge about the motion style may help to clarify the systematic relationship between groups of birds.

Why Do Birds Bob Their Heads? 113

A NB B NB NB NB NB NB NB NB

NB HB HB*

HB* HB* HB HB* HB* HB*

HB* HB HB NB

HB HB HB HB HB HB HB HB HB

HB

HB NB

NB C NB NB

NB NB

HB HB HB HB HB

HB

HB HB* HB* HB* HB* NB HB

Figure 33: Systematic relationships among major groups of charadrii birds based on:

A) Parsimony analyses of morphological data set (Chu, 1995).

B) Analyses of DNA-DNA hybridization data (Sibley & Ahlquist, 1990).

C) Nuclear analyses of DNA sequence data (Ericson, Envall, Irestedt, & Norman, 2003)

Each phylogenetic tree was complemented with the observed “bobbing” information. Each taxa was classified as head-bobbing (HB), occasional head-bobbing (HB*), or non-bobbing bird (NB).

Why Do Birds Bob Their Heads? 114

5.3.6 Ecological and behavioural factors under head-bobbing, body-bobbing and non- bobbing

The analysis of the ecological and behavioural factors under head-bobbing revealed that the prototypical head-bobbing bird is a ground-foraging bird, that feeds on seeds, small fruits, insects, small vertebrates, worms, and other small prey. It has low flying ability, prefers open woodlands, shores, and muddy areas to live, but is never observed in open oceans and seas. It can live either in solitary, in small or big groups. The predator pressure is normally medium.

Non-bobbing birds feed on a variety of sources, like molluscs, fruits, seeds, fish, squids, vertebrates, and carrion and they have very variable feeding habits, from foraging to hunting. They are distributed in all kinds of habitats from open lands to oceans. They are normal or good flyers and they can live in solitary, small or big groups. The predator pressure is generally low; often they are predators themselves.

The main differences between these prototypes of birds are the feeding source, feeding habits, flying abilities, and the predator pressure. Head-bobbing may have a double function, pattern discrimination at short to medium distances and predator detection. Head-bobbing may aid the detection of small stationary features such as seeds, or larger moving objects, such as predators, against textured background (Wohlschläger et al., 1993). Searching for food on the ground may increase the predator vulnerability of those birds. Furthermore, except for migrating birds, high flying abilities are not required.

Further analyses in body-bobbing birds are needed, but observations on 6 species of birds revealed that they feed on fruit, leaves, invertebrates, small vertebrates, etc. They are skilled and normal flyers with low predator pressure and prefer living in bushes and trees, in solitary or small groups. Furthermore, small birds like most of the Passeriformes, living in bushes and trees, have short legs and tend to hop rather than walk (Friedman, 1975).

Why Do Birds Bob Their Heads? 115

Head-bobbing birds may have lateralised eyes with big cyclopean fields in order to monitor predators. It has been found that the Ostrich, the Cattle egret, the Rock pigeons, the Eurasian Woodcock and three types of herons have binocular visual fields smaller than 22º and a cyclopean visual field bigger than 290º. Head-bobbing was not observed in birds with large binocular visual field and small cyclopean visual field. In comparison, non-bobbing birds’ visual field is very variable. There are birds with narrow binocular field and large cyclopean field like Mallard, and birds with large binocular field and small cyclopean visual field like Tawny owl. Mallard (Anas platyrhinchos), which have very narrow monocular field (5º) and large cyclopean field (360º), like other ducks, search for food by sieving the water. Ducks may require a large cyclopean field for searching for predators, but they do not need any mechanism, like head-bobbing, to identify small targets at short to medium distances. Tawny owl (Strix aluco) have a wide binocular fields (48º) and smaller cyclopean fields (210) (Martin et al., 2001; Martin & Coetzee, 2004; Martin & Katzir, 1993, 1994, 1995, 1999, 2000).

5.3.7 Interim Summary

Half of the modern birds may be head-bobbing birds, while the rest of the birds are equally divided between body-bobbing and non-bobbing birds. Head-bobbing and body-bobbing behaviours appear together in many families. They could be equivalent or related behaviours playing similar roles. Generally, all species within a family share the same “bobbing” behaviour. Detailed analyses of exceptions may provide useful data to understand the functional significance of head-bobbing. On the one hand, head-bobbing may be a characteristic inherited from a common ancestor of modern birds. During evolution, different selective pressures may have caused a loss of the head-bobbing behaviour. On the other hand, head-bobbing is a highly distributed trait in the avian phylogeny. Therefore, suppressions, and independent evolution of head-bobbing may easily have occurred during evolution. Comparisons between different phylogenetic trees revealed that birds closely related have the same bobbing behaviour. Furthermore, monophyletic groups normally share the same bobbing style, while polyphyletic groups may show different bobbing styles. Head-bobbing birds are normally ground foraging birds seeking for small feeding sources like insects, grains, worms, Why Do Birds Bob Their Heads? 116

etc. They are bad to normal flyers with medium predator pressure and prefer to live in open woodlands, marshes, and muddy areas. In addition, they may have lateralised eyes with narrow binocular fields and large cyclopean fields. Non-bobbing birds have less remarkable characteristics. They are good flyers, have low predator pressure, are highly distributed, and feed on a variety of food. Head-bobbing may be an earlier mechanism to solve visual demands, like food and predator detection. As soon as the visual demands are not present, or more efficient mechanisms appear, head-bobbing may decrease or be suppressed. Complex feeding habits may require more sophisticated and efficient visual mechanism like binocularity, which may replace head-bobbing.

General Discussion 11 7

6. GENERAL DISCUSSION

The flying ability together with the feeding behaviours and predator monitoring are probably among the greatest challenges for the avian visual system. Birds may need to solve different visual demands leading to different visual specialisations. Most birds have lateralised eyes, with a narrow binocular field and a broad monocular field (Martin, 1993; Martin & Katzir, 1999; Shimizu & Karten, 1993). Approximately half of the living birds are head-bobbing birds and it is well known that head-bobbing is visually controlled (Friedman, 1975b; Frost, Cavanagh, & Morgan, 1988). It may be a monocular mechanism to solve visual environmental demands in birds with lateralised eyes.

6.1 Intraocular and interocular transfer of information

The pigeon retina is divided into the dorso-temporal binocular red field and the remaining monocular yellow field. Two independent systems process the visual information in pigeons. The entire retina projects to the contralateral optic tectum and continues via the diencephalic nucleus rotundus to the entopallium (tectofugal pathway). The monocular area projects to the contralateral geniculate thalamic nucleus and continues to the wulst (thalamofugal pathway).

The first part of this work investigated how information retrieved from both eyes and in different retinal areas is generalised in the brain of walking pigeons. Intraocular transfer of information between the frontal red field and the lateral yellow field was tested in both directions: from frontal to lateral and vice-versa. Interocular transfer of information between the yellow visual fields of both eyes was also tested. The second part of the work examined the role of head-bobbing in pattern discrimination in walking pigeons by presenting the stimuli either during the hold or the thrust phase. The third part elaborated a list of head-bobbing and non head-bobbing birds. It investigated the development of head-bobbing in the evolution, as well as the ecological and behavioural characteristics of head-bobbing and its functional significance.

General Discussion 118

6.1.1 Intraocular transfer of information

In the first experiment it was found that pigeons are able to learn a pattern discrimination task in the lateral visual field by shifting the stimulus presentation from frontal to lateral positions in consecutive steps. Previous studies showed that pigeons are capable of pattern discrimination and stimulus detection in the lateral visual field (Bloch & Martinoya, 1982b; Goodale & Graves, 1982; Mallin & Delius, 1983; Remy & Emmerton, 1991b; Roberts et al., 1996). However, the only precedent using a similar methodology in freely walking pigeons, moving the stimulus presentation in consecutive steps from frontal to lateral direction, failed to train the birds (Nye, 1973). An impressive amount of training to learn the task was required in the current experiment. The increase of the distance between the pecking key and the stimulus presentation can constitute a handicap, increasing the difficulty of the task (Mallin & Delius, 1983).

A dramatic decrease of performance to chance level was found at 45º, by moving the stimulus presentation from the frontal to the lateral visual field. In addition, at 45º pigeons needed twice the amount of trials than in other positions to achieve a criterion of 70% of correct responses in four consecutive sessions. During the experiment, the stimulus presentation occurred right after the pigeon crossed a light barrier situated at 118 cm to the stimulus. The frontal binocular field of the pigeon in the horizontal plane has a maximum extension of 27º (Martin & Young, 1983). The edge of the frontal binocular field, when the observation point is set at 118 cm in the centre of the arena, corresponds to a stimulus located at 40º in the experimental arena. A 5x5 cm stimulus presented at 45º in the experimental arena falls entirely within the lateral visual field (Fig. 13). Therefore at 45º the pigeons may switch from processing the stimuli with the tectofugal to the thalamofugal pathway. Information transfer between the frontal and the lateral visual field was not found in freely walking pigeons. The finding confirm previous investigations in head-fixed and non walking pigeons (Mallin & Delius, 1983; Nye, 1973; Remy & Watanabe, 1993; Roberts et al., 1996).

General Discussion 119

The pigeons were also tested in the lateral-frontal direction. Three of the birds were tested in the initially naive visual field (see section 3.2), by moving the stimulus location from the lateral to the frontal visual field in ten consecutive steps. Three other pigeons that did not learn the task in the opposite visual field of the initially trained one were exclusively tested in the frontal visual field (0º). It should be considered that the initial training in the frontal visual field took place at least one year earlier. Contradictory results were found: on the one hand fewer amounts of trials were needed in the lateral-frontal direction than in the frontal-lateral direction. In contrast, no differences were found in the performance of the pigeons in the first 20 trials at each position between the frontal to lateral direction and vice versa. On the other hand the pigeons exclusively tested at 0º, having previous experience in the frontal visual field, did not remember the task and did not transferred information. They needed almost the same amount of trials to relearn the task as in the initial learning.

Taking all data together, probably no transfer or very poor transfer of information occurred from the lateral to the frontal visual field. Similarly, a poor intraocular transfer from lateral to frontal position was observed in a colour discrimination task (Mallin & Delius, 1983). However, other studies found intraocular transfer of information, from lateral to frontal position (Remy & Emmerton, 1991b; Roberts et al., 1996). Intraocular transfer of information from the lateral to the frontal visual field may be task dependent. Intraocular transfer was very poor or not observed in tasks with a more clear memory component, like colour discrimination (Mallin & Delius, 1983) and pattern discrimination. In contrast, intraocular transfer was observed in tasks without a clear memory component, like stimulus detection (Remy & Emmerton, 1991b) and matching to sample (Roberts et al., 1996). In natural conditions, the information received within the yellow field system may be transferred to the red visual field system in order to switch attention to interesting or important environmental stimuli, whereas transfer of memory from the frontal to the lateral visual field may not be needed. Projections from the visual wulst (thalamofugal pathway) into the optic tectum (tectofugal pathway) can mediate this transfer of information (Bagnoli et al., 1980; Karten et al., 1973; Miceli et al., 1987).

General Discussion 120

In conclusion, pigeons are not capable of intraocular transfer from the frontal to the lateral visual field, whereas information transfer from the lateral to the frontal visual field may depend on the task. Further investigations are required to confirm this last observation.

6.1.2 Interocular transfer of information

Interocular transfer of information between the two yellow fields was investigated. Pigeons initially trained in the LVF were tested in the RVF, and vice versa. In five of the six pigeons no traces of interocular transfer of information were observed. Only one pigeon was capable of performing the task in the untrained visual field. These data do not contradict any of the two hypothesis explaining interocular transfer: the “sensorimotor integration” hypothesis (Watanabe, 1986) and the “retinal locus” hypothesis (Goodale & Graves, 1982; Levine, 1945b; Mallin & Delius, 1983). Interestingly, the only pigeon capable of information transfer, showed in experiment 4 a different motion strategy than the other birds. It was the only bird that instead of varying the overall motion speed thorough the experiment was varying the head- bobbing frequency (see section 4.2.2). It may happen that only relevant information perceived within the yellow fields that triggers an instantaneous motor response would be transferred.

Right eye/left hemisphere dominance in pattern discrimination presented in the frontal binocular field has been observed in birds (Güntürkün, 1985, 1997a; Güntürkün & Bohringer, 1987). In contrast, in the intraocular transfer experiment, two animals trained in the right visual field never learned the task beyond 45º. Furthermore, in the interocular transfer experiments (excluding pigeon 51 that was capable of transfer) after a long lasting training, birds initially trained in the RVF relearned the task in the LVF, whereas birds initially trained in the LVF (except the animal capable of information transfer) did not relearn the task in the RVF. A weak left eye/right hemisphere dominance was observed in a symbolic delayed matching to sample task in pigeons (Roberts et al., 1996). Other left eye/right hemisphere dominances were observed in different tasks like predator detection, social recognition in chicks, and novel stimuli detection (Evans et al., 1993; Rogers, 2000; Vallortigara et al., 2001). It has been proposed that the left eye/right hemisphere may be dominant in the lateral visual field which is General Discussion 121

mainly focused at long distances, whereas the right eye/left hemisphere may be dominant in binocular vision which is specialised in short distance visual processes (Evans & Evans, 1999; Evans et al., 1993; Rogers, 2000; Vallortigara et al., 2001). A lateralised brain may allow dual attention to short distance tasks like feeding (using the right eye/left hemisphere system) and long distance tasks like vigilance for predators (left eye/right hemisphere system) (Rogers, 2000).

The distance at which a stimulus is observed may not be the only factor for a visual asymmetry. In fact, there are many evidences indicating that visual asymmetries may be task- dependent (Clayton & Krebs, 1994b; Güntürkün, 1997a; Nottelmann et al., 2002; Tommasi & Vallortigara, 2001; Vallortigara, 2000; von Fersen & Güntürkün, 1990). Hemispheric dominance may depend on both the distance at which the information is processed and on the task itself. Visual demands, like vigilance, which appeared during evolution more often at long distances, may be processed preferentially with the left monocular field/right hemisphere. In contrast, other visual demands, like feeding, which during evolution appeared at short distances, may be processed preferentially with the right binocular field/left hemisphere. If this is the case, the key for an asymmetry may not be the visual field area were a visual task is perceived, but the visual field area at which a task has been more often perceived during the evolution. For example, left eye specialisation was found for spatial distribution of attention in the frontal visual field (Diekamp et al., 2005). The left eye/right hemisphere may be specialised in processing easy shapes at long distances, like silhouettes of predators.

Asymmetries in the birds’ visual system may serve to increase the computational speed of certain processes by concentrating them into one hemisphere (Güntürkün et al., 2000; Rogers et al., 2004). To the ecological advantages of being capable to attend to both predators and feeding sources (Rogers, 2000), the computational advantages of processing information in one hemisphere should be added. Most probably the asymmetric avian brain is a consequence of the interaction between the ecological and computational advantages mediated by the anatomical substrate.

General Discussion 122

6.2 Pattern recognition during head-bobbing

Around 50% of the birds of the world are head-bobbing birds. By excluding birds with eyes located frontally, the proportion of head-bobbing birds is increased. Having lateralised eyes can be a prerequisite for head-bobbing. Head-bobbing may aid pattern recognition, movement detection and distance perception in birds with lateralised eyes.

A head-bobbing cycle is divided into a hold phase, when the head is stabilized, and a thrust phase, when the head is moved forward (Davies & Green, 1988; Friedman, 1975b; Frost, 1978; Troje & Frost, 2000). In addition, during the hold phase, saccadic eye movements have rarely been observed (Bloch et al., 1987; Pratt, 1982; Wohlschläger et al., 1993). As a consequence, it is broadly accepted that head-bobbing may aid in pattern recognition by stabilizing the retinal image. Pattern recognition in pigeons was tested by presenting the shapes used in the previous experiments during the hold phase, thrust phase and asynchronous. The pigeons were tested in both lateral visual fields, one lateral visual field, and frontal visual field. Unexpectedly, no significant differences were found between presenting the stimuli within the thrust or the hold phase in any stimulus location. The presentation of the stimuli within the thrust phase was strictly controlled. Therefore, pigeons were capable of pattern recognition during the entire head-bobbing cycle.

