Understanding the building blocks of avian complex cognition: the executive caudal nidopallium and the neuronal energy budget

by

Kaya von Eugen

A thesis submitted in partial fulfilment of the requirements for the degree of

Philosophiae Doctoris (PhD) in Neuroscience

From the International Graduate School of Neuroscience

Ruhr University Bochum

September 30th 2020

This research was conducted at the Department of Biopsychology, within the Faculty of Psychology at the Ruhr University under the supervision of Prof. Dr. Dr. h.c. Onur Güntürkün

Printed with the permission of the International Graduate School of Neuroscience, Ruhr University Bochum

Statement

I certify herewith that the dissertation included here was completed and written independently by me and without outside assistance. References to the work and theories of others have been cited and acknowledged completely and correctly. The “Guidelines for Good Scientific Practice” according to § 9, Sec. 3 of the PhD regulations of the International Graduate School of Neuroscience were adhered to. This work has never been submitted in this, or a similar form, at this or any other domestic or foreign institution of higher learning as a dissertation.

The abovementioned statement was made as a solemn declaration. I conscientiously believe and state it to be true and declare that it is of the same legal significance and value as if it were made under oath.

Bochum, 30.09.2020

Kaya von Eugen

PhD Commission

Chair: PD Dr. Dirk Jancke

1st Internal Examiner: Prof. Dr. Dr. h.c. Onur Güntürkün

2nd Internal Examiner: Prof. Dr. Carsten Theiß

External Examiner: Prof. Dr. Andrew Iwaniuk

Non-Specialist: Prof. Dr. Patrik Krieger

Date of Final Examination: 9th of December 2020

PhD Grade Assigned: magna cum laude

If a may think, its thoughts are not so small, For it may think of skies or hills or anything at all.

So a child may think, thoughts big and free and wide— It's good for and children, thoughts need not fit inside.

~Annette Wynne

Table of Contents

List of Figures ...... I List of Tables ...... III List of Abbreviations ...... V Abstract ...... IX

Chapter 1 General introduction ...... 1 1.1 Introduction ...... 3 1.2 Phylogeny of mammals, reptiles and birds ...... 5 1.2.1 The last common ancestor of amniotes ...... 5 1.2.2 Synapsids and mammals ...... 6 1.2.3 Sauropsids and reptiles ...... 7 1.3 Behaviour and cognition of mammals, reptiles and birds ...... 8 1.3.1 Cognition and complex cognition in birds, reptiles and mammals ...... 9 1.4 of mammals, reptiles and birds ...... 12 1.4.1 General lay-out ...... 12 1.4.2 The in reptiles and birds ...... 14 1.4.3 The pallium in mammals ...... 14 1.4.4 The question of homology ...... 15 1.4.5 A for (complex) cognition ...... 17 1.5 Aims and hypotheses of this thesis ...... 19

Chapter 2 Dopaminergic innervation reveals a possible primordial NCL-like structure in the Nile crocodile (Crocodylus niloticus) ...... 23 2.1 Introduction ...... 25 2.2 Material & Methods ...... 27 2.2.1 A note on nomenclature ...... 27 2.2.1.1 Elaboration on the use of ‘reptiles’ ...... 27 2.2.1.2 Neuroanatomy ...... 27 2.2.2 Immunohistochemistry ...... 28 2.2.3 TH-fibre density analysis ...... 28 2.3 Results ...... 29 2.4 Discussion ...... 32

Chapter 3 The dopaminergic innervation of the executive caudal nidopallium; a comparative study of pigeon, chicken, zebra finch and carrion crow...... 37 3.1 Introduction ...... 39 3.2 Material & Methods ...... 43 3.2.1 and tissue preparation ...... 43 3.2.2 Immunohistochemistry ...... 44 3.2.2.1 Tyrosine Hydroxylase (TH) ...... 44 3.2.2.2 Gallyas silver impregnation ...... 45 3.2.3 Data acquisition ...... 45 3.2.4 TH-fibre density estimation ...... 46 3.2.5 Close-up analysis ...... 46 3.3 Results ...... 47 3.3.1 TH+ fibre innervation ...... 47 3.3.1.1 Pigeon ...... 49 3.3.1.2 Chicken ...... 52 3.3.1.3 Carrion crow ...... 55 3.3.1.4 Zebra finch ...... 59 3.3.2 Myelinated trajectory of the dorsal arcopallial tract ...... 63 3.4 Discussion ...... 67 3.4.1 Pigeon ...... 67 3.4.2 Chicken...... 69 3.4.3 Carrion crow and zebra finch ...... 70 3.4.4 Functions of the NCL modulated by ...... 72 3.4.5 An evolutionary perspective of the caudal nidopallium ...... 75

Chapter 4 Neurons in the avian brain consume three times less glucose compared to mammals ...... 77 4.1 Introduction ...... 79 4.2 Material and methods ...... 80 4.2.1 Animals & housing ...... 80 4.2.2 PET measurements ...... 81 4.2.3 Input function ...... 83 4.2.4 Image analysis...... 83 4.2.5 Kinetic modelling ...... 84

4.2.6 Statistics ...... 86 4.3 Results ...... 86 4.4 Discussion ...... 95 4.4.1 Methodological considerations and validation ...... 95 4.4.1.1 Lumped constant ...... 96 4.4.1.2 Route of injection ...... 96 4.4.1.3 State and brain region ...... 97 4.4.1.4 Comparison to other bird PET study ...... 98 4.4.2 Neuronal energy budget of mammals and birds ...... 99

Chapter 5 General discussion ...... 105 5.1 Summary ...... 107 5.2 The evolution of an executive structure in the archosaurian lineage ...... 108 5.2.1 The NCL in reptiles, birds and dinosaurs ...... 108 5.2.2 Size and subdivisions ...... 109 5.2.3 Three proxies for reconstruction ...... 110 5.2.4 Endothermy, behaviour and the brain ...... 112 5.2.5 The origin and evolution of the NCL ...... 113 5.3 The brain in a flying body ...... 120 5.4 Conclusion and future directions ...... 124 5.4.1 The executive structure in sauropsids ...... 125 5.4.2 Avian neuronal energy budget ...... 125

References...... 131 Appendix ...... 171 A Curriculum Vitae ...... 173 B List of publications ...... 175 C Acknowledgements ...... 177

List of Figures

Figure 1 Simplified overview of phylogenetic tree of amniotes ...... 6 Figure 2 Overview of different brain structures in amniotes...... 13 Figure 3 Catecholaminergic innervation of the region of highest fibre density in the DVR of the Nile crocodile ...... 30 Figure 4 Quantification of baskets and boutons-en-passant in the dorsal ventricular ridge (DVR) of the Nile crocodile ...... 31 Figure 5 TH labelling in frontal sections of the mesencephalon...... 48 Figure 6 Series of exemplary heat maps representing TH+ fibres densities of one pigeon .. 50 Figure 7 Morphology of TH+ positive fibres and baskets of different structures in the caudal nidopallium in pigeon ...... 51 Figure 8 Series of exemplary heat maps representing TH+ fibre densities of one chicken .. 53 Figure 9 Morphology TH+ fibres and baskets in different structures of the caudal nidopallium in chicken ...... 54 Figure 10 Series of exemplary heat maps representing TH+ fibres of one carrion crow ...... 57 Figure 11 Morphology of TH+ fibres and baskets in the different NCL-like subareas of the caudal nidopallium in carrion crow...... 58 Figure 12 Series of exemplary heat maps representing TH+ fibre densities of zebra finch .. 61 Figure 13 Morphology of TH+ positive fibres and baskets in different structures in the caudal nidopallium in zebra finch ...... 62 Figure 14 Representative frontal brain slide of the caudal nidopallium stained with the Gallyas silver impregnation technique ...... 64 Figure 15 Close-up of DA in all species, visualized with the Gallyas silver impregnation technique ...... 66 Figure 16 Schematic representation of the NCL ...... 74 Figure 17 Overview of experimental set-up ...... 84 Figure 18 Example PET scan ...... 89 Figure 19 Arterial IF and TAC...... 90 Figure 20 Model fits...... 91 Figure 21 Boxplots of the average CMRglc of forebrain and cerebellum per group ...... 96 Figure 22 CMRglc per whole-brain neuron across species...... 101 Figure 23 Simplified overview of the origin and evolution of the NCL within the archosaurian lineage ...... 121

I

II

List of Tables

Table 1 Overview of four selected cognitive capacities in representative members of reptiles, birds and mammals...... 11 Table 2 Overview experimental subjects...... 84 Table 3 Lumped constant (LC) per group...... 92 Table 4 Whole-brain kinetic rate constants of the i.v. awake group ...... 93 Table 5 Whole-brain kinetic rate constants of the i.v. anesthetized group...... 93

Table 6 Whole-brain rate constant Ki per group...... 94 Table 7 CMRglc per group...... 95

III

IV

List of Abbreviations

14C-DG 2-deoxy-D-14C-glucose 18-FDG fluorodeoxyglucose F-18 AA anterior arcopallium AC anterior commissure AD dorsal arcopallium ADVR anterior dorsal ventricular ridge AI intermediate arcopallium AId dorsal intermediate arcopallium AIv ventral intermediate arcopallium AM medial arcopallium AP posterior arcopallium AV ventral arcopallium Bq becquerel

18 CB blood concentration F-FDG CDL dorsolateral corticoid area

CMRglc cerebral metabolic rate glucose 18 CP plasma concentration F-FDG 18 CT tissue concentration F-FDG DA dorsal arcopallial tract DLA dorsolateral amygdala dMI dorsal intermediate mesopallium DVR dorsal ventricular ridge E entopallium Eb entopallial belt GLUT glucose transporter HA apical part of the hyperpallium HD densocellular part of the hyperpallium HP HVC formal name i.p. intraperitoneal i.v. intravenous IHA interstitial part of HA

18 K1 F-FDG transfer rate plasma to tissue

18 k2 F-FDG transfer rate tissue to plasma 18 k3 F-FDG phosphorylation rate

V

VI

18 k4 F-FDG dephosphorylation rate 18 Ki F-FDG net uptake rate KP Kreide-Paleogene = Creataceous-Paleogene L Field L LAD dorsal arcopallial lamina lat NCL lateral subdivision of the NCL LC lumped constant LSt lateral M mesopallium MBq mega becquerel med NCL medial subdivision of the NCL Mya million years ago NC caudal nidopallium NCC caudo-central nidopallium NCIF island fields of the caudal nidopallium Ncl caudolateral part of the nidopallium in the crocodile NCL nidopallium caudolaterale NCLd dorsal nidopallium caudolaterale NCLdc caudal aspect of the dorsal nidopallium caudolaterale NCLdr rostral aspect of the dorsal nidopallium caudolaterale NCLl lateral nidopallium caudolaterale NCLm medial nidopallium caudolaterale NCLv ventral nidopallium caudolaterale NCM caudo-medial nidopallium NF frontal nidopallium Ov nucleus ovoidalis PDVR posterior dorsal ventricular ridge PET positron emission tomography PFC prefrontal cortex PT - RA robust nucleus of the arcopallium ROI region of interest SN substantia nigra TAC time activity curve TH tyrosine hydroxylase