The presented stimuli were easy to detect, in fact they were two 5x5 cm white shapes on a black background. It is plausible that pigeons may have difficulties to discriminate small features hidden in a low contrast complex background, like a grain hidden in the yellow grass of the savannah at close distances. Objects located at long distances are less affected by the optic flow provoked by the birds’ motion. In addition, image stabilization could be achieved in some birds rotating the eyes, although in relation to the skull, birds have large eyes that limit eye mobility (Martin and Katzir, 1995). The eye mobility of many birds may be enough to counteract the retina slip at medium to long distances, but the blur of small features near-by may require the contribution of head-bobbing.

General Discussion 123

The motions of pigeons during the hold and thrust phase, and during asynchronous presentations was analysed when the stimuli were presented either in both lateral visual fields or in the frontal field. For the hold and thrust phase, the duration, the amplitude, and the speed was calculated for the frontal and for the lateral visual fields presentations. Significant differences between the motion speeds were found between the frontal and the lateral visual fields. Two different strategies were observed. The majority of the pigeons varied the overall speed of the motion by decreasing the duration of the head-bobbing cycle, while keeping constant the distance. It was observed that an increase of the speed resulted in a reduced hold phase. This phenomenon has previously been observed in pigeons (Davies & Green, 1988). Significantly shorter hold phase durations and faster speed were observed in the frontal visual field with respect to the lateral visual fields.

Interestingly, one bird used a different strategy varying the head-bobbing frequency: it kept constant the motion speed across the conditions, varying the amplitude and the duration of the head-bobbing cycles. The amplitude and duration of the head-bobbing cycles were shorter in the frontal visual field presentations than in the lateral visual field presentations. Although this strategy only appears in one of the birds, it may be a common strategy of birds. An extreme example occurs when static pigeons and other birds are found doing head-bobbing while observing their environment. Variations in the head-bobbing cycle duration independently from the motion speed can be useful to increase the optic flow, and therefore enhance distance calculations.

In conclusion, pigeons are capable of recognizing easy features during the thrust phase. The capability of discriminating relevant stimuli like predators during the entire head-bobbing cycle may increase the survival rate of animals. Pigeons can modify the head-bobbing speed and frequency depending on the stimulus location in the visual field. In addition, different visual inputs may result in different head-bobbing styles.

General Discussion 124

6.3 Why do birds bob their heads?

In the last study, the ecological and behavioural factors under head-bobbing were investigated, and the development of head-bobbing in evolution was analysed. Field observation of 322 species of birds was done in their natural environment and/or semi-captivity conditions. It was found that around half of the living birds may have head-bobbing, around 25% are body- bobbing birds (hopping birds), and the rest are non-bobbing birds. Nocturnal birds and predator birds are normally non-bobbing birds, whereas birds with more lateralised eyes can be head- bobbing birds, non-bobbing birds, or body bobbing birds.

The question about the functional equivalence between body-bobbing birds and head-bobbing birds remains unsolved. The analysis of the birds’ collections suggested that in six families and the entire Passeriformes order, body-bobbing and head-bobbing were alternative behaviours. It may well happen that body-bobbing have a double function: it may enhance the mobility of small birds between tree branches and bushes and contribute to visual perception.

Generally, all species of birds belonging to a family share the same bobbing styles. The analysis of the few observed exceptions revealed that the feeding source, the speed of the motion, and the peak of activity, may be important factors for head-bobbing. Furthermore, all species of birds observed within a monophyletic group share the same bobbing style. Head- bobbing may help to distinguish monophyletic groups of birds from polyphyletic groups of birds. Furthermore, the bobbing behaviour can serve as a cue for understanding the systematic relationship among a group of birds. The analyses of birds with occasional or rare head- bobbing pointed out that head-bobbing may serve to distinguish complicated patterns hidden on the floor. For example, in ostriches head-bobbing may aid to detect a threatening lion in the long grasses of the savannah, whereas in plovers and lapwings head-bobbing may aid to find small invertebrates camouflaged in the mud. In addition, head-bobbing may occur to enhance optic flow in order to calculate distances through motion parallax. For example, bustards have been observed performing a very peculiar head-bobbing with long thrust phases while walking very slowly searching for food.

General Discussion 125

Head-bobbing may have evolved initially in a common ancestor of all living birds. It may appear early in the evolution as a mechanism to aid visual processing. Most recently evolved species like birds of prey, are non head-bobbing birds with eyes located frontally. Both reptiles and birds have lateralised eyes (Husband & Shimizu, 2001) and birds have descended from reptiles (Shimizu & Karten, 1991). It has been observed that head-movements in lizards and African chameleon are related to visual processing (Flanders, 1985, 1988; LeBas & Marshall, 2000; Torr & Shine, 1994). However, it cannot be discarded that head-bobbing have evolved independently several times in evolutions. We found evidences for both suppressions and independent evolutions of head-bobbing by feeding our observations in Sibley and Ahlquist’s taxonomic tree (see section 5.3.5).

The prototype head-bobbing bird feeds on seeds, fruits, and small feeding of sources. It prefers small sized food over big sized food and seeks for food by wading or ground foraging. It lives either in woodlands or shallow waters. It can be either an awkward flyer or a normal flyer. Actually, head-bobbing birds are normally worse flyers than non-bobbing birds. Head-bobbing birds suffer higher predator pressure than non-bobbing birds. Although it is commonly accepted that predator pressure is related to the flock size (Alcock, 1997), preferences were observed between living in solitary, small groups, or flocks neither in head-bobbing nor in non- bobbing birds (Fig. 31 and Appendix C: Table C-3). Non-bobbing birds have very variable sources and feeding habitats. However, foraging-birds are rarely non-bobbing birds. They are broadly distributed in different kind of habitats. Non-bobbing birds are generally elegant and skilled flyers with low predator pressure. As in head-bobbing birds, no preferences in group size were observed. Further analyses in body-bobbing birds are needed, but the observations revealed that they feed on fruits, leaves, invertebrates, small vertebrates, etc. They are small birds that prefer bushes and trees to live. They live normally in solitary or small groups and they are skilled and normal flyers with low predator pressure.

The ultimate causes of head-bobbing in evolution may be the feeding habits and the predator pressure. Feeding on small targets, hard to detect lying on the floor may require both accurate pattern detection and distance calculation. Ground-foraging birds may be more exposed to predators, therefore a visual mechanism, like head-bobbing, compatible with a large cyclopean General Discussion 12 6

visual field is required. Flying abilities may not be a demand for these ground feeding birds, although it would develop in migrating birds like pigeons, cranes, and storks.

A prerequisite for head-bobbing behaviour is probably the lateral position of the eyes in the skull. Birds with lateralised eyes as well as reptiles, have a more developed tectofugal pathway, whereas predator birds (non-bobbing birds) have a more developed thalamofugal pathway (Husband & Shimizu, 2001; Karten & Shimizu, 1989; Shimizu & Karten, 1991; Shimizu& Karten, 1993). The need for complex and fast visual information analyses, as a consequence of more complex feeding habits could explain the specialisation of the thalamofugal pathway to the frontal visual field in birds of prey (Güntürkün & Hahmann, 1999). A relationship between the lack of head-bobbing and the thalamofugal pathway development should be considered. Although, several examples of non-bobbing species with lateralised eyes were observed. These birds use methods of feeding that do not require a high visual specialisation. For example, many species of ducks and flamingos search for food by filtering the water. Head-bobbing may be an early mechanism for solving visual demands, like detection of small hidden features, as soon as these demands are absent or are solved in a more efficient way, head-bobbing would occur rarely or even disappear

6.4 Summary

Intraocular transfer of information was not observed from the frontal visual field to the lateral visual field. From the lateral to the frontal visual field, information transfer may be task- dependent. These results may be a consequence of two independent visual systems in the pigeon brain: the tectofugal pathway, receiving information from the entire retina, and the thalamofugal pathway, processing information from the lateral visual field. Interocular transfer of information between the two yellow fields was rarely observed. Interocular transfer between the yellow fields may occur when a relevant visual input for the animal is accompanied by an instantaneous motor response. Pigeons have difficulties to learn a pattern discrimination task in the lateral visual field. It may be explained by the distance between the response key and the stimulus. The right eye/left hemisphere may be specialised in processing stimuli located at General Discussion 12 7

short to medium distances usually perceived within the frontal binocular field, whereas the left/eye right hemisphere may be specialised in processing far away stimuli usually perceived within the lateral visual field. Pigeons are capable of pattern discrimination during the thrust phase of head-bobbing. Being able to discriminate relevant visual information during the entire head-bobbing cycle may constitute an ecological advantage. The analysis of the head motion revealed that pigeons walked slower when the stimuli were presented in the lateral visual field than in the frontal visual field. Increases of the motion speed result into shorter hold phases. Head-bobbing parameters may change according to the characteristics of the visual target. It may be a visual mechanism in birds with lateralised eyes that appeared early in evolution to solve visual demands like feature recognition and distance calculation. Head-bobbing may also have other visual contributions. Around half of the modern birds are head-bobbing birds, one fourth of them are body-bobbing birds, and the remaining birds are non-bobbing birds. The importance of head-bobbing behaviour may be underestimated until now. Head-bobbing has been observed broadly distributed in the taxonomy of Aves. Head-bobbing may have evolved first in a common ancestor of the entire class Aves, but several independent evolutions and suppressions should be considered. Species of birds closely related normally share the same bobbing style. Ground foraging, low flying abilities and medium to high predator pressure are correlated with head-bobbing behaviour. The development of stereopsis, nocturnal vision, or a lack of the necessary visual challenge may be related with a suppression of head-bobbing.

Appendix 128

APPENDIX

A. Intraocular and interocular transfer

Pigeon 0º 14º 23º 29º 36º 41º 45º 48º 61º 76º 95º 251-LVF 80 80 75 85 85 80 60 75 80 85 85 259-LVF 75 95 95 75 70 85 55 78 65 80 84 512-LVF 70 60 70 50 55 65 65 50 60 70 65 988-LVF 80 70 70 74 56 80 60 75 80 85 85 51-RVF 90 75 70 60 65 65 65 60 78 70 85 321-RVF 85 65 45 45 55 55 65 75 70 85 68 333-RVF 85 84 84 70 82 75 46 56 73 55 73 347-RVF 85 90 84 75 75 75 50 85 75 100 55 Av. Total 81,2 77,4 74,1 66,8 67,9 72,5 58,3 69,2 72,7 78,7 75 Av. LVF 76,2 76,2 77,5 71 66,5 77,5 60 69,5 71,2 80 79,7 Av. RVF 86,2 78,5 70,8 62,5 69,3 67,5 56,3 69 74 77,5 70,2 246-LVF 70 70 70 90 85 60 60 246-RVF 75 60 70 65 80 60 50 259-RVF 90 52 65 52 73 70 60

Table A-1: Percentage of correct responses for each pigeons in the first 20 trials at different positions of the screen. In blue appears the percentage of correct responses of the pigeons in the last 20 trials or the initial training. Half of the birds were tested in the left visual field (LVF), whereas the other half was tested in the right visual field (RVF).

Pigeon 14º 23º 29º 36º 41º 45º 48º 61º 76º 95º 251-LVF 4 4 11 5 4 14 4 4 6 7 259-LVF 4 4 7 8 4 19 4 13 4 15 512-LVF 5 4 5 10 14 11 10 5 4 7 988-LVF 4 4 4 14 4 25 12 7 4 4 51-RVF 4 8 5 18 12 20 4 4 5 11 321-RVF 5 10 25 5 10 8 4 4 4 14 333-RVF 4 4 5 4 4 17 28 4 12 11 347-RVF 4 4 4 4 4 7 4 5 4 5 Av. Total 4,25 5,25 8,25 8,50 7,00 15,13 8,75 5,75 5,38 8,00 Av. LVF 4,25 4,00 6,75 9,25 6,50 17,25 7,50 7,25 4,50 8,25 Av. RVF 4,25 6,50 9,75 7,75 7,50 13,00 10,00 4,25 6,25 7,75 246-LVF 8 4 9 4 6 30 246-RVF 4 9 19 10 12 30 259-LVF 26 9 4 4 12 30

Table A-2: Number of sessions needed to achieve the criterion at different positions of the screen (1 session = 20 trials). Half of the birds were tested in the left visual field (LVF), whereas the rest of the birds were tested in the right visual field (RVF). Appendix 129

S.Position 14º 23º 29º 36º 41º 45º 48º 61º 76º 95º % of C.R. 14º n.s. .014 .034 n.s. .032 .n.s. n.s. n.s. n.s. 23º n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. 29º .014 n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. 36º .034 n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. 41º n.s. n.s. n.s. n.s. .046 n.s. n.s. n.s. n.s. 45º .032 n.s. n.s. n.s. .046 n.s. .011 .011 .009 48º n.s. n.s. n.s. n.s. n.s. n.s. n.s. .009 n.s. 61º n.s. n.s. n.s. n.s. n.s. .011 n.s. n.s. n.s. 76º n.s. n.s. n.s. n.s. n.s. .011 0.09 n.s. n.s. 95º n.s. n.s. n.s. n.s. n.s. .009 n.s. n.s. n.s.

Table A-3: Post hoc analysis of the percentage of correct responses for the factor Screen Position. Statistical test were considered non significant (n.s.) if p>.05

S.Position 14º 23º 29º 36º 41º 45º 48º 61º 76º 95º Trials 14º n.s. n.s. n.s. n.s. .004 .n.s. n.s. n.s. .011 23º n.s. n.s. n.s. n.s. .006 n.s. n.s. n.s. .018 29º n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. 36º n.s. n.s. n.s. n.s. .009 n.s. n.s. n.s. n.s. 41º n.s. n.s. n.s. n.s. .036 n.s. n.s. n.s. n.s. 45º .004 .006 n.s. .009 .036 n.s. .005 .004 n.s. 48º n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. n.s. 61º n.s. n.s. n.s. n.s. n.s. .005 n.s. n.s. 0.29 76º n.s. n.s. n.s. n.s. n.s. .004 n.s. n.s. n.s 95º .011 .018 n.s. n.s. n.s. n.s. n.s. .029 n.s.

Table A-4: Post hoc analysis of the number of trials required to achieve the criterion for the factor Screen Position. Statistical test were considered non significant (n.s.) if p>.05

Appendix 130

100 100 100 100 95 90 90

85 85 85 82 80 80 80 80 80 80 78 75 70 70 70 s

se 60 n 60 60 60 o

sp 50 Learning e r t Trained visual field

ec 40 r 40 40

r Naive visual field co

f 30 o

% 20 20

10

0

51 59 88 12 21 47 51 33 F2 F2 F9 F5 F3 F3 VF F3 LV LV LV LV RV RV R RV Visual field and pigeon

Figure A-5: Percentage of correct responses for each pigeon in 220 learning trials, 10 catch trials in the trained visual field and 10 catch trials in the naive visual field (Fig. 21A and 21B).

240 200 160 120

m) 80 m

( 40 0 ace n -40 0 100 200 300 400 500 600 st i

d -80 Y -120 -160 -200 -240 X distance (mm)

Table A-6: Sample of pigeon 251 walking trajectory towards the pecking keys when the screen was located in the frontal visual field at 0º. The pecking keys are represented by two isolated blue points. The red circle represents the point at which the trajectory started to change. The centre of the coordinate system was located in the middle of the arena at the pecking key level

Appendix 131

240 200 160 120 80

mm) 40

ce ( 0 an -40 0 100 200 300 400 500 600 st i

d -80 Y -120 -160 -200 -240 X distance (mm)

Table A-7: Sample of pigeon 251 walking trajectory towards the pecking keys when the screen was located in the corner at 45º. The pecking keys are represented by two isolated blue points. The red circle represents the point at which the trajectory started to change. The centre of the coordinate system was located in the middle of the arena at the pecking key level

Appendix 132

B. Pattern recognition during head-bobbing

600 600

500 500

482 469 456 400 400

338 300 300 308 295

250 235 200 200 211 187 188

144 y y 100 100 Std. Dev = 5,21 Std. Dev = 4,34 enc 74 Mean = 9,1 uenc Mean = 10,6 62 eq equ r 39 r 0 N = 1907,00 0 32 N = 1907,00 F F 2,5 7,5 12,5 17,5 22,5 2,5 7,5 12,5 17,5 22,5 5,0 10,0 15,0 20,0 25,0 5,0 10,0 15,0 20,0 25,0

Number of frames: hold phase Number of frames: thrust phase

Figure B-1: Distributions used for presenting the stimuli in the random catch trials and during the learning trials. The temporal resolution of the system is 120 Hz. one frame is equivalent to 8,33 ms. The distribution for the hold phase had a mean of 78 ms (9,4 frames) and a standard deviation of 47 ms (5,71 frames). The distribution for the thrust phase had a mean of 88 ms (10.6 frames) and a standard deviation of 36 ms (4,34 frames).