VB fractional blood volume VTA ventral tegmental area

VII

VIII

Abstract

Only little more than half a century ago did the scientific community start to understand the complexity of the avian pallium. The wealth of behavioural data that was gathered in parallel pointed towards the same conclusion; in contrast to what one might expect from such a small brain, the birdbrain is capable of complex cognition. In order to understand how the avian brain generates complex cognition it is crucial to both consider socio-ecological challenges that necessitated the development of certain structures and circuits, and place this within the framework of the ancestral bauplan that informs of both the possibilities, but also the limitations. This can be achieved by taking a comparative approach and study reptiles and mammals in parallel to birds. In both the synapsid (mammals) and sauropsid (reptiles and birds) lineage, complex cognition evolved as a successful strategy, and, in the ~300 million years of separate evolution, developed different brains to facilitate this. The cerebral divergence occurred with the split of the synapsids and sauropsids, and the corresponding extant descendants of mammals and birds/reptiles are characterized by either a six-layered cortex or a densely packed collection of nuclei respectively. This thesis investigated the building blocks of avian complex cognition by studying two specific phenomena: the anatomy of the executive caudal nidopallium and the neuronal energy budget. The crucial structure in the caudal nidopallium involved in complex cognition is the nidopallium caudolaterale (NCL). This area is analogous to the mammalian prefrontal cortex (PFC), and has up till now been identified and described in pigeons only. Nothing is currently known about a possible NCL in any of the reptilian sister branches to the birds, nor how it might vary across the avian lineage. In this thesis, in order to pinpoint a possible origin of the NCL, I investigated the presence of this area in the Nile crocodile (Crococylia niloticus). To get a better understanding of the extent and diversification of the NCL across the crown birds, I analysed the location and trajectory in the different bird taxa pigeon (Columba livia), chicken (Gallus gallus domesticus), zebra finch (Taeniopygia guttata), and carrion crow (Corvus corone). The NCL can be delineated with a immunohistochemical stain against tyrosine hydroxylase (TH), the rate- limiting enzyme in the production of dopamine, and it is visible as the area of highest TH+ fibre innervation within the caudal nidopallium. The number of fibres were counted with a custom- made quantification program, and combined with a close-up analysis of NCL-specific dopaminergic cell and fibre morphology. Moreover, I visualized the myelinated fibre tract that runs between the dorsal arcopallium (AD) and NCL with the Gallyas silver impregnation stain. In the Nile crocodile, I discovered an area of higher dopaminergic innervation in a topologically equivalent location to the NCL in birds. This possibly demonstrates the presence of an NCL-

IX

like structure in at least the last common ancestor of birds and crocodiles ~245 million years ago. In addition, the four different avian taxa showed variation in both location and extent of the NCL depending on order and even family. Namely, compared to pigeon and chicken, the Passeriformes can be characterized by an NCL that spans across the entire back of the caudal forebrain and can be partitioned in at least four subdivisions. This indicates that avian species with more complex capacities have an extended and more parcellated NCL. In the second part of my thesis, I researched the neuronal energy budget. Recently it was demonstrated that birds, and passerines and psittacines in particular, are characterised by much higher neuronal densities compared to similarly-sized mammals. It was proposed that this might be one of the key components in the facilitation of avian complex cognition. However, neural tissue is extremely metabolically expensive, and this raised the question, how are they able to sustain such high numbers of neurons? Here, I visualized glucose uptake in the awake and anesthetized pigeon (Columba livia) with positron emission tomography (PET) and fluorodeoxyglucose F-18 (18F-FDG) as radiotracer. Combined with kinetic modelling, it is possible to quantify the exact cerebral metabolic rate of glucose consumption (CMRglc), and divided by the known neuron numbers it will provide the neuronal energy budget. I discovered that compared to mammals, a neuron in the pigeon brain is three times more energy-efficient, which possibly explains how they are able to pack such high numbers of neurons in their brain. The exact mechanism of how the neurons attain these low metabolic costs is currently not understood. The findings in my thesis allowed for a reconstruction of the origin and evolution of the NCL. This posits the NCL first emerged in the basal archosaur ~245 mya, and evolved in concert to several brain expansions that occurred in parallel to the evolution of endothermy and the benefits of enhanced behavioural flexibility. This reconstruction places the capacity for sustained flight in a key position in the development of a large and neuron-dense brain, as observed in crows and parrots. Possibly, flight might have been a driving factor in the development of an aerobic physiology, and facilitated an increased action radius and more complex behavioural repertoire. In parallel, this required a highly developed nervous system, which in turn could be sustained by increased metabolic rates. The energy-efficient neurons are a crucial component in the facilitation of complex cognition, since with higher neuronal costs it would have been difficult to sustain both flight and high neuron numbers. Thus, this thesis showed that complex cognition in the small avian brain is attained by an expanded and diversified executive caudal nidopallium that is high in energy-efficient neuron numbers, and it posits flight as one of the key factors.

X

Chapter 1 General introduction

1

2

1.1 Introduction

The term birdbrain came into in the 1930s to describe ‘a stupid person’ (Cambridge Dictionary, n.d.). Bird brain are small, which invites the easy conclusion that therefore birds must also be stupid. Almost a century later, this term needs revision: birds have been found to rank among the most cognitively able classes among modern fauna. For example, pigeons (Columba livia) are masters of categorization (Güntürkün, Koenen, Iovine, Garland, & Pusch, 2018; Levenson, Krupinski, Navarro, & Wasserman, 2015), and chickens (Gallus gallus domesticus) show basic numerical understanding (Rugani, Vallortigara, Priftis, & Regolin, 2015). Even more striking are corvids, such as ravens (Corvus corax), who know what another raven can and cannot see and hide food accordingly (Bugnyar, Reber, & Buckner, 2016), and parrots , who besides the capacity of learning and adapting vocal responses (Bradbury & Balsby, 2016) can readily solve complex tasks by selecting from different available tools (Auersperg, von Bayern, Gajdon, Huber, & Kacelnik, 2011). Especially these latter capacities were for a long time considered uniquely primate, but today corvids and parrots are deemed behaviourally and cognitively on par (Güntürkün & Bugnyar, 2016).

Because the last common ancestor of birds and mammals lived more than 300 million years ago (mya, Benton & Donoghue, 2006), these cognitive similarities are considered a clear example of convergent evolution (Emery & Clayton, 2004; Seed, Emery, & Clayton, 2009). It is thought that early mammals and birds faced comparable socio-ecological challenges for which cognition was a successful strategy (Lefebvre, Reader, & Sol, 2004a; Sol, 2009a). This success is visible in the fact that birds and mammals underwent a parallel explosive radiation, and in modern times comprise of approximately 10,000 and 6,500 species respectively that have invaded almost all available ecological niches around the globe (Burgin, Colella, Kahn, & Upham, 2018; Jetz, Thomas, Joy, Hartmann, & Mooers, 2012). It is, however, remarkable that birds display an equal evolutionary success story in terms of cognition, with brains that are significantly smaller compared to mammals. How does the birdbrain churn out such cognitive prowess?

3

Of course, absolute brain size has not been considered the appropriate proxy for cognitive capacities for a long time (Dicke & Roth, 2016). And indeed, size is not the only difference between avian and mammalian brains. In the millions of years of separate evolution, the forebrain of both lineages organized in radically different ways. The cerebral divergence occurred with the split of the synapsids and sauropsids, and the corresponding extant descendants of mammals and birds/reptiles are characterized by either a six-layered cortex or a densely packed collection of nuclei respectively (Jarvis, 2009). Despite the sauropsid- specific cerebral organization, there are numerous examples of comparable cerebral processes that rely on analogous functional areas which operate within a similar circuitry as the neocortex (Kröner & Güntürkün, 1999; Shanahan, Bingman, Shimizu, Wild, & Güntürkün, 2013; Stacho et al., 2020). These similarities underpin the comparability of cognition and behaviour in birds and mammals (Güntürkün & Bugnyar, 2016; Iwaniuk, 2017). Next to prominent similarities in computation between the brains of birds and mammals, striking differences are also present. And some of these avian-specificities bring about certain advantages. Namely, compared to the neocortex of mammals, birds pack much higher numbers of neurons in their forebrain. Since a neuron is the basic unit of computation, it has been suggested that this is a key factor in how such a small brain can generate complex behaviours (Olkowicz et al., 2016).

In order to better understand how the small brains of birds give rise to such impressive cognitive prowess, it is thus important to discuss brain and behaviour in the context of phylogenetic history and socio-ecological challenges (Striedter, 2005; Wylie, Gutierrez-Ibanez, & Iwaniuk, 2015). This can be achieved by taking a comparative approach and study reptiles and mammals in parallel to birds. Reptiles are the sister branch to birds, and have a comparable cerebral bauplan, but did not develop equal cognitive feats (Jarvis, 2009). In contrast, mammals are phylogenetically more distant, have a radically dissimilar pallial organization, but are cognitively on par (Güntürkün & Bugnyar, 2016). The following introduction will discuss the phylogeny, behaviour and brains of birds, reptiles, and mammals within the context of environment and ancestry. This will identify specific similarities and differences between mammals and birds, that can be linked to (complex) cognition. Specifically, two topics are identified that will be discussed in more detail in the introduction of the chapters that follow. This chapter will close with the aims and hypotheses of this thesis.

4

1.2 Phylogeny of mammals, reptiles and birds

1.2.1 The last common ancestor of amniotes Birds, reptiles and mammals all belong to the group of amniotes, this is a monophyletic group that diverged from amphibians approximately 340 mya in late (Benton & Donoghue, 2006). The basal amniotes are characterized by the development of the cleidoic egg, in which the embryo is enclosed by foetal membranes, which enabled reproduction outside of the aquatic environment and was thus the main facilitator of terrestrialization (Güntürkün, Stacho, & Ströckens, 2017; Mess, Blackburn, & Zeller, 2003). In addition, amniotes lost specific ‘fish’ characteristics in the form of a larger body, flexible neck, stronger limbs, sense organs and new metabolic pathways (Carroll, 1988; Malatesta et al., 2020). They also had a larger skull, which could indicate a relatively larger brain compared to ancestral anamniotes (Evans, 2008), though still small in comparison to the extant amniotes (Nomura, Murakami, Gotoh, & Ono, 2014).

The oldest unequivocal amniote fossils are the Hylonomus lyelli and Protoclepsydrops haplous from Joggins, Nova Scotia (Carroll, 1988; Reisz & Modesto, 1996; Reisz, 1997). These two fossils represent the basal dichotomy of the amniotes that occurred early in the amniote phylogenetic tree (Figure 1). Hylonomus is considered the oldest known sauropsid; the lineage that gave rise to extant reptiles, including birds, and all extinct reptilian sister branches. Protoclesydrops is the oldest known ‘mammal-like reptile’, or synapsid, which includes all pre- mammals and extant mammals (Reisz, 2014). From morphological evidence, the dichotomy of sauropsids and synapsids was originally based on a specific feature of the skull, namely the number of cranial arches (-/ἁψίς, ‘arch’) visible as symmetrical lateral openings (temporal fenestrae) posterior to the orbit. Since the temporal and other cranial fenestrae are linked to jaw adductor musculature, the variation in fenestrae emergence is thought to represent differences in feeding styles of early amniotes (Ford & Benson, 2020; Frey, Tarsitano, Oelofsen, & Riess, 2001). The sauropsids include the anapsids and diapsids, which means no (an-/ ἀν-, ‘not’) and two (di-/δί, ‘two’) fenestrae respectively, and the synapsid condition is the presence of only one (syn-/σίν-, ‘one’) opening (Frey et al., 2001). Traditionally, the anapsids were thought to represent the ancestral condition from which one or two fenestrae were acquired in the synapsid and diapsid lineage respectively. The only living representatives of the anapsids are the turtles, and this hypothesis placed them outside of the reptilian clade as a basal amniote before the split of modern mammals and reptiles. Modern genomic and paleontological evidence has shown that the anapsid condition represents a secondary loss of two temporal fenestrae, and that they emerged from the diapsid lineage (Bever, Lyson, Field, & Bhullar, 2015; Hedges & Poling, 1999; Field et al., 2014; Wang et al., 2013). What is important to note, and that is exactly what this classical example of the turtle shows, is that the

5

early amniote evolution is far from being fully resolved. For example, most recently it was suggested that the extinct group of varanopids, classically defined as synapsids, is in fact a sister group to the diapsids, thus belonging to early reptiles (Ford & Benson, 2020).