Appendix 133

Visual field Presentation Phase Vf.*Pr. Vf.*Ph. Pr.*Ph. Vf.*Pr.*Ph. (Vf) (Pr) (Ph) Pigeon Parameter F (1,29) Sig. F(2,58) Sig. F (1,29) Sig. F(2,58) Sig. F(1,29) Sig. F(2,58) Sig. F(2,58) Sig.

Duration 66.982 .000 .311 .734 1083.8 .000 1.222 .302 10.353 .003 1.461 .241 .135 .874 51 Distance 2.208 .148 1.037 .361 2313.6 .000 .509 .604 2.774 .107 1.144 .326 .524 .595 Speed 8.530 .007 .175 .840 3258.8 .000 .430 .653 1.669 .207 .205 .815 .624 .540

Duration 24.982 .000 1.492 .236 4714.4 .000 .124 .884 13.271 .001 1.164 .321 2.518 .092 251 Distance 42.292 .000 4.970 .011 4209.3 .000 1.036 .363 34.234 .000 5.983 .005 .863 .428 Speed 1.503 .233 .056 .945 5046.8 .000 .084 .920 4.419 .077 .646 .529 .490 .616

Duration 4.785 .037 .338 .714 3625.1 .000 .319 .728 .003 .954 .277 .759 1.276 .287 259 Distance .263 .612 2.615 .082 1849.7 .000 .813 .449 .624 .436 2.761 .072 .824 .444 Speed 5.489 .026 2.378 .102 3926.4 .000 .168 .845 24.442 .000 3.349 .042 .303 .740

Duration 8.190 .008 1.060 .353 1777.1 .000 .013 .987 9.219 .005 .228 .797 1.121 .333 347 Distance .028 .869 .980 .381 3346.6 .000 .352 .705 .099 .756 .948 .393 .350 .706 Speed 28.100 .000 8.063 .001 9528.3 .000 7.139 .002 29.467 .000 15.371 .000 15.219 .000

Duration .001 .981 .155 .856 1728.0 .000 .085 .919 1.126 .298 .283 .755 .184 .832 512 Distance .921 .346 .040 .960 893.63 .000 1.224 .272 .275 .605 .269 .765 .386 .682 Speed .911 .349 3.809 .029 5302.8 .000 6.668 .003 .077 .783 8.645 .001 10.980 .000

F (1,4) Sig. F(2,8) Sig. F (1,4) Sig. F(2,8) Sig. F (1,4) Sig. F(2,8) Sig. F(2,8) Sig.

Duration 9.654 .036 .058 .944 4.120 .112 .852 .462 3.984 .117 .061 .941 .177 .841 All Distance 2.568 .184 .185 .835 813.954 .000 1.335 .316 1.651 .268 .377 .765 1.460 .288 Speed 2.003 .230 .817 .475 586.841 .000 3.510 .080 1.981 .232 1.476 .285 3.544 .079

Table B-2: Three within-subject 2*3*2 ANOVA analyses for the parameters duration, distance and speed were performance for each pigeon. The first factor is the factor visual field with two levels: lateral visual fields and frontal visual field. The second factor is the factor presentation with 3 levels: hold phase presentation, thrust phase presentation and random presentation. The third factor is the factor phase with two levels: hold phase and thrust phase values of the parameters. The same data analysis was calculated for the average duration, distance and speed for all pigeons. Significant effects (p<.05) are marked in blue.

Appendix 134

C. Why birds bob their heads?

Behavioural observations following Clements’ 2000 checklist (2004 update) Order Family Family members Common name Scientific name Source Bobbing Struthioniformes Struthionidae Ostriches Ostrich Struthio camelus Video recording HB* Rheiformes Rheidae Rheas Greater Rhea Rhea americana Video recording NB Lesser Rhea Rhea pennata Video recording NB Casuariiformes Casuariidae Cassowaries Northern Cassowary Causarius unappendiculatus Video recording HB Dwarf Cassowary Casuarius bennetti Video recording HB Dromaiidae Emus Australian Emu Dromaius novaehollandiae Video recording HB* Dinornithiformes Apterygidae Kiwis Tinamiformes Tinamidae Tinamous Elegant Crested Tinamou Eudromia elegans Video recording HB Sphenisciformes Spheniscidae Penguins Humboldt Penguin Spheniscus humboldti Field observation NB Gaviiformes Gaviidae Loons Common Loon Gavia immer Video recording NB Podicipediformes Podicipedidae Grebes Great Crested Grebe Podiceps cristatus (Dagg, 1977) HB* (ws) Little Grebe Tachybaptus ruficollis Field observation HB* (ws) Procellariiformes Diomedeidae Albatrosses Black-footed Albatross Diomedea nigripes Field observation OHM Procellariidae Shearwaters and Petrels Hydrobatidae Storm-Petrels Pelecanoididae Diving-Petrels Pelecaniformes Phaethontidae Tropicbirds Red-tailed Tropicbird Phaeton rubricauda (IBC,2005) BB? Pelecanidae Pelicans Great White Pelican Pelecanus onocrotalus Field observation NB Dalmatian Pelican Pelecanus crispus Video recording NB Sulidae Boobies and Gannets Northern Gannet Morus bassanus Video recording NB Phalacrocoracidae Cormorants Great cormorant Phalacrocorax carbo Field observation NB Guanay cormorant Phalacrocorax bougainvillii Video recording NB Anhingidae Anhingas and Darters Anhinga Anhinga anhinga (IBC,2005) HB? Fregatidae Frigatebirds (IBC,2005) Ciconiiformes Ardeidae Herons, Egrets and Bitterns Gray Heron Ardea cinerea (Fujita, 2003) HB (wl) Great Blue heron Ardea herodias Field observation HB Black Headed Heron Ardea melanocephala Video recording HB Little Egret Egretta garzetta (Fujita, 2003) HB Intermediate Egret Egretta intermedia Video recording HB Appendix 135

Order Family Family members Common name Scientific name Source Bobbing Snowy Egret Egretta thula Field observation HB Western Reef-Heron Egretta gularis Video recording HB Little Bittern Ixobrychus minutus Video recording HB Cattle Egret Bubulcus ibis Video recording HB Boat-billed Heron Cochlearius cochlearius Field observation HB Scopidae Hamerkop Hamerkop Scopus umbretta Video recording HB Ciconiidae Storks White Stork Ciconia ciconia Video recording HB Woolly-necked Stork Ciconia episcopus Video recording HB Abdim's Stork Ciconia abdimii Video recording HB Oriental Stork Ciconia boyciana Video recording HB African Openbill Anastomus lamelligerus Video recording HB Marabou Leptoptilos crumeniferus Video recording NB? Yellow-billed Stork Mycteria ibis Video recording HB ephippiorhynchus Saddle-billed Stork senegalensis Video recording HB Balaenicipitidae Shoebill Shoebill balaeniceps rex Video recording NB Threskiornithidae Ibises and Spoonbills Scarlet Ibis Eudocimus ruber Video recording HB American White Ibis Eudocimus albus Video recording HB Sacred Ibis Threskiornis aethiopicus (Dagg , 1977) HB Glossy Ibis Plegadis falcinellus Field observation HB Indian Black Ibis Pseudibis papillosa Field observation HB Roseate Spoonbill Platalea ajaja Video recording HB Eurasian Spoonbill Platalea leucorodia Video recording HB Madagascar Crested Ibis Lophotibis cristata Video recording HB Hadada Ibis Bostrychia hagedash Video recording HB Phoenicopteriformes Phoenicopteridae Flamingos Greater Flamingo Phoenicopterus roseus Video recording NB Lesser Flamingo Phoenicopterus minor Field observation NB Chilean Flamingo Phoenicopterus chilensis Video recording NB Anseriformes Anhimidae Screamers Southern Screamer Chauna torquata Field observation NB (ws) Anatidae Ducks, Geese and Swans Greylag Goose Anser anser Field observation NB (ws) Bar-headed Goose Anser indicus Field observation NB (ws) Chiloe Wigeon Anas sibilatrix Field observation NB (ws) Specled Teal Anas flavirostris Field observation NB (ws) Red-billed Duck Anas erythrorhyncha Field observation NB (ws) Hottentot Teal Anas hottentota Field observation NB (ws) Gadwall Anas strepera Field observation NB (ws) Appendix 136

Order Family Family members Common name Scientific name Source Bobbing Mallard Anas platyrhynchos Field observation NB (ws) Northen Pintail Anas acuta Field observation NB (ws) Yellow Billed Pintail Anas georgica Field observation NB (ws) Common Teal Anas crecca Field observation NB (ws) Egyptian Goose Alopochen aegyptiacus Field observation NB (ws) Wood Duck Aix sponsa Field observation NB (ws)? Lesser Scaup Aythya affinis Field observation NB (ws) Tufted Duck Aythya fuligula Field observation NB (ws) Pochard Aythya ferina Field observation NB (ws) Ferruginous Pochard Aythia nyroca Field obsrvation NB (ws) Bean Goose Anser fabalis Field observation NB (ws) Red-breasted Goose Branta ruficollis Field observation NB (ws) Black Brant Branta bernicla nigricans Field observation NB (ws) Goldeneye Bucephala clangula Field observation NB (ws) Bufflehead Bucephala albeola Field observation NB (ws) White-winged Duck Cairina scutulata Field observation NB (ws) Long-tailed Duck Clangula hyemalis Field obsrvation NB (ws) Mute Swan Cygnus olor Field observation NB (ws) Whooper Swan Cygnus cygnus Field observation NB (ws) White-faced Whistling Duck Dendrocygna viduata Field observation NB (ws) Harlequin Duck Histrionicus histrionicus Video recording NB (ws) Smew Mergus Albellus Video recording NB (ws) Goosander Mergus merganser Field observation NB (ws) African Pygmy-Goose Nettapus auritus Field observation NB (ws) Marbled Teal Marmaronetta angustirostris Field observation NB (ws) Ruddy Duck Oxyura jamaicensis Field observation NB (ws) White-Headed Duck Oxiura leucocephala Field observation NB (ws) Red-crested Pochard Netta rufina Field observation NB (ws) King Eider Somateria spectabilis Field observation NB (ws) Spectacled Eider Somateria fischeri Field observation NB (ws) Common Eider Somateira m. mollisima Field observation NB (ws) Common Shelduck Tadorna tadorna Field observation NB (ws) Paradise Shelduck Tadorna variegata Field observation NB (ws) Falconiformes Cathartidae New World Vultures Black Vulture Coragyps atratus Field observation NB Andean Condor Vultur gryphus Field observation NB Appendix 137

Order Family Family members Common name Scientific name Source Bobbing Pandionidae Osprey Osprey Pandion haliaetus (IBC,2005) NB Accipitridae Hawks, Eagles and Kites Eagles Aquila sp. Field observation NB Buzzards Buteo sp. Field observation NB Harriers Circus sp. Field observation NB Kites Milvus sp. Field observation NB Eurasian Griffon Gyps fulvus Field observation NB White-backed Vulture Gyps africanus Video recording NB Ruppell´s griffon vulture Gyps rueppelli Video recording NB Black Vulture Aegypius monachus Field observation NB lammergeier Gypaetus barbatus Video recording NB Steller's Sea-eagle Haliaeetus pelagicus Video recording NB Bateleur Terathopius ecaudatus Video recording NB Lappet-faced Vulture Torgos tragheliotus Video recording NB Hooded Vulture Necrosyrtes monachus Video recording NB White-Headed Vulture Trigonoceps occipitalis Video recording NB Sagittariidae Secretary-bird Secretary bird Sagittarius serpentarius Video recording HB Falconidae Falcons and Caracaras Eurasian Krestel Falco tinnunculus Video recording NB Southern Caracara Caracara plancus Video recording NB Galliformes Megapodiidae Megapodes Australian Brush-turkey Alectura lathami (IBC,2005) HB Guans, Chachalacas and Cracidae Allies Razor-billed Curassow Mitu tuberosa Video recording HB Nocturnal Curassow Nothocrax urumutum Field observation HB Black Curassow Crax alector Field observation HB Horned Guan Oreophasis derbianus Video recording HB Blue-throated Piping-guan Pipile cujubi Video recording HB Meleagridae Turkeys Wild Turkey Meleagris gallopavo Video recording HB Ocellated Turkey Meleagris ocellata Video recording HB Tetraonidae Grouses Hazel Grouse Bonasa bonasia Video recording HB Eurasian Capercaillie Tetrao urogallus Video recording HB Black Grouse Tetrao tetrix Video recording HB Odontophoridae New World Quail California Quail Callipepla californica Video recording HB Phasianidae Pheasants and Partridges Partridges Perdix s.p. (Dagg, 1977) HB Golden Pheasant Chrysolophus pictus Video recording HB Domestic Chicken Gallus domesticus (Pratt, 1982) HB Common Quail Coturnix coturnix Field observation HB Blue-breasted Quail Coturnix chinensis Video recording HB Appendix 138

Order Family Family members Common name Scientific name Source Bobbing Brown Quail Coturnix ypsilophora (Dagg, 1977) HB Vulturine Guineafowl Acryllium vulturinum Video recording HB Indian Peafowl Pavo cristatus Video recording HB Green Peafowl Pavo muticus Video recording HB Gray Brested Patridge Arborophila orientalis Video recording HB Crested Fireback Lophura ignita Video recording HB Satyr-tragopan Tragopan satyra Video recording HB Malayan Peacock Pheasant Polyplectron malacense Video recording HB Crested Francolin Francolinus sephaena Video recording HB Yellow Necked Spurfowl Francolinus leucoscepus Video recording HB Red Necked Spurfowl Francolinus afer Video recording HB Numididae Guineafowl Helmeted Guineafowl Numida meleagris Field observation HB Opisthocomiformes Opisthocomidae Hoatzin Gruiformes Mesitornithidae Mesites Turnicidae Button Quails Barred Buttonquail Turnix suscitator Field observation HB Gruidae Cranes Demoiselle Crane Anthropoides Virgo Video recording HB Black crowned-Crane Balearica pavonina Video recording HB Red-crowned Crane Grus japonensis Video recording HB Hooded Crane Grus monacha Video recording HB Common Crane Grus grus Video recording HB Aramidae Limpkin Psophiidae Trumpeters Gray-winged Trumpeter Psophia crepitans Video recording HB Rallidae Rails, Gallinules and Coots Common Moorhen Gallinula chloropus Video recording HB* (b) Eurasian Coot Fulica atra Video recording HB* (b) Giant Wood-rail Aramides ypecaha Video recording HB Black crake Amaurornis flavirostris Video recording HB Corncrake Crex crex Video recording HB Heliornithidae Sungrebes and Finfoots Rhynochetidae Kagu Kagu Rhynochetos jubatus Field observation NB Eurypygidae Sunbittern Sunbittern Eurypyga helias Video recording HB Cariamidae Seriemas Red-legged Seriema Cariama cristata Video recording HB* Otididae Bustards Kori Bustard Ardeotis kori Video recording HB* Black Bustard Eupodotis afra Video recording HB* Charadriiformes Jacanidae Jacanas Wattled Jacana Jacana jacana Video recording HB African Jacana Actophilornis africanus Video recording HB Appendix 139