Figure 1 Simplified overview of phylogenetic tree of amniotes. The main divergence between synapsids and sauropsids occurred around 320 mya, giving rise to mammals and reptiles, including birds, respectively. Within the mammalian lineage, the Monotremata diverges around 166 mya, and marsupials split off placental mammals approximately 140 mya. Most of the placental radiation occurred around the -Triassic border 66 mya. In the sauropsid lineage, Lepodosauria diverged approximately 278 mya, this branch gave rise to the Squamata (lizards and snakes) and Rhynocephalia (Sphenodon) who split around 240 mya. From the other branch, the extant orders are the Testudines and the Archosauria, who diverged at 255 mya. The Archosauria split into Crocodylia and the Aves approximately 245 mya. Since this thesis focusses on Aves, specific species have been highlighted. The Galliformes diverged from the around 72 mya, and the Columbiformes branched off from the Passeriformes around 64 mya. The split between the Estrildidae and Corvidae occurred approximately 54 mya. Silhouettes are from phylopic.org, who operates under a public license.

1.2.2 Synapsids and mammals After the split from sauropsids, synapsids dominated during the Permian. This represents the first case of successful radiation in amniotes, and during this time the most dominant groups were pelycosaurs and therapsids. These species developed a parasagittal stance as opposed

6

to the sprawling locomotion of reptiles, a plant-based diet high in fibres, and larger bodies up to the size of a hippopotamus. Towards the end of the Permian, ~80% of the terrestrial vertebrates, and almost all synapsid lineages, went extinct in the Permian-Triassic (PT) extinction set off by volcanic outbursts. Only a few smaller species of therapsids survived, among which the cynodonts from which basal mammals derived in Late Triassic (Benton, 2012; Brocklehurst, Kammerer, & Fröbisch, 2013; Evans, 2008; Goffinet, 2017). During this time, most of the terrestrial fauna was dominated by dinosaurs, which forced stem mammals into ecological insignificance (Benton, Forth, & Langer, 2014). These early mammals were small shrew-sized creatures with a nocturnal life-style and an insect-based diet (Kemp, 2006). During this less turbulent time, two larger points of divergence occurred within the mammalian lineage. The first split is dated to approximately 166 mya, and separated Prototheria, which gave rise to the extant monotremes, from Theria. A bit later around 140 mya, a dichotomy of the Theria resulted in contemporary subclades marsupials and placentals (Lefèvre, Sharp, & Nicholas, 2010; Luo, 2007). The fossil evidence suggests that the second pulse of explosive radiation, of especially the placental mammals, only occurred after mass extinction at the Cretaceous-Paleogene border 66 mya that wiped out all the dinosaurs (KPg extinction, Renne et al., 2013). More recent molecular data pinpoint several major diversifications, especially of interordinal cladogenesis, within the mammalian lineage before the mass extinction event (dos Reis et al., 2012; Springer et al., 2017).

1.2.3 Sauropsids and reptiles Immediately after their emergence, sauropsids were in evolutionary terms not that successful and remained small compared to the synapsids. The living representatives of this lineage include the Lepidosauria, Testudines and Archosauria, which diverged early in sauropsid evolutionary history. The Lepidosauria diverged in Permian approximately 278 mya, and gave rise to the squamates and Rhynocephalia which split approximately 240 mya. Squamates are now made up by lizards and snakes, and only one genus represents the Rhynocephalia, called Sphenodon or tuatara. Testudines, now turtles and tortoises, split from archosaurs just before the Permian-Triassic boundary around 255 mya. Lastly, the archosaurs, made up by modern Crocodylia and birds, diverged around 245 mya (Benton & Donoghue, 2006; Benton, Donoghue, & Asher, 2009; Chiari, Cahais, Galtier, & Delsuc, 2012; Mulcahy et al., 2012). What becomes apparent is that the term ‘reptiles’ in its conventional use refers to a paraphyletic group which includes all reptilian branches except for the monophyletic group of birds. Thus, the correct term for the collective of turtles, squamates, tuatara and crocodiles, while excluding birds, would be ‘non-avian reptiles’. However, for the ease of reading in this thesis the term ‘reptiles’ was adopted.

7

After the PT extinction event, the terrestrial fauna was predominantly dominated by the ancestors of birds, dinosaurs (Goffinet, 2017). The first dinosaurs emerged in Middle Triassic and produced a diverse group that varied tremendously in body size, preferred locomotion (bipedal versus quadrupedal), and diet (herbivorous versus carnivorous, (Benton et al., 2014; Sereno, 1999)). The first birds emerged ~160 mya and were about chicken-sized, had long arms with feathered wings and a relatively larger brain compared to contemporaries. They evolved from a subgroup of the predominantly carnivorous theropods, specifically the smaller bodied and large brained maniraptors, exemplified by the Velociraptor (Brusatte, O’Connor, & Jarvis, 2015). This group of early birds diversified for about 100 mya alongside the reign of dinosaurs, and produced a variety of ectomorphs (O’Connor, Chiappe, & Bell, 2011). The first split within the extant lineages of the modern crown group, Neornithes, occurred in early Cretaceous giving rise to the Paleognathae and (Yonezawa et al., 2017), and a second split between the Galloanserae and Neoaves occurred around 85 mya (Agnolín, Egli, Chatterjee, Marsà, & Novas, 2017). Thus, three distinct orders survived the mass extinction event at the KPg border. The explosive radiation of the Neoaves did not happen until after the KPg extinction (Brusatte et al., 2015; Prum et al., 2015).

1.3 Behaviour and cognition of mammals, reptiles and birds

Before I can examine the differences in (complex) cognition between the extant species of birds, reptiles and mammals, it is important to define the contested terms of cognition and complex cognition. ‘Cognition’ is the whole process of perception through the senses, the processing and retention of the obtained information, and the decision to act translated in to a motor command (Shettleworth, 2010). This can be viewed on a gradient, with on one end a process such as simple associative learning, and on the other end innovation or future planning. Simple associative learning describes behaviour that is solely controlled by the environment, and the product of an association of stimuli. Here, a specific set of stimuli will always evoke the same behavioural response (Dickinson, 2012). The key element for the latter processes involves the mental representation of either a goal or an object, to which the behavioural output can flexibly be adapted to (Shettleworth, 2001). This type of flexible adaptation is often described as a -down process that relies on the core of inhibition, working memory, and cognitive flexibility (Diamond, 2013). ‘Complex cognition’ then refers to the collection of higher-order cognitive faculties, and entails causal reasoning, imagination, flexibility, and prospection, which interact in a variety of ways to generate complex behaviour (Emery & Clayton, 2004).

8

1.3.1 Cognition and complex cognition in birds, reptiles and mammals Discussing and contrasting all available research is unfortunately outside the scope of this thesis, therefore I will highlight four key behaviours that are representative for what we currently know of cognition and complex cognition in birds, reptiles and mammals. For this analysis I discuss classical conditioning, reversal learning, working memory and flexible planning. The first can be explained solely by associative learning, the other three incorporate additional (complex) cognitive domains. Also, since there is a large amount of intra-class variation in terms of cognitive capacities, I will discuss representative species to accurately exemplify this diversity (Table 1).

Learning is a key component to behaviour and cognition. Non-associative learning processes, such as habituation and sensitization, are often framed as simple stimulus-response behaviours. These processes are observed in all Animalia, and do not even require the involvement of a brain-like structure (Cloninger & Gilligan, 1987). Associative learning, in which the organism pairs two stimuli or events, does involve a cognitive process (Shettleworth, 2001). For example, in classical conditioning a previously unrewarded stimulus is associated with a potent natural stimulus (such as food or a shock), and over time the previously neutral stimulus elicits a conditioned response (Gormezano, Prokasy, & Thompson, 1987). Classically, this is tested by teaching subjects that a certain colour, sound, smell or action is associated with a food reward or electrical shock. This capacity is present in all three classes and has been assessed in a multitude of ways. For example, a turtle species (Pseudemys nelsoni) could learn to knock over a bottle to obtain a food reward, and even discriminate between two types of bottles with only one rewarded (Davis & Burghardt, 2007). Birds are usually trained to peck a certain stimulus (e.g. a Picasso painting) on a screen to obtain a food reward, and can also discriminate between two presented options (e.g. Picasso versus Monet; Watanabe, Sakamoto, & Wakita, 1995), whereas rats can be trained to press a lever to receive food or water (Hull, 1977).

A more complex type of associative learning is reversal learning. Here, a subject is trained on a forced choice task between for example a green cup, containing a food reward, and a red cup, which is empty. After a certain learning criterion is reached, the reward contingencies are reversed, and now the red cup contains the reward. This task tests behavioural flexibility, which requires inhibition, and can also contain a type of abstraction. Cognitively, a distinction is made between a win-stay-lose-shift strategy or a generalized approach (Macphail, 1982). As the name implies, with the first strategy the subject will stay with a choice until it receives an error feedback signal, and then it will switch. This approach is observed in pigeons, rats, turtles, and crocodiles (Bond, Kamil, & Balda, 2007; Day, Crews, & Wilczynski, 1999; Gossette & Hombach, 1969). The general principle of reversal implies a more abstract representation of

9

the reversal of reward contingencies. If this is the case, the behavioural approach can be transferred to a different set of stimuli (e.g. from a visual to a spatial reversal task). This is a more complex capacity, since it entails the formation of a concept or a rule independent of, or, even in the absence of a concrete stimulus. This strategy is visible in primates and corvids (Bond et al., 2007; Izquierdo, Brigman, Radke, Rudebeck, & Holmes, 2017).

Next to inhibition, another core executive function is working memory, which is conceptualized as the capacity to maintain relevant information ‘online’ for a limited amount of time (Baddeley & Hitch, 1974). This capacity is commonly tested with a delayed-match-to-sample task, in which the first presented stimulus (e.g. visual, olfactory) should be remembered to after a delay period select the appropriate matching stimulus in a forced-choice task. The capacity of working memory is limited, both in number of stimuli that can be maintained and the time- period over which it can be remembered (Hahn & Rose, 2020). There are currently no studies that have tested working memory in reptiles. Comparisons of birds and mammals show comparable load capacity (Balakhonov & Rose, 2017; Buschman, Siegel, Roy, & Miller, 2011; Wright & Elmore, 2016), and pigeons perform less well than different mammalian species with increasing delay lengths (Leising et al., 2013; Lind, Enquist, & Ghirlanda, 2015). Corvids have not been tested with respect to long delay capacities.

Flexible planning is the capacity to make current decisions that will have a beneficial outcome at a different time and place. The capacity includes two cognitive domains: ‘inhibition’, or self- control of not acting on immediate needs or desires, and ‘mental time travel’, in which the subject is able to have a mental representation of later potential situations or events (Osvath & Osvath, 2008). This capacity thus includes all three core capacities of complex cognition as defined by Diamond (2013). Future planning has not been tested in reptiles, and experiments in rodents and pigeons predominantly target episodic-like memory (Babb & Crystal, 2006b; Rattenborg & Martinez-Gonzalez, 2011), which tests only the second cognitive domain of future planning. Whereas corvids and great-apes are capable to plan for the future (Kabadayi & Osvath, 2017), pigeons and rats do not show the capacity to form a coherent what-where- when memory required to demonstrate mental time travel (Bird, Roberts, Abroms, Kit, & Crupi, 2003; Skov-Rackette, Miller, & Shettleworth, 2006).

10

Table 1 Overview of four selected cognitive capacities in representative members of reptiles, birds and mammals.