Order Family Family members Common name Scientific name Source Bobbing Rostratulidae Painted-Snipe Dromadidae Crab Plover (Whiteside, Haematopodidae Oystercatchers Oystercatcher Haematopus ostralegus 1967) HB* Blackish Oystercatcher Haematopus ater IBC HB* Magellanic Oystercatcher Haematopus leucopodus IBC HB* Ibidorhynchidae Ibisbill Ibisbill Ibidorhyncha struthersii IBC HB Recurvirostridae Avocets and Stilts Avocet Recurvirostra avosetta Video recording HB Black-winged stilt Himantopus himantopus Video recording HB Burhinidae Thick-knees Stone-curlew Burhinus oedicnemus Video recording NB Glareolidae Pratincoles and Coursers Charadriidae Plovers and Lapwings Little Ringed Plover Charadrius dubius Field observation HB*& OHM Grey Plover Pluvialis squatarola Video recording HB Field Egyptian Plover Pluvianus aegyptius Observation HB*& OHM Northen Lapwing Vanellus vanellus Video recording HB* Southern Lapwing Vanellus chilensis Video recording NB, OHM Crowned Lapwing Vanellus coronatus Video recording HB* Blacksmith Lapwing Vanellus armatus Video recording HB* Pluvianellidae Magellanic Plover (Whiteside, Scolopacidae Sandpipers and Allies Turnstones Arenaria interpres 1967) HB Sharp-tailed Sandpiper Calidris acuminata (Dagg, 1977) HB Red-necked Stint Calidris ruficollis (Dagg, 1977) HB Common Snipe Gallinago gallinago (IBC,2005) HB Wilson's Phalarope steganopus tricolor (IBC,2005) HB Common Redshank Tringa totanus Video recording HB Common Sandpiper Actitis hypoleucos Video recording HB Ruff Philomachus pugnax Video recording HB Pedionomidae Plains-wanderer Thinocoridae Seedsnipes Rufous-bellied Seedsnipe Attagis gayi (IBC,2005) HB Chionididae Sheathbills Pale-faced Sheathbill Chionis alba (IBC,2005) HB Stercorariidae Skuas and Jaegers Great Skua Stercorarius skua (IBC,2005) NB Laridae Gulls Gulls Larus sp. (Dagg, 1977) NB Black Tailed Gull Larus crassirostris Video recording NB Sternidae Terns Inca Tern Larosterna inca Video recording NB Rynchopidae Skimmers Appendix 140

Order Family Family members Common name Scientific name Source Bobbing Alcidae Auks, Murres and Puffins Atlantic Puffin Fratercula arctica (IBC,2005) NB Pterocliformes Pteroclididae Sandgrouse Namaqua Sandgrouse Pterocles namaqua (IBC,2005) HB (Dunlap & Columbiformes Columbidae Pigeons and Doves Rock Dove Columba livia Mowrer,1930) HB (wl) Speckled Pigeon Columba guinea Video recording HB African Olive Pigeon Columba arquatrix Video recording HB (Dunlap & Doves Streptopelia sp. Mowrer,1930) HB Zoe Imperial Pigeon Ducula zoeae Field observation HB Squatter Pigeon Geophaps scripta Field observation HB Victoria Crowned-Pigeon Goura victoriae (Dagg, 1977) HB Souther Crowned-pigeon Goura scheepmakeri Field observation HB Squatter Pigeon Geophaps scripta Video recording HB Laughing Dove Streptopelia senegalensis Video recording HB African Collared-Dove Streptopelia roseogrisea Video recording HB Bronze-naped Pheasant Pigeon Otidiphaps nobilis Video recording HB Wompoo Fruit-Dove Ptilinopus magnificus Video recording HB Emerald Dove Chalcophaps indica Video recording HB Cinnamon Ground Dove Gallicolumba rufigula Video recording HB Bar-shouldered Dove Geopelia humeralis Video recording HB Socorro Dove Zenaida graysoni Video recording HB Crested Pigeon Geophaps Iophotes Video recording HB Diamond Dove Geopelia cuneata Video recording HB Mindanao Bleeding-heart Gallicolumba criniger Video recording HB Ruddy Quail-Dove Geotrygon montana Video recording HB Crested Quail-Dove Geotrygon versicolor Video recording HB Orange-bellied Fruit-Dove Ptilinopus iozonus Video recording HB Namaqua Dove Oena capensis Video recording HB Psittaciformes Cacatuidae Cockatoos and Allies Psittacidae Parrots, Macaws and Allies Budgerigar Melopsittacus undulatus Field observation OHM Musophagiformes Musophagidae Turacos Violet Turaco Musophaga violacea Video recording HB& BB Purple-crested Turaco Tauraco porphyreolophus (IBC,2005) HB& BB Cuculiformes Cuculidae Cuckoos Greater Roadrunner Geococcyx californianus NB Coral-billed Ground- cuckoo Carpococcyx renauldi Video recording HB Crested Coua Coua cristata Video recording BB Appendix 141

Order Family Family members Common name Scientific name Source Bobbing Guira Cuckoo Guira guira Video recording HB& BB Coppery-tailed Coucal Centropus cupreicaudus (IBC,2005) HB White-browed Coucal Centropus superciliosus (IBC,2005) HB Strigiformes Tytonidae Barn-Owls Barn Owl Tyto alba (IBC,2005) OHM Strigidae Typical Owls Tawny Owl Strix aluco Field observation OHM Caprimulgiformes Steatornithidae Oilbird Aegothelidae Owlet-Nightjars Podargidae Frogmouths Nyctibiidae Potoos Caprimulgidae Nightjars and Allies Apodiformes Apodidae Swifts Hemiprocnidae Treeswifts Trochiliformes Trochilidae Hummingbirds Coliiformes Coliidae Mousebirds Blue-naped Mousebird Urocolius macrourus Video recording BB Trogoniformes Trogonidae Trogons Coraciiformes Alcedinidae Kingfishers Common Kingfisher Alcedo atthis Field observation OHM Todidae Todies Momotidae Motmots Meropidae Bee-eaters Coraciidae Typical Rollers Brachypteraciidae Ground-Rollers Pitta-like Ground-roller Atelornis pittoides Video recording NB Long-tailed Ground-roller Uratelornis chimaera Video recording NB Leptosomidae Cuckoo-Roller Upupidae Hoopoes Hoopoe Upupa epops Video recording HB Phoeniculidae Woodhoopoes Abyssinian Ground- Bucerotidae Hornbills hornbill Bucorvus abyssinicus Video recording NB Southern Ground-hornbill Bucorvus leadbeateri Video recording NB Piciformes Galbulidae Jacamars Bucconidae Puffbirds Capitonidae Barbets Red-and-yellow Barbet Trachyphonus erythrocephalus Video recording BB D'arnaud's Barbet Trachyphonus darnaudii Video recording BB Ramphastidae Toucans Keel-billed toucan Ramphastos sulfuratus Video recording BB Indicatoridae Honeyguides Picidae Woodpeckers and Allies Green Woodpecker Picus viridis Field observation BB Northern Flicker Colaptes auratus Video recording BB Appendix 142

Order Family Family members Common name Scientific name Source Bobbing Golden-fronted Woodpecker Melanerpes aurifrons Video recording BB Passeriformes Eurylaimidae Broadbills Green Broadbill Calyptomena Viridis Video recording BB Philepittidae Asities Furnariidae Ovenbirds Pale-legged Hornero Furnarius leucopus Video recording HB Rufous-banded Miner Geositta rufipennis Video recording HB Cordilleran Canastero Asthenes modesta (IBC,2005) BB Ochre-cheeked Spinetail Poecilurus scutatus (IBC,2005) BB

Narrow-billed (IBC,2005) Dendrocolaptidae Woodcreepers Woodcreeper Lepidocolaptes angustirostris BB Thamnophilidae Typical Antbirds Formicariidae Antthrushes and Antpittas Conopophagidae Gnateaters Rhinocryptidae Tapaculos Magellanic Tapaculo Scytalopus magellanicus (IBC,2005) BB Phytotomidae Plantcutters Cotingidae Cotingas Long-wattled Umbrellabird Cephalopterus penduliger Video recording Pipridae Manakins Tyrannidae Tyrant Flycatchers Oxyruncidae Sharpbill Pittidae Pittas Pittas Pitta sp. (Dagg, 1977) BB Blue-tailed Pitta Pitta arcuata Video recording BB Black Crouned pitta Pitta venusta Video recording BB Banded pitta Pitta guajana Vidoe recordings BB Atrichornithidae Scrub-birds Menuridae Lyrebirds Lyre Bird Menura superba (Dagg, 1977) HB Acanthisittidae New Zealand Wrens Alaudidae Larks Hirundinidae Swallows Lesser Striped Swallow Hirundo abyssinica Field observation BB Motacillidae Wagtails and Pipits White Wagtail Motacilla alba Field observation HB African Pied Waigtail Motacilla Aguimp Field observation HB Campephagidae Cuckoo-shrikes

Pycnonotidae Bulbuls Redwhiskered Bulbul Pycnonotus jocosus Video recording HB Common Bulbul Pycnonotus barbatus Video recording HB

Regulidae Kinglets Appendix 143

Order Family Family members Common name Scientific name Source Bobbing

Chloropseidae Leafbirds Aegithinidae Ioras Long Tailed Tit Aeghithalos caudatos Field observation BB Ptilogonatidae Silky-flycatchers Bombycillidae Waxwings Hypocoliidae Hypocolius Dulidae Palm Chat Cinclidae Dippers Troglodytidae Wrens Mimidae Mockingbirds and Thrashers Prunellidae Accentors Turdidae Thrushes and Allies Song thrush Turdus philomenos Field observation BB Fieldfare Turdus philaris (Jonsson, 1993) BB Robin Erithacus rubecula (Jonsson, 1993) BB Mistle thrust turdus viscivorus Field observation BB Redstarts Phoenicurus sp. (Jonsson, 1993) BB (Whiteside, Blackbirds Turdus meruls 1967) HB& BB Cisticolidae Cisticolas and Allies Sylviidae Old World Warblers Cetti’s Warblers Cettia cetti Field observation BB Polioptilidae Gnatcatchers Muscicapidae Old World Flycatchers Chats Cossypha sp. (Jonsson, 1993) BB Cape robin-chat Cossypha caffra Video recording BB Red-capped Robin-Chat Cossypha natalensis Video recording BB White-browed Scrub-robin Cercotrichas leucophrys Video recording BB Spotted flycatcher muscicapa striata Field observation BB Capped wheatear Oenanthe pileata Video recording BB Platysteiridae Wattle-eyes Rhipiduridae Fantails Monarchidae Monarch Flycatchers Petroicidae Australasian Robins

Pachycephalidae Whistlers and Allies Picathartidae Rockfowl White Creasted Laughing- Timaliidae Babblers thrush Garrulax leucolophus Video recording BB Appendix 144

Order Family Family members Common name Scientific name Source Bobbing Yellow-throated Laughing- thrush Garrulax galbanus Video recording BB Pomatostomidae Pseudo-babblers Paradoxornithidae Parrotbills Bearded Reedling Panurus biarmicus Orthonychidae Logrunner and Chowchilla Whipbirds and Quail- Cinclosomatidae thrushes Aegithalidae Long-tailed Tits Maluridae Fairywrens Acanthizidae Thornbills and Allies Epthianuridae Australian Chats Neosittidae Sitellas Climacteridae Australasian Treecreepers Paridae Chickadees and Tits Tits Parus sp. Field observation BB Sittidae Nuthatches Tichodromidae Wallcreeper Certhiidae Creepers Rhabdornithidae Philippine Creepers Remizidae Penduline Tits Nectariniidae Sunbirds and Spiderhunters Melanocharitidae Berrypeckers and Longbills Tit and Crested Paramythiidae Berrypeckers Dicaeidae Flowerpeckers Pardalotidae Pardalotes Zosteropidae White-eyes Promeropidae Sugarbirds Meliphagidae Honeyeaters Oriolidae Old World Orioles Irenidae Fairy-bluebirds Laniidae Shrikes Magpie Shrike Corvinella melanoleuca Field observation BB Malaconotidae Bushshrikes and Allies Common Gonolek laniarius barbarus Field observation BB Prionopidae Helmetshrikes Vangidae Vangas Dicruridae Drongos Callaeidae Wattlebirds Appendix 145

Order Family Family members Common name Scientific name Source Bobbing Grallinidae Mudnest-builders White-winged Corcoracidae Chough&Apostlebird Artamidae Woodswallows Pityriaseidae Bristlehead Cracticidae Bellmagpies and Allies Paradisaeidae Birds of Paradise Ptilonorhynchidae Bowerbirds Corvidae Crows, Jays and Magpies White-winged Magpie Platysmurus leucopterus (Frost, 1978) HB& BB Magpie Gymnorhina tibicen (Dagg, 1977) HB& BB Rook Corvus frugilegus (Frost, 1978) HB& BB Carrion Crow Corvus corone (Dagg, 1977) HB& BB Jay Garrulus glandarius (Dagg, 1977) BB Eurasian Magpie Pica pica Video recording HB& BB Red-billed Chough Pyrrhocorax pyrrhocorax Field observation HB& BB Alpine Chough Pyrrochoras gragulus Video recording HB& BB Azure-winged magpie Cyanopica cyana Video recording HB& BB Euroasian Jackdaw Corvus monedula Video recording HB& BB Cape Rock Corvus Capensis Field observation HB& BB Sturnidae Starlings Superb starling lamprotornis superbus Video recording HB*& BB (Whiteside, Starlings Sturnus sp. 1967) HB*& BB Mynah Bird Acrhidotheres tristis (Dagg, 1977) HB*& BB European Starling Sturnus vulgaris HB*& BB Ashy Starling Cosmopsarus unicolor HB*& BB Red-billed Oxpecker Buphagus erythrorhynchus BB White-headed Buffalo- Ploceidae Weavers and Allies Weaver Dinemellia dinemelli Video recording HB& BB Village Weaver Ploceus cucullatus Video recording HB& BB Long-tailed Widowbird Euplectes progne Video recording HB& BB Spleckle-fronted Weaver Sporopipes frontalis Video recording HB& BB Red billed Buffalo-weaver Bubalornis niger Video recording HB& BB

Baglafecht weaver Ploceus baglafecht Field observation BB Estrildidae Waxbills and Allies Zebra Finch Taeniopygia guttata Field observation BB Diamond Firetail Stagonopleura guttata BB

Pink-throated Twinspot Hypargos margaritatus BB Appendix 146

Order Family Family members Common name Scientific name Source Bobbing Black-crowned Waxbill Estrilda nonnula BB Blue-capped Cordon-bleu Uraeginthus cyanocephalus BB Viduidae Indigobirds Vireonidae Vireos and Allies Drepanididae Hawaiian Honeycreepers Peucedramidae Olive Warbler Parulidae Wood Warblers Coerebidae Banaquit Thraupidae Tanager and Allies Blue-necked Tanager Tangara cyanicollis BB Golden Tanager Tangara arthus BB

Paradise Tanager Tangara chilensis BB Buntings, Sparrows, Emberizidae Seedeater and Allies Yellowhammer Emberiza citrinella Video recording BB Saffron Finch Sicalis flaveola Video recording BB Black-hooded Sierra- Finch Phrygilus atriceps Video recording BB Grey-hooded Sierra-Finch Phrygilus gayi Video recording BB Patagonian Sierra-Finch Phrygilus patagonicus Video recording BB Saltators, Cardinals and Cardinalidae Allies Icteridae Blackbirds, orioles, crackles Scarlet Headed Blackbird Amblyramphus holosericeus BB Siskins, Crossbills and (Whiteside, Fringillidae Finches Chaffinch Fringilla coelebs 1967) HB Serin Serinus serinus Field observation BB Crossbill Loxia leucoptera Field observation BB Common Crossbill Loxia curvirostra Video recording BB Eurasian Bullfinch pyrrhula pyrrhula Video recording BB Coccothraustes Hawfinch coccothraustes BB Oriole Finch Linurgus olivaceus Video recording BB Passeridae Old World Sparrows House Sparrow Passer domesticus Video recording BB Yellow-spotted Petronia Petronya pyrgita Video recording BB Grey-headed Sparrow Passer griseus Video recording BB

Table C-1: Behavioural observations following Clements’ 2000 checklist (2004 update)

Appendix 147

Abbreviations:

HB: head-bobbing while walking HB(s): head-bobbing while swimming. HB(ws): head-bobbing while walking & swimming HB(wl): head-bobbing while walking & landing HB*: rare or occasional head-bobbing birds

NB: neither head-bobbing nor hoping birds NB(ws): Non-bobbing during walking and swimming BB: hoping birds OHM: other head movement ? Further observations needed Appendix 148

Behavioural observations following Sibley`s phylogenetic tree (1990)

Class Aves Subclass Neornithes Infraclass Eoaves Parvclass Ratitae Order Struthioniformes Suborder Struthioni Infraorder Struthionides Family Struthionidae: Ostrich HB* Infraorder Rheides Family Rheidae: Rheas NB? Suborder Casuarii Family Casuariidae Tribe Casuariini: Cassowaries HB Tribe Dromaiini: Emus HB* Family Apterygidae: Kiwis Order Tinamiformes Family Tinamidae: Tinamous HB Infraclass Neoaves Parvclass Galloanserae Superorder Gallomorphae Order Craciformes Suborder Craci Family Cracidae: Guans, Chachalacas, etc. HB Suborder Megapodii Family Megapodiidae: Megapodes HB Order Galliformes Parvorder Phasianida Superfamily Phasianoidea Family Phasianidae: Grouse, Pheasants, etc. HB Superfamily Numidoidea Family Numididae: Guineafowls HB Parvorder Odontophorida Family Odontophoridae: New World Quails HB Superorder Anserimorphae Order Anseriformes Infraorder Anhimides Superfamily Anhimoidea Family Anhimidae: Screamers NB Superfamily Anseranatoidea Family Anseranatidae: Magpie Goose NB Infraorder Anserides Family Dendrocygnidae: Whistling-Ducks NB Family Anatidae Subfamily Oxyurinae: Stiff-tailed Ducks NB Subfamily Cygninae: Swans NB Subfamily Anatinae Tribe Anserini: Geese NB Tribe Anatini: Typical Ducks NB Parvclass Turnicae Order Turniciformes Family Turnicidae: Buttonquails(Turnix, Ortyxelos) HB Parvclass Picae Order Piciformes Infraorder Picides Family Indicatoridae: Honeyguides Family Picidae: Woodpeckers, Wrynecks BB

Infraorder Ramphastides Superfamily Megalaimoidea Family Megalaimidae: Asian Barbets Superfamily Lybioidea Family Lybiidae: African Barbets BB Superfamily Ramphastoidea Family Ramphastidae Subfamily Capitoninae: New World Barbets BB Subfamily Ramphastinae: Toucans BB Parvclass Coraciae Superorder Galbulimorphae Order Galbuliformes Infraorder Galbulides Family Galbulidae: Jacamars Infraorder Bucconides Appendix 149

Family Bucconidae: Puffbirds Superorder Bucerotimorphae Order Bucerotiformes Family Bucerotidae: Typical Hornbills BB Family Bucorvidae: Ground-Hornbills NB Order Upupiformes Infraorder Upupides Family Upupidae: Hoopoes HB Infraorder Phoeniculides Family Phoeniculidae: Wood-Hoopoes Family Rhinopomastidae: Scimitarbills Superorder Coraciimorphae Order Trogoniformes Family Trogonidae Subfamily Apaloderminae: African Trogons Subfamily Trogoninae Tribe Trogonini: New World Trogons Tribe Harpactini: Asian Trogons Order Coraciiformes Suborder Coracii Superfamily Coracioidea Family Coraciidae: Typical Rollers Family Brachypteraciidae: Ground-Rollers NB Superfamily Leptosomoidea Family Leptosomidae: Cuckoo-Rollers Suborder Alcedini Infraorder Alcedinides Parvorder Momotida Family Momotidae: Motmots Parvorder Todida Family Todidae: Todies Parrorder Alcedinida Family Alcedinidae: Alcedinid Kingfishers Parrorder Cerylida Superfamily Dacelonoidea Family Dacelonidae: Dacelonid Kingfishers Superfamily Ceryloidea Family Cerylidae: Cerylid Kingfishers Infraorder Meropides Family Meropidae: Bee-eaters Parrclass Coliae Order Coliiformes Family Coliidae Subfamily Coliinae: Typical Mousebirds BB Subfamily Urocoliinae: Long-tailed Mousebirds Parvclass Passerae Superorder Cuculimorphae Order Cuculiformes Infraorder Cuculides Parvorder Cuculida Superfamily Cuculoidea Family Cuculidae: Old World Cuckoos HB, BB Superfamily Centropodoidea Family Centropodidae: Coucals HB Parvorder Coccyzida Family Coccyzidae: American Cuckoos Infraorder Crotophagides Parvorder Opisthocomida Family Opisthocomidae: Hoatzin Parvorder Crotophagida Family Crotophagidae Tribe Crotophagini: Anis Tribe Guirini: Guira Cuckoo HB&BB Parvorder Neomorphida Family Neomorphidae: Roadrunners, etc. NB Superorder Psittacimorphae Order Psittaciformes Family Psittacidae: Parrots, Macaws, etc. OHM Superorder Apodimorphae Order Apodiformes Family Apodidae: Typical Swifts Family Hemiprocnidae: Crested Swifts Order Trochiliformes Family Trochilidae Subfamily Phaethornithinae: Hermits Subfamily Trochilinae: Typical Hummingbirds Appendix 150

Superorder Strigimorphae Order Musophagiformes Family Musophagidae Subfamily Musophaginae: Turacos HB&BB Subfamily Criniferinae: Plaintain-eaters Order Strigiformes Suborder Strigi Parvorder Tytonida Family Tytonidae: Barn and Grass owls OHM Parvorder Strigida Family Strigidae: Typical Owls OHM Suborder Aegotheli Family Aegothelidae: Owlet-nightjars Suborder Caprimulgi Infraorder Podargides Family Podargidae: Australian Frogmouths Family Batrachostomidae: Asian Frogmouths Infraorder Caprimulgides Parvorder Steatornithida Superfamily Steatornithoidea Family Steatornithidae: Oilbird Superfamily Nyctibioidea Family Nyctibiidae: Potoos Parvorder Caprimulgida Superfamily Eurostopodoidea Family Eurostopodidae: Eared Nightjars Superfamily Caprimulgoidea Family Caprimulgidae Subfamily Chordeilinae: Nighthawks Subfamily Caprimulginae: Nightjars Superorder Passerimorphae Order Columbiformes Family Columbidae: Pigeons, Doves HB Order Gruiformes Suborder Grui Infraorder Eurypygides Family Eurypygidae: Sunbittern HB Infraorder Otidides Family Otididae: Bustards HB* Infraorder Gruides Parvorder Gruida Superfamily Gruoidea Family Gruidae: Cranes HB Family Heliornithidae Tribe Aramini: Limpkin Tribe Heliornithini: New World Sungrebe (Podica& Heliopais ) Superfamily Psophioidea Family Psophiidae: Trumpeters HB Parvorder Cariamida Family Cariamidae: Seriemas HB* Family Rhynochetidae: Kagu NB Suborder Ralli Family Rallidae: Rails, Gallinules, Coots HB Suborder Mesitornithi Family Mesitornithidae: Mesites Order Ciconiiformes Suborder Charadrii Infraorder Pteroclides Family Pteroclidae: Sandgrouse HB Infraorder Charadriides Parvorder Scolopacida Superfamily Scolopacoidea Family Thinocoridae: Seedsnipe HB Family Pedionomidae: Plains-wanderer Family Scolopacidae Subfamily Scolopacinae: Woodcock, Snipe HB Subfamily Tringinae: Sandpipers, Curlews, Phalaropes HB Superfamily Jacanoidea Family Rostratulidae: Paintedsnipe Family Jacanidae: Lily-trotters, Jacanas HB Parvorder Charadriida Superfamily Chionidoidea Family Chionididae: Sheathbills HB Superfamily Charadrioidea Family Burhinidae: Thick-knees NB Family Charadriidae Appendix 151

Subfamily Recurvirostrinae Tribe Haematopodini: Oystercatchers HB* Tribe Recurvirostrini: Avocets, Stilts HB Subfamily Charadriinae: Plovers, Lapwings HB* Superfamily Laroidea Family Glareolidae Subfamily Dromadinae: Crab-plover NB Subfamily Glareolinae: Pratincoles NB Family Laridae Subfamily Larinae Tribe Stercorariini: Jaegers, Skuas NB Tribe Rynchopini: Skimmers Tribe Larini: Gulls NB Tribe Sternini: Terns NB Subfamily Alcinae: Auks, Murres, Puffins NB Suborder Ciconii Infraorder Falconides Parvorder Accipitrida Family Accipitridae Subfamily Pandioninae: Osprey NB Subfamily Accipitrinae: Hawks, Eagles, O.W. vultures, kites NB Family Sagittariidae: Secretarybird HB Parvorder Falconida Family Falconidae: Falcons, Caracaras NB Infraorder Ciconiides Parvorder Podicipedida Family Podicipedidae: Grebes HB Parvorder Phaethontida Family Phaethontidae: Tropicbirds BB? Parvorder Sulida Superfamily Suloidea Family Sulidae: Boobies, Gannets NB Family Anhingidae: Anhingas, Darters HB? Superfamily Phalacrocoracoidea Family Phalacrocoracidae: Cormorants NB Parvorder Ciconiida Superfamily Ardeoidea Family Ardeidae: Herons, Bitterns, Egrets HB Superfamily Scopoidea Family Scopidae: Hammerhead HB Superfamily Phoenicopteroidea Family Phoenicopteridae. Flamingos NB Superfamily Threskiornithoidea Family Threskiornithidae: Ibises, Spoonbills HB Superfamily Pelecanoidea Family Pelecanidae Subfamily Balaenicipitinae: Shoebill NB Subfamily Pelecaninae: Pelicans NB Superfamily Ciconioidea Family Ciconiidae Subfamily Cathartinae: New World vulture NB Subfamily Ciconiinae: Storks HB (unless Marabou) Superfamily Procellarioidea Family Fregatidae: Frigatebirds Family Spheniscidae: Penguins NB Family Gaviidae: Loons NB Family Procellariidae Subfamily Hydrobatinae: Storm-Petrels Subfamily Procellariinae: Shearwaters, Petrels, Diving-Petrels Subfamily Diomedeinae: Albatrosses NB (OHM) Order Passeriformes HB, BB, HB&BB Suborder Tyranni (Suboscines) Infraorder Acanthisittides Family Acanthisittidae: New Zealand Wrens Infraorder Eurylaimides Superfamily Pittoidea Family Pittidae: Pittas BB Superfamily Eurylaimoidea Family Eurylaimidae: Broadbills Family Philepittidae: Asities Infraorder Tyrannides Parvorder Tyrannida Family Tyrannidae Subfamily Corythopinae: Corythopis, Mionectes, etc. Subfamily Tyranninae: Tyrant Flycatchers Subfamily Tityrinae Appendix 152

Tribe Schiffornithini: Schiffornis Tribe Tityrini, Tityras: Becards Subfamily Cotinginae: Cotingas, Plantcutters, Sharpbills Subfamily Piprinae: Manakins Parvorder Thamnophilida Family Thamnophilidae: Typical Antbirds Parvorder Furnariida Superfamily Furnarioidea Family Furnariidae Subfamily Furnariinae: Ovenbirds Subfamily Dendrocolaptinae: Woodcreepers BB Superfamily Formicarioidea Family Formicariidae: Ground Antbirds Family Conopophagidae: Gnateaters Family Rhinocryptidae: Tapaculos BB Suborder Passeri (Oscines) Parvorder Corvida Superfamily Menuroidea Family Climacteridae: Australo-Papuan Treecreepers Family Menuridae Subfamily Menurinae: Lyrebirds HB Subfamily Atrichornithinae: Scrubbirds Family Ptilonorhynchidae: Bowerbirds Superfamily Meliphagoidea Family Maluridae Subfamily Malurinae Tribe Malurini: Fairywrens Tribe Stipiturini: Emuwrens Subfamily Amytornithinae: Grasswrens Family Meliphagidae: Honeyeaters, incl. Ephthianura, Ashbyia Family Pardalotidae Subfamily Pardalotinae: Pardalotes Subfamily Dasyornithinae: Bristlebirds Subfamily Acanthizinae Tribe Sericornithini: Scrubwrens Tribe Acanthizini: Thornbills, Whitefaces, etc. Superfamily Corvoidea Family Eopsaltriidae: Australo-Papuan robins, Drymodes Family Irenidae: Fairy-bluebirds, Leafbirds Family Orthonychidae: Log-runners or Chowchillas Family Pomatostomidae: Australo-Papuan babblers Family Laniidae: True Shrikes (Lanius, Corvinella, Eurocephalus) BB Family Vireonidae: Vireos, Greenlets, Peppershrikes Family Corvidae Subfamily Cinclosomatinae: Quail-thrushes, Whipbirds Subfamily Corcoracinae: Australian Chough, Apostlebird Subfamily Pachycephalinae Tribe Neosittini: Sittellas Tribe Mohouini: New Zealand Mohoua, Finschia Tribe Falcunculini: Shrike-tits, Oreoica, Rhagologus Tribe Pachycephalini: Whistlers, Shrike-thrushes Subfamily Corvinae Tribe Corvini: Crows, Magpies, Jays, Nutcrackers HB&BB Tribe Paradisaeini: Birds of Paradise, Melampitta Tribe Artamini: Currawongs, Woodswallows, Peltops, Pityriasis Tribe Oriolini, Orioles: Cuckooshrikes Subfamily Dicrurinae Tribe Rhipidurini: Fantails Tribe Dicrurini: Drongos Tribe Monarchini: Monarchs, Magpie-larks Subfamily Aegithininae: Ioras BB Subfamily Malaconotinae Tribe Malaconotini: Bush-shrikes BB Tribe Prionopini: Helmet-shrikes, Batis, Platysteira, Vangas Family Callaeatidae: New Zealand wattlebirds Parvorder Passerida Superfamily Muscicapoidea Family Bombycillidae Tribe Dulini: Palmchat Tribe Ptilogonatini: Silky-flycatchers Tribe Bombycillini: Waxwings Family Cinclidae: Dippers Family Muscicapidae Subfamily Turdinae: True thrushes, incl. Chlamydochaera, BB, HB&BB Brachypteryx, Alethe Subfamily Muscicapinae Appendix 153

Tribe Muscicapini: Old World Flycatchers Tribe Saxicolini. Chats, Erithacus, etc. Family Sturnidae Tribe Sturnini: Starlings, Mynas HB&BB Tribe Mimini: Mockingbirds, Thrashers, American Catbirds Superfamily Sylvioidea Family Sittidae Subfamily Sittinae: Nuthatches Subfamily Tichodromadinae: Wallcreeper Family Certhiidae Subfamily Certhiinae Tribe Certhiini: Northern Creepers Tribe Salpornithini: Afro-Asian Creeper Subfamily Troglodytinae: Wrens Subfamily Polioptilinae: Gnatcatchers, Verdin, Gnatwrens Family Paridae Subfamily Remizinae: Penduline-Tits Subfamily Parinae: Titmice, Chickadees BB Family Aegithalidae: Long-tailed Tits, Bushtits Family Hirundinidae Subfamily Pseudochelidoninae: River-Martins Subfamily Hirundininae: Swallows, Martins BB Family Regulidae: Kinglets Family Pycnonotidae: Bulbuls HB Family Hypocoliidae: Gray Hypocolius Family Cisticolidae: African Warblers Family Zosteropidae: White-eyes Family Sylviidae Subfamily Acrocephalinae: Leaf Warblers, Reed Wrablers, etc Subfamily Megalurinae: Grassbirds, Songlarks, Fernbird Subfamily Garrulacinae: Laughingthrushes, Liocichlas BB Subfamily Sylviinae Tribe Timaliini: Babblers, Minlas, Fulvettas, Yuhinas, Parrotbills Tribe Chamaeini: Wrentit Tribe Sylviini: Sylviine Warblers (Sylvia, Parisoma) BB Superfamily Passeroidea Family Alaudidae: Larks Family Nectariniidae Subfamily Promeropinae: African Sugarbirds Subfamily Nectariniinae Tribe Dicaeini: Flowerpeckers Tribe Nectariniini: Sunbirds, Spiderhunters Family Melanocharitidae Tribe Melanocharitini: Melanocharis, Rhamphocharis Tribe Toxorhamphini: Toxorhamphus, Oedistoma Family Paramythiidae: Paramythia, Oreocharis Family Passeridae Subfamily Passerinae: Sparrows, Rock-Sparrows, etc. Subfamily Motacillinae: Wagtails and Pipits HB Subfamily Prunellinae: Accentors, Dunnock Subfamily Ploceinae. Weaverbirds, incl. Amblyospiza, HB&BB, BB Bubalornis, etc. Subfamily Estrildinae Tribe Estrildini: Waxbills, Estrildines BB Tribe Viduini: Indigobirds, Whydahs Family Fringillidae Subfamily Peucedraminae: Olive Warbler Subfamily Fringillinae Abrebiations: Tribe Fringillini: Chaffinches, Brambling HB, BB? Tribe Carduelini: Goldfinches, Crossbills, etc. BB HB: head-bobbing birds Tribe Drepanidini: Hawaiian Honeycreepers Subfamily Emberizinae HB*: rare or occasional head-bobbing birds Tribe Emberizini: Buntings, Longspurs, Towhees NB: neither head-bobbing nor hoping birds Tribe Parulini: Wood Warblers, incl. Zeledonia BB: hoping birds Tribe Thraupini: Tanagers, Swallow-tanager, Neotropical OHM: other head movement Honeycreepers, Plushcap,Tanager-finches ?: Further observations needed Tribe Cardinalini: Cardinals Tribe Icterini: Troupials, Meadowlarksß American Blackbirdsß Oropendolas

Table C-2: Behavioural observations according to Sibley`s phylogenetic tree (1990). Appendix 154

Analysis of the ecological and behavioural characteristics of HB and NB birds Feeding Feeding Group Flying Predator Order Family Common name Scientific name Bobbing Activity Habitat source method size abilities pressure Casuariiformes Casuariidae Northern Cassowary Causarius unappendiculatus HB 1 1 0 3 2 0 0 Tinamiformes Tinamidae Elegant Crested Tinamou Eudromia elegans HB 1 1 1 3 1 1 0 Ciconiiformes Ardeidae Grey Blue Heron Ardea herodias HB 3 4 1 3 3 3 0 Ciconiiformes Scopidae Hamerkop Scopus umbretta HB 3 4 1 3 3 2 0 Ciconiiformes Ciconiidae White Stork Ciconia ciconia HB 3 4 0 3 3 3 0 Ciconiiformes Threskiornithidae Scarlet Ibis Eudocimus ruber HB 1 6 2 3 3 2 1 Falconiformes Sagittariidae Secretary Bird Sagittarius serpentarius HB 3 5 0 3 1 1 0 Galliformes Megapodiidae Australian Brush-turkey Alectura lathami HB 1 1 0 3 2 1 1 Galliformes Cracidae Razor-billed Curassow Mitu tuberosa HB 1 1 1 3 2 1 1 Galliformes Cracidae Horned Guan Oreophasis Derbianus HB 1 2 1 3 2 1 1 Galliformes Meleagridae Wild Turkey Meleagris gallopavo HB 1 1 2 3 1 1 2 Galliformes Tetraonidae Black Grouse Tetrao tetrix HB 1 1 0 3 1 1 1 Galliformes Odontophoridae California Quail Callipepla californica HB 1 1 2 3 2 2 1 Galliformes Phasianidae Peafowls Pavo sp. HB 1 1 2 3 2 1 2 Galliformes Numididae Helmeted Guineafowl Numida meleagris HB 1 1 2 3 1 1 2 Gruiformes Turnicidae Barred Buttonquail Turnix suscitator HB 1 1 1 3 1 2 1 Gruiformes Gruidae Demoiselle Crane Anthropoides Virgo HB 1 1 2 3 1 2 2 Gruiformes Aramidae Limpkin Aramus guarauna HB 1 1 1 3 3 1 1 Gruiformes Psophiidae Gray-winged Trumpeter Psophia crepitans HB 1 1 2 3 2 1 1 Gruiformes Eurypygidae Sunbittern Eurypyga Helias HB 2 1 0 3 2 1 1 Charadriiformes Jacanidae Wattle Jacana Jacana jacana HB 2 1 0 3 3 1 1 Charadriiformes Ibidorhynchidae Ibisbill Ibidorhyncha struthersii HB 1 1 2 3 3 2 1 Charadriiformes Recurvirostridae Avocet Recurvirostra avosetta HB 1 1 2 3 3 2 1 Charadriiformes Scolopacidae Common Redshank Tringa totanus HB 1 1 2 3 3 3 1 Charadriiformes Thinocoridae Rufous-bellied Seedsnipe Attagis gayi HB 1 1 2 3 3 3 1 Pterocliformes Pteroclididae Namaqua Sandgrouse Pterocles namaqua HB 1 1 1 3 1 1 1 Cuculiformes Cuculidae White-browed Coucal Centropus superciliosus HB 1 1 0 3 1 1 0 Coraciiformes Upupidae Hopooe Upupa epops HB 1 1 0 3 1 2 0 Appendix 155

Feeding Feeding Group Flying Predator Order Family Common name Scientific name Bobbing Activity Habitat source method size abilities pressure Podicipediformes Podicipedidae Great Crested Grebe Podiceps cristatus HB 2 3 1 3 4 1 1 Gruiformes Rallidae Common Moorhen Gallinula chloropus HB 1 1 2 3 3 1 1 Ciconiiformes Ardeidae Gray Heron Ardea cinerea HB 3 4 1 3 3 3 0 Columbiformes Columbidae Rock dove Columba livia HB 1 1 0 3 1 2 1 Struthioniformes Struthionidae Ostrich Struthio camelus HB* 1 1 0 3 1 0 0 Rheiformes Rheidae Greater Rhea Rhea americana HB* 1 1 0 3 1 0 0 Casuariiformes Dromaiidae Australian Emu Dromaius novaehollandiae HB* 1 1 0 3 1 0 1 Gruiformes Otididae Kori Bustard Ardeotis kori HB* 3 1 0 3 1 1 2 Charadriiformes Haematopodidae Oystercatcher Haematopus ostralegus HB* 2 1 2 3 3 2 1 Charadriiformes Charadriidae Northen Lapwing Vanellus vanellus HB* 1 1 1 3 3 3 1 Charadriiformes Charadriidae Little ringed Plover Charadrius dubius HB* 1 1 1 3 3 3 0 Sphenisciformes Spheniscidae Humboldt Penguin Spheniscus humboldti NB 2 3 2 3 5 0 2 Gaviiformes Gaviidae Common Loon Gavia immer NB 2 3 0 3 4 2 1 Pelecaniformes Pelecanidae Great White Pelican Pelecanus omocrotalus NB 2 6 2 3 5 2 1 Pelecaniformes Sulidae Northern Gannet Morus bassanus NB 2 3 1 3 5 2 0 Pelecaniformes Phalacrocoracidae Great Cormorant Phalacrocorax carbo NB 2 3 0 3 4 3 0 Ciconiiformes Ciconiidae Marabou Leptoptilos crumeniferus NB 5 5 2 3 1 2 0 Ciconiiformes Balaenicipitidae Shoebill balaeniceps rex NB 2 4 0 3 3 1 0 Phoenicopteriformes Phoenicopteridae Greater Flamingo Phoenicopterus roseus NB 0 7 2 3 3 3 0 Anseriformes Anhimidae Common teal Anas rubripes NB 1 7 2 3 3 3 1 Falconiformes Cathartidae Black Vulture Coragyps atratus NB 5 5 1 3 1 3 0 Falconiformes Pandionidae Osprey Pandion haliaetus NB 2 5 0 3 4 3 0 Falconiformes Accipitridae Eagles Aquila sp. NB 5 5 0 3 2 3 0 Falconiformes Accipitridae Cinereous Harrier Circus cinereus NB 3 5 0 3 1 3 0 Falconiformes Accipitridae Cinereous Vulture Aegypius monachus NB 5 5 1 3 1 3 0 Falconiformes Falconidae Eurasian Krestel Falco tinnunculus NB 3 5 0 3 1 3 0 Gruiformes Rhynochetidae Kagu Rhynochetos jubatus NB 1 1 0 3 2 0 1 Charadriiformes Burhinidae Stone-curlew Burhinus oedicnemus NB 1 1 1 0 1 1 0 Charadriiformes Stercorariidae Great Skua Stercorarius skua NB 3 1 0 3 5 3 0 Charadriiformes Laridae Gulls Larus sp. NB 2 1 2 3 5 3 0 Charadriiformes Sternidae Inca Tern Larosterna inca NB 2 3 2 3 5 3 0 Appendix 156

Feeding Feeding Group Flying Predator Order Family Common name Scientific name Bobbing Activity Habitat source method size abilities pressure Charadriiformes Alcidae Atlantic Puffin Fratercula arctica NB 2 3 0 3 5 3 1 Cuculiformes Cuculidae Greater Roadrunner Geococcyx californianus NB 3 2 0 3 2 2 0 Coraciiformes Brachypteraciidae Pitta-like Ground-roller Atelornis pittoides NB 1 1 0 3 2 1 0 Coraciiformes Bucerotidae Southern Ground-hornbill Bucorvus leadbeateri NB 3 1 1 3 1 1 0 Anseriformes Anhimidae Southern Screamer Chauna torquata NB 1 1 1 3 3 1 1 Anseriformes Anhimidae Harlequin Duck Histrionicus histrionicus NB 1 3 0 3 5 2 1 Pelecaniformes Phaethontidae Red-tailed Tropicbird Phaeton rubricauda BB 2 3 1 3 5 3 0 Cuculiformes Cuculidae Crested Coua Coua cristata BB 3 2 0 3 2 1 0 Coliiformes Coliidae Blue-naped Mousebird Urocolius macrourus BB 1 2 1 3 2 3 0 Piciformes Capitonidae Red-and-yellow Barbet Trachyphonus erythrocephalus BB 1 1 0 3 1 2 0 Piciformes Ramphastidae Keel-billed Toucan Ramphastos sulfuratus BB 1 2 0 3 2 2 0 Piciformes Picidae Green Woodpecker Picus viridis BB 1 2 0 3 2 2 0

LEGEND

Feeding source: Feeding method: 0 = small molluscs 1 = foraging or wading Group size: 1 = Fruits, leaves, seeds, small vertebrates, 2 = collecting in trees and bushes 0 = solitary or pairs invertebrates, insects, etc. 3 = diving in water 1 = small groups (from 2 to 10 birds) 2 = small fishes, squid, crustaceans 4 = stalking and spearing 2 = big groups (up to 10 birds) 3 = small vertebrates, reptiles, frogs 5 = hunting and searching from the air 4 = vertebrates: birds and small mammals 6 = digging and scooping Activity: Bobbing Style: 5 = carrion 7 = filtering 1 = nocturnal 2 = crepuscular HB = head-bobbing 3 = diurnal BB = body-bobbing Habitat: Flying ability: NB = non-bobbing 1 = savannah and other open woodlands 0 = terrestrial Predator pressure: 2 = forests 1 = fly short distances (awkward) 0 = low predator pressure 3 = swamps, marshes and muddy shores 2 = normal flyer 1 = medium pressure 4 = lakes and rivers 3 = elegant and skilled. 2 = high predator pressure 5 = see, ocean and coastal areas

Table C-3: Analysis of the ecological and behavioural characteristics of HB and NB birds: feeding source, feeding method, group size, habitat, flying abilities, and predator pressure. References 157

REFERENCES

Alcock, J. (1997). Adaptative response to predators. In Animal Behaviour: an evolutionary approach (6th edition). Massachusset: Sinauer associates. Alonso, Y. (1998). Lateralization of visual guided behaviour during feeding in zebra finches (Taeniopygia guttata). Behavioural Processes, 43(3), 257-263. Andrew, R. J. (1988). The development of visual lateralization in the domestic chick. Behav Brain Res, 29(3), 201-209. Andrew, R. J., & Dharmaretnam, M. (1993). Lateralization and strategies of viewing in the domestic chick. In H. P. Zeigler & H. J. Bischof (Eds.), Vision, brain and behavior in Birds (pp. 319-332). Cambridge: the Mit Press. Bagnoli, P., Grassi, S., & Magni, F. (1980). A direct connection between visual Wulst and Tectum opticum in the pigeon (Columba livia) demonstrated by horseradish peroxidase. Arch Ital Biol, 118(1), 72-88. Baran-Marescot, M. (2005). Birds of the world. New York: Harry N Abrams. Bell, G. A., & Gibbs, M. E. (1979). Interhemispheric engram transfer in chick. Neurosci Lett, 13(2), 163-168. Beritov, I. S., & Chichinadse. (1935). Localisation of visual perception in the pigeon. bulletin de biologie et de médicine expérimentale, 11, 103-104. Bischof, H. J., & Watanabe, S. (1997). On the structure and function of the tectofugal visual pathway in laterally eyed birds. European Journal of Morphology, 35, 246- 254. Bloch, S., Jaeger, R., Lemeignan, M., & Martinoya, C. (1988). Correlation between ocular saccades and head movements in walking pigeons. Journal of Physiology, 406: 1-2 July(July), 173. Bloch, S., Lemeignan, M., & Martinoya, C. (1987). Coordinated vergence for frontal fixation, but independent eye movements for lateral viewing, in the pigeon. Eye movements: from physiology to cognition, 48-56. References 158

Bloch, S., & Martinoya, C. (1982a). Comparing frontal and lateral viewing in the pigeon. I. Tachistoscopic visual acuity as a function of distance. Behav Brain Res, 5(3), 231-244. Bloch, S., & Martinoya, C. (1982b). Specialization of visual function for different retinal areas in the pigeon. In J. P. Erwert, R. R. Capranica & D. J. Ingle (Eds.), Advances in Vertebrate Neuroethology (pp. 359-368). New York: Plehum Press. Brooks, B., & Holden, A. L. (1973). Suppression of visual signals by rapid image displacement in the pigeon retina: a possible mechanism for "saccadic" suppression. Vision Res, 13(7), 1387-1390. Budzynski, C. A., & Bingman, V. P. (2004). Participation of the thalamofugal visual pathway in a coarse pattern discrimination task in an open arena. Behav Brain Res, 153(2), 543-556. Casperson, L. W. (1999). Head movement and vision in underwater-feeding birds of stream, lake, and seashore. Bird Behaviour, 13, 31-46. Catania, A. C. (1965). Interocular Transfer of Discriminations in the Pigeon. J Exp Anal Behav, 47, 147-155. Chu, P. C. (1995). Phylogenetic reanalyis of Strauch's osteological data set for the Charadriiformes. Condor, 97, 174-196. Clayton, N. (1992). Lateralization and unilateral transfer of spatial memory in marsh tits. Journal of Comparative Physiology A, 171, 799-806. Clayton, N., & Krebs, J. R. (1994). Lateralization and unilateral transfer of spatial memory mash tits: are two eyes better than one? Journal of Comparative Physiology A, 174, 769-773. Clayton, N. S., & Krebs, J. R. (1994a). Hippocampal growth and attrition in birds affected by experience. Proc Natl Acad Sci U S A, 91(16), 7410-7414. Clayton, N. S., & Krebs, J. R. (1994b). memory for spatial and object-specific cues in food-storing and non-storing birds. Journal of Comparative Physiology A, 174, 371-379. Clements, J. F. (2000). Birds of the World: A Checklist (5th edition, 2004 update ed.): Ibis Pub Co. References 159