Reptiles Bird Mammal Species Squamates Turtle Crocodile Pigeon Corvid Rodent Primate Classical Yes [1] Yes [2] ? Yes [3] Yes [4] Yes [5] Yes [6] conditioning Reversal Yes [7] Yes [8] Yes [9] Yes [10] Yes [11] Yes [12] Yes [13] learning

Working ? ? ? Yes [14] Yes [15] Yes[16] Yes [17] memory

Flexible ? ? ? No [18] Yes [19] No [20] Yes [21] planning

[1] Day, Ismail, & Wilczynski, 2003; [2] Davis & Burghardt, 2007; [3] Watanabe et al., 1995; [4] B. Wilson, Mackintosh, & Boakes, 1985a); [5] Hull, 1977; [6] R. E. Clark & Zola,9 1998; [7] Day et al., 1999; [8] Holmes & Bitterman, 1966; [9] Gossette & Hombach, 1969; [10] Laude, Stagner, Rayburn-Reeves, & Zentall, 2014; [11] Yes Bond et al., 2007; [12] Rayburn-Reeves, Stagner, Kirk, & Zentall, 2013; [13] Butter, 1969; [14] Bettina Diekamp, Kalt, & Gü, 2002; [15] Veit & Nieder, 2013; [16] Lu, Slotnick, & Silberberg, 1993; [17] Buschman et al., 2011; [18] Skov-Rackette et al., 2006; [19] Kabadayi & Osvath, 2017; [20] Bird et al., 2003; [21] N. J. Mulcahy & Call, 2006

From this overview two things can be concluded. First, even though the last decade has seen great progress, reptiles remain a severely understudied species, and second the evidence that is available shows that inter- and intra-class differences in cognitive capacities exist. The first point is important to consider, since this type of ‘taxonomic chauvinism’ can contribute to our current view of cognition in reptiles (sensu , Shine, & Lourdais, 2002; Burghardt, 2014). Indeed, absence of evidence is not evidence of absence, and caution is warranted when drawing conclusions on capacities such as working memory and flexible planning for which there is no data. What can be concluded from this overview is that basic cognitive capacities such as different types of associative learning are present in all three classes. The first differences start to appear in reversal learning and working memory, where either different strategies are adopted or species diverge in load or delay capacities. This does demonstrate at least one feature of complex cognition, namely working memory, is present in all birds and mammals studied so far. Clear intra-class differences appear for the capacity of flexible planning, which involves all three complex cognitive domains as defined by Diamond (2013). Importantly, corvids and great apes are behaviourally on par. With the caveat that future

11

studies might prove me wrong, based on the current evidence I am inclined to conclude birds and mammals have a more rich behavioural repertoire, facilitated by (complex) cognition, compared to reptiles. Moreover, within each lineage, there are species that demonstrate more complex behavioural repertoires that require distinct cognitive feats such as abstraction and conceptualization. Even though only four behavioural capacities were discussed here, this pattern is corroborated by a wealth of behavioural data that show comparable disparities (Güntürkün & Bugnyar, 2016; Güntürkün, Ströckens, Scarf, & Colombo, 2017b). What are the underlying brain structures that give rise to these behaviours? And are the differences in behaviour reflected by diverting neural structures?

1.4 Brains of mammals, reptiles and birds

1.4.1 General lay-out Like all vertebrates, reptiles, birds, and mammals have the same trichotomy of embryonic brain lay-out consisting of the hindbrain (rhombencephalon), midbrain (mesencephalon), and forebrain (prosencephalon). The adult brain morphology diversifies depending on which parts of the three sections enlarge. For example, one of the characteristics of birds and reptiles is a dorsal bulging of the mesencephalon that gives rise to the optic tectum, an important visual area. The mammalian homologue is known as the superior colliculus, and is much less pronounced. In both mammals and birds, and to a lesser extent reptiles, the telencephalon is relatively expanded (Yamamoto & Bloch, 2017). This is a structure that derives from the prosencephalon, and can be subdivided in the subpallium and the pallium. The subpallium gives rise to the basal ganglia, which is a highly conserved structure across vertebrates with respect to connectivity, function, and genetic developmental program (Grillner, Robertson, & Stephenson-Jones, 2013; Kuenzel, Medina, Csillag, Perkel, & Reiner, 2011; Reiner, 2016; Strausfeld & Hirth, 2013). The pallium varies extensively across amniotes, and its evolution, with associated question of homologies, remains a matter of debate among comparative scholars (Güntürkün & Bugnyar, 2016; Puelles et al., 2017). This structure is of special interest for research on the origin of complex cognition, and thus for the scope of this thesis, since in both birds and mammals it has been implicated in higher order functions such as learning, memory, motor control, among many other capacities (Loreta Medina & Abellán, 2009). Below, I will first discuss differences of the pallium in morphological lay-out in reptiles, birds and mammals, and next elaborate on the question of homologies.

12

Figure 2 Overview of different brain structures in amniotes. The brains of reptiles (A,B), birds (C,D) and mammals (E,F) are all characterized by the same layout of telencephalon, midbrain, cerebellum and brainstem. The telencephalon consists of a subpallium and pallium that can easily be identified in all three vertebrates, here depicted are coronal slices of one hemisphere stained against myelin (B,D,F). The pallium varies extensively between birds and reptiles, and mammals. Birds and reptiles have a nucleated structure known as the DVR, that can be subdivided in a mesopallium and nidopallium. Moreover, the DVR is overlaid by an additional territory known as the hyperpallium/Wulst in birds and the dorsal cortex in reptiles. In contrast, the pallium of mammals is dominated by the six-layered neocortex. Schematic brains of lizard, pigeon and hedgehog (A,C,E) taken from Naumann et al. (2015), coronal slice of crocodile (B) modified from Billings et al. (2020), coronal slice of pigeon (D) and ferret (F) modified from Bugnyar & Güntürkün (2016).