Collett, T. S. (1978). Peering-a locust behaviour pattern for obtaining motion parallax information. J. exp. Biol., 76, 237-241. Cooper, A., & Penny, D. (1997). Mass survival of birds across the Cretaceous-Tertiary boundary: molecular evidence. Science, 275(5303), 1109-1113. Cronin, T. W., Kinloch, M. R., & Olsen, G. H. (2005). Head-bobbing behavior in foraging whooping cranes favors visual fixation. Curr Biol, 15(7), R243-244. Daanje, A. (1951). On locomotory movements in birds and the intention movements derived from them. Behaviour, 3, 48-96. Dagg, A. I. (1977a). Running, Walking and Jumping: The Science of Locomotion. London: Wykeham. Dagg, A. I. (1977b). The walk of the Silver gull (Larus novaehollandiae) and of other birds. Journal of Zoology, London, 182, 529-540. Davies, M. N. O., & Green, P. R. (1988). Head-bobbing during walking, running and flying: relative motion perception in the pigeon. Journal of Experimental Biology, 138, 71-91. Davies, M. N. O., & Green, P. R. (1991). The adaptability of visuomotor control in the pigeon during landing flight. Zool Jb Physiol, 95, 331-338. Dawkins, M. S. (2002). What are birds looking at? Head movements and eye use in chickens. Anim. Behaviour, 63, 991-998. Deng, C., & Rogers, L. J. (2002). Factors affecting the development of lateralization in chicks. In L. J. Rogers & R. J. Andrew (Eds.), Comparative Vertebrate lateralization (pp. 206-246). Cambridge: Cabridge University Press. Di Stefano, M., Kusmic, C., & Musumeci, D. (1987). Binocular interactions measured by choice reaction times in pigeons. Behav Brain Res, 25(2), 161-165. Diekamp, B., Hellmann, B., Troje, N. F., Wang, S. R., & Güntürkün, O. (2001). Electrophysiological and anatomical evidence for a direct projection from the nucleus of the basal optic root to the nucleus rotundus in pigeons. Neuroscience Letters, 305, 103-106. References 160

Diekamp, B., Prior, H., & Güntürkün, O. (1999). Functional lateralization, interhemispheric transfer and position bias in serial reversal learning in pigeons (Columbia livia). Anim. Cogn., 2, 187-196. Diekamp, B., Regolin, L., Güntürkün, O., & Vallortigara, G. (2005). A left-sided visuospatial bias in birds. Curr Biol, 15(10), R372-373. Dunlap, K., & Mowrer, O. H. (1930). Head movements and eye functions of birds. Journal of Comparative Psychology, 11, 99-113. Ehrlich, D., & Saleh, C. N. (1982). Composition of the tectal and posterior commissures of the chick (Gallus domesticus). Neurosci Lett, 33(2), 115-121. Engelage, J., & Bischof, H. J. (1994). Visual wulst influences on flash evoked responses in the ectostriatum of the zebra finch. Brain Res, 652(1), 17-27. Ericson, P. G., Envall, I., Irestedt, M., & Norman, J. A. (2003). Inter-familial relationships of the shorebirds (Aves: Charadriiformes) based on nuclear DNA sequence data. BMC Evol Biol, 3(1), 16. Evans, C. S., & Evans, L. (1999). Chicken food calls are functionally referential. Anim Behav, 58(2), 307-319. Evans, C. S., Evans, L., & Marier, P. (1993). On the meaning of alarm calls: functional references in the avian vocal system. Anim Behav, 46(1), 23-28. Fernández-Juricic, E., Erichsen, J. T., & Kacelnik, A. (2004). Visual perception and social foraging in birds. Trends in ecology and evolution, 19(1), 25-31. Flanders, M. (1985). Visually guided head movement in the African chameleon. Vision Res, 25(7), 935-942. Flanders, M. (1988). Head movement co-ordination in the African chameleon. Neuroscience, 24(2), 511-517. Folta, K., Diekamp, B., & Güntürkün, O. (2004). Asymmetrical modes of visual bottom- up and top-down integration in the thalamic nucleus rotundus of pigeons. J Neurosci, 24(43), 9475-9485. Friedman, M. B. (1975). How birds use their eyes. Neural and endocrine aspect of behaviour in birds, in Wright, P. Caryl, P.G. Wowles, M., 182-204. References 161

Friedman, M. B. (1975b). Visual control of head movements during avian locomotion. Nature, 255(5503), 67-69. Frost, B. J. (1978). The optokinetic basis of head-bobbing in the pigeon. Journal of Experimental Biology, 74, 187-195. Frost, B. J., Cavanagh, P., & Morgan, B. (1988). Deep tectal cells in pigeons respond to kinematograms. Journal of Comparative Physiology, 162, 639-647. Fujita, M. (2002). Head bobbing and the movement of the centre of gravity in walking pigeons (Columba Livia). J. Zool., Lond., 257, 373-379. Fujita, M. (2003). Head bobbing and the body movement of little egrets ( Egretta garzetta) during walking. J Comp Physiol A Neuroethol Sens Neural Behav Physiol, 189(1), 53-58. Fujita, M. (2004). Kinematic parameters of the walking of herons, ground-feeders, and waterfowl. Comp Biochem Physiol A Mol Integr Physiol, 139(1), 117-124. Galifret, Y. (1968). [The various functional areas of the retina of pigeons]. Z Zellforsch Mikrosk Anat, 86(4), 535-545. Gamlin, P. D., & Cohen, D. H. (1986). A second ascending visual pathway from the optic tectum to the telencephalon in the pigeon (Columba livia). J Comp Neurol, 250(3), 296-310. Garamszegi, L. Z., Moller, A. P., & Erritzoe, J. (2002). Coevolving avian eye size and brain size in relation to prey capture and nocturnality. Proc Biol Sci, 269(1494), 961-967. Gaston, K. E. (1984). Interocular transfer of pattern discrimination learning in chicks. Brain Res, 310(2), 213-221. Goodale, M. A. (1982). Visuomotor organization of pecking in the pigeon. In J. P. Erwert, R. R. Capranica & D. J. Ingle (Eds.), Advances in Vertebrate Neuroethology (pp. 349-357). New York: Plehum Press. Goodale, M. A. (1983). Visually guided pecking in the pigeon (Columba livia). Brain, Behavior and Evolution, 22(1), 22-41. References 162

Goodale, M. A., Ellard, C. G., & Booth, L. (1990). The role of image size and retinal motion in the computation of absolute distance by the Mongolian gerbil (Meriones unguiculatus). Vision Res, 30(3), 399-413. Goodale, M. A., & Graves, J. A. (1982). Interocular transfer in the pigeon: Retinal locus as a factor. In D. Ingle, M. A. Goodale & R. Mansfield (Eds.), Advances in analysis of visual behavior (pp. 211-240). London: MIT Press. Green, P. R. (1998). Head orientation and trajectory of locomotion during jumping and walking in domestic chicks. Brain, Behavior and Evolution, 51(1), 48-58. Green, P. R., Davies, M. N., & Thorpe, P. H. (1992). Head orientation in pigeons during landing flight. Vision Research, 32(12), 2229-2234. Green, P. R., Davies, M. N. O., & Thorpe, P. H. (1994). Head-bobbing and head orientation during landing flights of pigeons. Journal of Comparative Physiology, 174, 249-256. Güntürkün, O. (1985). Lateralization of visually controlled behavior in pigeons. Physiol Behav, 34(4), 575-577. Güntürkün, O. (1997a). Avian visual lateralization: a review. NeuropReport, 8(6). Güntürkün, O. (1997b). Morphological asymmetries of the tectum opticum in the pigeon. Experimental Brain Research, 116, 561-566. Güntürkün, O. (2002). Hemispheric Asymmetry in the Visual System of Birds. In K. D. Hugdahl, R. J. (Ed.), Brain Asymmetry (pp. pp. 3-36). Cambridge: MIT Press. Güntürkün, O. (2003). Hemispheric Asymmetry in the Visual System of Birds. In K. H. R. j. Davidson (Ed.), The Asymmetrical Brain (pp. 4-36). London: MIT Press. Güntürkün, O., & Bohringer, P. G. (1987). Lateralization reversal after intertectal commissurotomy in the pigeon. Brain Res, 408(1-2), 1-5. Güntürkün, O., Diekamp, B., Manns, M., Nottelmann, F., Prior, H., Schwarz, A., et al. (2000). Asymmetry pays: visual lateralization improves discrimination success in pigeons. Current Biology, 10(17), 1079-1081. Güntürkün, O., & Hahmann, U. (1994). Visual acuity and hemispheric asymmetries in pigeons. Behav Brain Res, 60(2), 171-175. References 163

Güntürkün, O., & Hahmann, U. (1999). Functional subdivisions of the ascending visual pathways in the pigeon. Behav Brain Res, 98(2), 193-201. Güntürkün, O., Hellmann, B., Melsbach, G., & Prior, H. (1998). Asymmetries of representation in the visual system of pigeons. Neuroreport, 9(18), 4127-4130. Güntürkün, O., & Kesch, S. (1987). Visual lateralization during feeding in pigeons. Behav Neurosci, 101(3), 433-435. Güntürkün, O., Miceli, D., & Watanabe, M. (1993). Anatomy of the avian thalamofugal pathway. In H. P. Zeigler & H. J. Bischof (Eds.), Vision, Brain, and Behavior in Birds (pp. 115-135). Cambridge, MA: MIT Press. Harlid, A., Janke, A., & Arnason, U. (1997). The mtDNA sequence of the ostrich and the divergence between paleognathous and neognathous birds. Mol Biol Evol, 14(7), 754-761. Harlid, A., Janke, A., & Arnason, U. (1998). The complete mitochondrial genome of Rhea americana and early avian divergences. J Mol Evol, 46(6), 669-679. Hart, N. S., Partridge, J. C., & Cuthill, I. C. (2000). Retinal asymmetry in birds. Curr Biol, 10(2), 115-117. Hayes, B. (1982). The structural organization of the pigeon retina. In N. Osborne & G. Chader (Eds.), Progress in retinal research (Vol. 1). Oxford: Pergamon press. Hodos, W., & Erichsen, J. T. (1990). Lower-Field Myopia in Birds an Adaptation That Keeps the Ground in Focus. Vision Research, 30(5), 653-658. Horn, G. (1991). Imprinting and recognition memory: A review of neural mechanisms. In R. J. Andrew (Ed.), Neural and Behavioral Plasticity: The Domestic Chick as a Model (pp. 219-261). Oxford: Oxford University Press. Hughes, J. M. (1996). Phylogenetic analysis of the cuculidae (aves, cuculidae) using behavioral, and ecological characters. The Auk, 112(1), 10-22. Hughes, J. M., & Baker, A. J. (1999). Phylogenetic relationships of the enigmatic hoatzin (Opisthocomus hoazin) resolved using mitochondrial and nuclear gene sequences. Mol Biol Evol, 16(9), 1300-1307. Husband, S., & Shimizu, T. (2001). Evolution of the avian visual system. In R. G. Cook (Ed.), Avian visual cognition. On line: www.pigeon.psy.tufts.edu/avc/husband. References 164

Husband, S., & Shimizu, T. (1999). Efferent projections of the ectostriatum in the pigeon (Columba livia). Journal of Comparative Neurology, 406(3), 329-345. Hyde, P. S., & Knudsen, E. I. (2002). The optic tectum controls visually guided adaptive plasticity in the owl's auditory space map. Nature, 415(6867), 73-76. Jarvis, E. D., Güntürkün, O., Bruce, L., Csillag, A., Karten, H., Kuenzel, W., et al. (2005). Avian brains and a new understanding of vertebrate brain evolution. Nat Rev Neurosci, 6(2), 151-159. Jonsson, L. (1999). Birds of Europe: With North Africa and the Middle East. London: Chistopher Helm. Karten, H. J., & Hodos, W. (1970). Telencephalic projections of the nucleus rotundus in the pigeon (Columba livia). Journal of Comparative Neurology, 140, 35-52. Karten, H. J., Hodos, W., Nauta, W. J., & Revzin, A. M. (1973). Neural connections of the "visual wulst" of the avian telencephalon. Experimental studies in the piegon (Columba livia) and owl (Speotyto cunicularia). J Comp Neurol, 150(3), 253-278. Karten, H. J., & Shimizu, T. (1989). The origins of the neocortex: Connections and lamination as distinct events in evolution. Journal of Cognitive Neuroscience, 1, 291-301. Keysers, C., Diekamp, B., & Güntürkün, O. (2000). Evidence for physiological asymmetries in the intertectal connections of the pigeon (Columba livia) and their potential role in brain lateralisation. Brain Research, 852(2), 406-413. Kral, K. (1998). Side-to-sie head movements to obtain motion depth cues: A short review a researach on the praying mantis. Behavioral processes, 43, 71-77. Kral, K. (2003). Behavioural-analytical studies of the role of head movements in depth perception in insects, birds and mammals. Behav Processes, 64(1), 1-12. Land, M. F. (1999). Motion and vision: why animals move their eyes. Journal of Comparative Physiology A Sensory Neural & Behavioral Physiology, 185, 341- 352. LeBas, N. R., & Marshall, N. J. (2000). The role of colour in signalling and male choice in the agamid lizard Ctenophorus ornatus. Proc Biol Sci, 267(1442), 445-452. References 165

Lehrer, M., Srinivasan, M., Zhang, S., & Horridge, G. (1988). Motion cues provide bee's visual world with a third dimension. Nature, 332, 356-357. Levine, J. (1945a). Studies in the interrelations of central nervous structures in binocular vision: I. The lack of bilateral transfer of visual discriminative habits acquired monocularly by the pigeons. The Journal of Genetic Psychology, 67, 105-129. Levine, J. (1945b). Studies in the interrelations of central nervous structures in binocular vision: II. The conditions under which interocular transfer of discriminative habits takes place in the pigeon. The Journal of Genetic Psychology, 67, 131-142. Levine, J. (1952). Studies in the interrelations of central nervous structures in binocular vision: III. Localization of the memory trace as evidenced by the lack of inter- and intraocular habit transfer in pigeons. The Journal of Genetic Psychology, 81, 131- 142. Mallin, H. D., & Delius, J. (1983). Inter- and intraoccular transfer of colour discriminations with mandibulation as an operant in the head-fixed pigeon. Behaviour Analysis Letters, 3, 297-309. Manns, M., & Güntürkün, O. (1999). 'Natural' and artificial monocular deprivation effects on thalamic soma sizes in pigeons. Neuroreport, 10(15), 3223-3228. Mariani, A. P. (1982). Association amacrine cells could mediate directional selectivity in pigeon retina. Nature, 298(5875), 654-655. Mariani, A. P. (1987). Neuronal and synaptic organization of the outer plexiform layer of the pigeon retina. Am J Anat, 179(1), 25-39. Martin, G. R. (1993). Producing the image. In H. P. Zeigler, & Bischof, H. J. (Ed.), Vision, brain, and behavior in birds (pp. 5-25). Cambridge, MA: MIT Press. Martin, G. R., Ashash, U., & Katzir, G. (2001). Ostrich ocular optics. Brain Behav Evol, 58(2), 115-120. Martin, G. R., & Brooke Mde, L. (1991). The eye of a procellariiform seabird, the Manx shearwater, Puffinus puffinus: visual fields and optical structure. Brain Behav Evol, 37(2), 65-78. Martin, G. R., & Coetzee, H. (2004). Visual field in hornbills: precision-grasping and sunshades. ibis, 146, 18-26. References 166

Martin, G. R., & Katzir, G. (1993). Visual fields in the Stone-curlew Burhinus oedicnemus. Ibis, 136, 448-453. Martin, G. R., & Katzir, G. (1994). Visual fields and eye movements in herons (Ardeidae). Brain Behav Evol, 44(2), 74-85. Martin, G. R., & Katzir, G. (1995). Visual fields in ostriches. Nature, 374, 19-20. Martin, G. R., & Katzir, G. (1999). Visual fields in Short-toed Eagles, Circaetus gallicus (Accipitridae), and the function of binocularity in birds. Brain Behav Evol, 53(2), 55-66. Martin, G. R., & Katzir, G. (2000). Sun shades and eye size in birds. Brain Behav Evol, 56(6), 340-344. Martin, G. R., & Young, S. R. (1983). The retinal binocular field of the pigeon (Columba livia: English racing homer). Vision Research, 23(9), 911-915. Martinoya, C., Rivaud, S., & Bloch, S. (1983). Comparing frontal and lateral viewing in the pigeon. II. Velocity thresholds for movement discrimination. Behav Brain Res, 8(3), 375-385. McKenzie, R., Andrew, R. J., & Jones, R. B. (1998). Lateralization in chicks and hens: new evidence for control of response by the right eye system. Neuropsychologia, 36(1), 51-58. Mench, J. A., & Andrew, R. J. (1986). Lateralization of a food search task in the domestic chick. Behav Neural Biol, 46(2), 107-114. Miceli, D., Reperant, J., Villalobos, J., & Dionne, L. (1987). Extratelencephalic projections of the avian visual Wulst. A quantitative autoradiographic study in the pigeon Columbia livia. J Hirnforsch, 28(1), 45-57. Mihara, M., & Watanabe, S. (1982). Interocular transfer of cardiac conditioning in the pigeon. Behav Brain Res, 4(4), 411-416. Morgan, B., & Frost, B. J. (1981). Visual response characteristics of neurons in nucleus of basal optic root in pigeons. Experimental Brain Research, 42, 181-188. Nicol, A. U., Brown, M. W., & Horn, G. (1995). Neurophysiological investigations of a recognition memory system for imprinting in the domestic chick. European Journal of Neuroscience, 7(4), 766-776. References 167

Nottelmann, F., Wohlschlager, A., & Güntürkün, O. (2002). Unihemispheric memory in pigeons-knowledge, the left hemisphere is reluctant to share. Behav Brain Res, 133(2), 309-315. Nye, P. W. (1973). On the functional differences between frontal and lateral visual fields of the pigeon. Vision Res, 13(3), 559-574. Parsons, C. H., & Rogers, L. J. (1993). Role of the tectal and posterior commissures in lateralization of the avian brain. Behav Brain Res, 54(2), 153-164. Patterson, T. A., Gilbert, D. B., & Rose, S. P. (1990). Pre- and post-training lesions of the intermediate medial hyperstriatum ventrale and passive avoidance learning in the chick. Exp Brain Res, 80(1), 189-195. Pearlman, A. L., & Hughes, C. P. (1976). Functional role of efferents to the avian retina. I. Analysis of retinal ganglion cell receptive fields. J Comp Neurol, 166(1), 111- 122. Perrins, D. (1991). The photographic guide to birds of the world. New York: Reed Unternational Books Limited. Peterson, R. M. G., & Hollon, P. A. D. (1983). A field guide to the birds of Britain and Europe. Poteser, M., & Kral, K. (1995). The significance of head movements in distance discrimination in praying mantis larvae. J Exp Biol, 198(Pt 10), 2127-2137. Poteser, M., Pabst, M., & Kral, K. (1998). Proprioceptive contribution to distance estimation by motion parallax in a praying mantid. J Exp Biol, 201 (Pt 9), 1483- 1491. Pratt, D. W. (1982). Saccadic eye movements are coordinated with head movements in walking chickens. Journal of Experimental Biology, 97, 217-223. Prior, H., & Güntürkün, O. (2001). Parallel working memory for spatial location and food-related object cues in foraging pigeons: binocular and lateralized monocular performance. Learn Mem, 8(1), 44-51. Prior, H., Lingenauber, F., Nitschke, J., & Güntürkün, O. (2002). Orientation and lateralized cue use in pigeons navigating a large indoor environment. J Exp Biol, 205(Pt 12), 1795-1805. References 168

Prior, H., Wiltschko, R., Stapput, K., Güntürkün, O., & Wiltschko, W. (2004). Visual lateralization and homing in pigeons. Behav Brain Res, 154(2), 301-310. Remy, M., & Emmerton, J. (1991a). Behavioral spectral sensitivities of different retinal areas in pigeons. Behav Neurosci, 103(1), 170-177. Remy, M., & Emmerton, J. (1991b). Directional dependence of intraocular transfer of stimulus detection in pigeons (Columba livia). Behav Neurosci, 105(5), 647-652. Remy, M., & Güntürkün, O. (1991). Retinal afferents to the tectum opticum and the nucleus opticus principalis thalami in the pigeon. J Comp Neurol, 305(1), 57-70. Remy, M., & Watanabe, S. (1993). Two eyes and One World: Studies of Interocular and Intraocular Transfer in Birds. In H. P. Zeigler & H. J. Bischof (Eds.), Vision, brain and behavior in Birds (pp. 333-350). Cambridge: the Mit Press. Roberts, W. A., Phelps, M. T., Macuda, T., Brodbeck, D. R., & Russ, T. (1996). Intraocular transfer and simultaneous processing of stimuli presented in different visual fields of the pigeon. Behav Neurosci, 110(2), 290-299. Rogers, L. J. (2000). Evolution of hemispheric specialization: advantages and disadvantages. Brain Lang, 73(2), 236-253. Rogers, L. J. (2002). Advantages and disadvantages of lateralization. In L. J. Rogers & R. J. Andrew (Eds.), Comparative Vertebrate lateralization (pp. 126-153). Cambridge: Cabridge University Press. Rogers, L. J., & Andrew, R. J. (2002). Comparative Vertebrate lateralization. Cambridge: Cabridge University Press. Rogers, L. J., Zucca, P., & Vallortigara, G. (2004). Advantages of having a lateralized brain. Proc Biol Sci, 271 Suppl 6, S420-422. Saleh, C. N., & Ehrlich, D. (1984). Composition of the supraoptic decussation of the chick (Gallus gallus). A possible factor limiting interhemispheric transfer of visual information. Cell Tissue Res, 236(3), 601-609. Sandi, C., Patterson, T. A., & Rose, P. R. (1993). Visual input and lateralization of brain function in learning in the chick. Neuroscience, 52(2), 393-401. References 169

Shimizu, T., Cox, K., & Karten, H. J. (1995). Intratelencephalic projections of the visual wulst in pigeons (Columba livia). Journal of Comparative Neurology, 359(4), 551-572. Shimizu, T., & Karten, H. J. (1991). Central visual pathways in reptiles and birds: Evolution of the visual system. In J. Cronly-Dillon & R. Gregory (Eds.), Vision and Visual Disfuction, Vol 2: Evolution of the eye and visual system (pp. 421- 441). London: Macmillan Press. Shimizu, T., & Karten, H. J. (1993). The avian visual system and the evolution of the neocortex. In H. P. Zeigler, & Bischof, H. J. (Ed.), Vision, brain, and behavior in birds (pp. 103-114). Cambridge, MA: MIT Press. Sibley, C. G., & Ahlquist, J. E. (1990). Phylogeny and classification of birds: a study in molecular evolution. New Haven and London: Yale University Press. Sibley, C. G., Ahlquist, J. E., & Monroe, B. L. (1988). A classification of the living birds of the world based on DNA-DNA hybridation studies. The AUK, 105(3), 409-423. Simpson, J. I. (1984). The accessory optic system. Annual Review of Neuroscience, 7, 13- 41. Skiba, M., Diekamp, B., Prior, H., & Güntürkün, O. (2000). Lateralized interhemispheric transfer of color cues: evidence for dynamic coding principles of visual lateralization in pigeons. Brain & Language, 73(2), 254-273. Stevens, J. V., & Kirsch, W. R. (1980). Interocular transfer in pigeons of color discrimination but not motor response training. Animal Learning & Behavior, 8(1), 17-21. Stevens, J. V., & Klopfer, F. (1977). Interocular transfer of conditioning and extinction in birds. Journal of comparative and physiological psychology, 91(5), 1074-1081. Stevenson, T., & Fanshawe, J. (2001). Field Guide to the Birds of East Africa: Kenya, Tanzania, Uganda, Rwanda, Burundi. London: Princeton University Press. Tommasi, L., & Vallortigara, G. (2001). Encoding of geometric and landmark information in the left and right hemispheres of the Avian Brain. Behav Neurosci, 115(3), 602-613. References 170

Torr, G. A., & Shine, R. (1994). An ethogram for the small scincid lizar. Amphibia- Reptilia, 15, 21-34. Troje, N. F., & Frost, B. J. (1999). Evidence for active vision during the thrust-phase of the pigeon's head-bobbing. Paper presented at the Ninth Annual Meeting of the Canadian Society for Brain, Behaviour, and Cognitive Science. Troje, N. F., & Frost, B. J. (2000). Head-bobbing in pigeons: How stable is the hold phase? Journal of Experimental Biology, 203, 935-940. Ulrich, C., Prior, H., Duka, T., Leshchins'ka, I., Valenti, P., Güntürkün, O., et al. (1999). Left-hemispheric superiority for visuospatial orientation in homing pigeons. Behav Brain Res, 104(1-2), 169-178. Valenti, A., Sovrano, V. A., Zucca, P., & Vallortigara, G. (2003). Visual lateralisation in quails (Coturnix coturnix). Laterality, 8(1), 67-78. Vallortigara, G. (1992). Right hemisphere advantage for social recognition in the chick. Neuropsychologia, 30(9), 761-768. Vallortigara, G. (2000). Comparative neuropsychology of the dual brain: a stroll through animals' left and right perceptual worlds. Brain Lang, 73(2), 189-219. Vallortigara, G., Cozzutti, C., Tommasi, L., & Rogers, L. J. (2001). How birds use their eyes: Opposite left-right specialization for the lateral and frontal visual hemifield in the domestic chick. Curr Biol, 11(1), 29-33. 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 Neuroethol Sens Neural Behav Physiol, 187(12), 997-1007. van Tuinen, M., Sibley, C. G., & Hedges, S. B. (2000). The early history of modern birds inferred from DNA sequences of nuclear and mitochondrial ribosomal genes. Mol Biol Evol, 17(3), 451-457. Verstappen, M., Aerts, P., & Van Damme, R. (2000). Terrestrial locomotion in the black- billed magpie: kinematic analysis of walking, running and out-of-phase hopping. Journal of Experimental Biology, 203 Pt 14, 2159-2170. von Euler, F. (2001). Selective extinction and rapid loss of evolutionary history in the bird fauna. Proc R Soc Lond B Biol Sci, 268(1463), 127-130. References 171 von Fersen, L., & Güntürkün, O. (1990). Visual memory lateralization in pigeons. Neuropsychologia, 28(1), 1-7. Voss, J., & Bischof, H. J. (2003). Regulation of ipsilateral visual information within the tectofugal visual system in zebra finches. J Comp Physiol A Neuroethol Sens Neural Behav Physiol, 189(7), 545-553. Wallace, G. K. (1959). Visual scanning in the desert locust schistocerca gregaria Forskal. J. exp. Biol., 36, 512-525. Watanabe, M., Ito, H., & Ikushima, M. (1985). Cytoarchitecture and ultrastructure of the avian ectostriatum: afferent terminals from the dorsal telencephalon and some nuclei in the thalamus. J Comp Neurol, 236(2), 241-257. Watanabe, S. (1986). Interocular transfer of learning in the pigeon: visuo-motor integration and separation of discriminanda and manipulanda. Behav Brain Res, 19(3), 227-232. Weidner, C., Reperant, J., Miceli, D., Haby, M., & Rio, J. P. (1985). An anatomical study of ipsilateral retinal projections in the quail using radioautographic, horseradish peroxidase, fluorescence and degeneration techniques. Brain Res, 340(1), 99-108. Westheimer, G., & Blair, S. M. (1974). Unit activity in the accessory optic system in alert monkeys. Investigative Ophthalmology, 13, 533-534. Wild, J. M. (1989). Pretectal and tectal projections to the homologue of the dorsal lateral geniculate nucleus in the pigeon: an anterograde and retrograde tracing study with cholera toxin conjugated to horseradish peroxidase. Brain Res, 479(1), 130-137. Wohlschläger, A., Jäger, R., & Delius, J. (1993). Head and Eye Movements in Unrestrained Pigeons (Columba livia). Journal of Comparative Psychology, 107(3), 313-319. Wylie, D. R., & Frost, B. J. (1990). The visual response properties of neurons in the nucleus of the basal optic root of the pigeon: a quantitative analysis. Experimental Brain Research, 82, 327-336. Wylie, D. R. W., Bischof, W. F., & Frost, B. J. (1998). Common reference frame for coding translational and rotational optic flow. Nature, 392, 278-282. References 172

Wylie, D. R. W., Linkenhoker, B., & Lau, K. L. (1997). Projections of the nucleus of the basal optic root in pigeons (Columba livia) revealed using biotinylated dextran amine. Journal of Comparative Neurology, 384, 517-536. Zeier, H. (1970). Lack of eye to eye transfer of an early response modification in birds. Nature, 225(5234), 708-709. Zeigler, H. P., & Bischof, H. J. (Eds.). (1993). Vision, brain, and behavior in birds. Cambridge, MA: Mit Press.

Acknowledgements

I am very grateful to all the people who have helped me in innumerable ways through the past three years of my PhD. I also would like to thank the protagonists of this work: the birds.

Professor Nikolaus F. Troje gave me the opportunity to work in this stimulating project, without him, this work would have never been possible. I thank him for his enthusiasm, encouragement, honesty, and bright ideas during this time.

I would like to thank Professor Onur Güntürkün for his wise advice, support, insights and suggestions that helped me carry out the different studies of this work.

I am very thankful to the International Graduate School of Neuroscience (IGSN) of the Ruhr- University Bochum, for providing everything that a PhD student may need: research and educational opportunities, support, and funding.

Special thanks to the enthusiastic research assistants that helped collect the data. Katrin Stoppa and Antigoni Marioli helped me with the pigeons training. The time we shared training the birds turned into priceless data for my thesis and into fruitful friendships. Evgeny Bobrov did a wonderful bibliographic search and helped to rank the birds’ characteristics. Luisa Moratalla patiently organized and digitalized a great deal of the birds’ video collection.

Many thanks to all that participated in the construction of the experimental device. The workshop of the Psychology department constructed the chamber; Thomas Jakubowski, programmed the first version of the experimental arena software; Andre Meiske kept the computers and the data safe and healthy. Special thanks to Tobias Otto who constructed and programmed the motion capture system and developed a new software for the experimental arena. He helped me in so many aspects during these three years that without him my live in Germany would be more complicated.

I am grateful to Elena Giakoumaki, Lea Soei, and Cord Westhoff for their valuable hints, suggestions, and helpful comments on a preliminary version of my dissertation. In addition, I would also like to thanks Cord Westhoff for his time, patient and brilliant sense of humor.

I am thankful to all members of the Biomotion Lab, Biopsychology department, and IGSN, particularly to all who helped me on the background without being noticed: colleges, administration staff, technicians, pigeon keepers, etc.

Thanks to my former adviser in Spain Dr. Pilar Herreros de Tejada, she introduced me to world of science.

Juan M. Sánchez Peral, my husband, taught me all I know about birds’ identification and observation, without him the last part of this job would not have been possible: all my love for him.

My foremost thank to all that make my life extremely happy: my husband, parents, brothers, nephews, Selim, and friends.

Curriculum Vitae

1. Personal Data:

Name: Laura Jiménez Ortega Date of birth: 20.06.1975 Place of birth: Madrid, Spain Address: Inst. für kognitive Neurowissenschaft Fakultät für Psychologie Ruhr-Universität-Bochum 44780 Bochum, Germany

2. Educational background:

- Primary school (1979-1990). Colegio J.A.B.Y.: Madrid - High school (1990-1995). Instituto Palas Atenea: Madrid - Degree in General Psychology (1995-1998). Universidad Complutense de Madrid - Specialization in Cognitive Psychology (1998-2000). Universidad Complutense de Madrid - M.Sc. Courses: Research Method in Psychology (2000-2001). University College of London - Advanced research studies in Pshycology (2001). Universidad Complutense de Madrid - PhD in Neuroscience: Fellowship (2003). International Graduate School of Neuroscience: Ruhr-Universität-Bochum

3. Research experience

- Research assistant (1998-1999). Universidad Complutense de Madrid, Department of Psychobiology, Animal Visual Lab. Researching and teaching in animal behaviour

- Contrast sensibility function in mice (1999-2000). Universidad Complutense de Madrid, Department of Psychobiology, Animal Visual Lab. Research project for the master thesis

- Research collaboration (September-December 2000). “University College of London, Visual Research Unit”. Adult visual asymmetries using VEP

- Honorific collaborator in Psychobiology Department (January 2001-May 2002). Universidad Complutense de Madrid, Department of Psychobiology, Animal Visual Lab. Electrophysiology in rats and mice.