13

1.4.2 The pallium in reptiles and birds In reptiles and birds, the pallium is dominated by a dense aggregation of distinct nuclei, known as the dorsal ventricular ridge (DVR), which bulges into the lateral ventricle from the dorsolateral wall of the hemisphere. The DVR is dorsally overlaid by a thinner structure separated from the DVR by the lateral ventricle, known as the dorsal cortex in reptiles and the hyperpallium / Wulst in birds. Traditionally, the reptilian DVR was subdivided in an anterior (ADVR) and posterior (PDVR) part (Ulinski, 1983). However, recently Briscoe et al. (2018) demonstrated a high comparability in genetic markers of the alligator DVR to specified areas of the avian DVR. This analysis was corroborated in the lizard (Desfilis, Abellán, Sentandreu, & Medina, 2018), indicating is it not a specialization of archosaurs. These findings are supported by established literature demonstrating similarities in topology and connectivity, and a recent structural MRI analysis could delineate all identified subfields (Behroozi et al., 2018). Following Briscoe et al. (2018), I will employ part of the avian terminology where an explicit proposal for homology is present. Thus, the reptilian DVR can be subdivided into a mesopallium, nidopallium, and arcopallium, with additional pallial amygdala territories. In reptiles, the dorsal structure can be subdivided into the lateral, dorsal and medial cortex that consist of three layers that differ in neuronal density. Parts of this structure are comparable to the hippocampus and the piriform cortex (Striedter, 2016).

As mentioned, the avian DVR can be further subdivided in mesopallium, nidopallium and arcopallium, and additional olfactory and pallial amygdala areas (Reiner, Perkel, Bruce, et al., 2004). No clear laminar structure can be distinguished in the avian DVR, but distinct nuclei or areas can be discerned based on neuronal densities or separation by lamina or large fibre bundles (Rehkämper & Zilles, 1991). A recent analysis demonstrated the presence of columns, which is a morphological and functional feature classically only identified in mammals, see below (Stacho et al., 2020). The dorsal structure is labelled the hyperpallium or Wulst, and also holds the hippocampal complex, which has been described as semi-laminar (Reiner, Perkel, Bruce, et al., 2004; Striedter, 2016).

1.4.3 The pallium in mammals In mammals the dominant pallial structure is an ordered six-layered cellular sheet, known as the neocortex or isocortex. Like the reptilian dorsal cortex, the six layers can be morphologically distinguished based on neuronal density and cellular composition. Perpendicular to the layers, a columnar organization is characterized by a typical vertical interconnectivity (Kaas, 2012). The entire neocortex can be subdivided in distinct areas, known as Brodmann areas, based on cytoarchitectural features (Brodmann, 1909). The neocortex sits as a mantle over other pallial structures such as the pallial claustro-amygdalar complex and the hippocampus. The claustrum and amygdala do not have a layered organization, and

14

the hippocampus is characterized by three layers. Across mammalian species, the neocortex can vary in degree of cortical folding, and number of subdivisions (Kaas, 2013).

1.4.4 The question of homology This obvious dissimilarity of adult-stage brain morphology has fuelled an over-a-century long debate on the exact evolutionary homologous relationships. It seems that the hypothesized homologies are strongly influenced by the experimental approach (Montiel & Aboitiz, 2018). Back in the early 20th century, it was proposed that only the dorsal cortex / Wulst was homologous to the neocortex, and the DVR was in fact a hypertrophied striatum, and thus homologous to the mammalian basal ganglia. This was based on a histological analysis of cellular morphology, where a microscopic view of the DVR shows its basal ganglia-like features of a nucleated as opposed to laminar organization (Reiner, Perkel, Bruce, et al., 2004). At the time, a layered cortex was considered a pre-requisite for cognition, and the poorly understood basal ganglia were thought to solely be involved in instinctive behaviour. This had as a result the sauropsids were thought to only be capable of stereotyped behaviours (Reiner, 2005).

This view was overturned by the seminal work of Harvey Karten in the 1960s. With the advent of immunohistochemistry it became possible to reveal that labelling of dopamine and acetyl cholinesterase almost exclusively stained the ventral part of the forebrain. In mammals, this staining pattern was solely observed for the basal ganglia. Moreover, experiments with tract- tracing demonstrated the avian ‘neostriatum’ and parts of the Wulst received thalamic visual, auditory and somatosensory input (Karten, 1969). This type of sensory innervation was later also shown in different species of reptiles (Ulinski, 1983). This implied a crucial role of sensory processing up to then only ascribed to the neocortex. This seminal work was corroborated by contemporaries and later scientists, and this movement culminated in the Avian Brain Nomenclature Forum that revised avian brain terminology only 18 years ago (Reiner, Perkel, Bruce, et al., 2004). As mentioned, this is only now, where applicable, being adapted for the reptilian DVR also (Briscoe et al., 2018).

Related to the question of homology, Karten proposed two views: 1) The neocortex has a dual- linked homology to parts of both the DVR and the dorsal cortex / Wulst. 2) The homology is present at the level of cell types where input, interstitial and output neurons were present in the last common ancestor but organized in highly dissimilar ways in extant descendants. These views culminate into the nuclear-to-layered hypothesis, which postulates that distinct neocortical layers are homologous to different nuclei in the DVR (Karten, 2013). For example, the input neurons in layer 4 of different sensory cortices correspond to the primary pallial thalamic recipient nuclei in the sauropsid nidopallium (Jarvis, 2009). This hypothesis recently obtained a wave of experimental support from expression patterns of molecular markers during

15

development. Sauropsid input, interstitial and output cells shared conserved features with corresponding input, inter, and output cells and layers in mammals (Briscoe, Albertin, Rowell, & Ragsdale, 2018; Briscoe & Ragsdale, 2018a; Dugas-Ford, Rowell, & Ragsdale, 2012a).

The dominant alternative hypothesis is the nuclear-to-claustrum/amygdala hypothesis (Jarvis, 2009). It postulates that large parts of the sauropsid DVR are homologous to mammalian claustrum and amygdala. It was first proposed on the basis of comparable nuclear morphology and connectivity (Bruce & Neary, 1995). The recent two decades of research on genoarchitecture identified homologous transcription factors that are involved in cerebral lay- out and morphology during development (Puelles, Kuwana, Puelles, & Rubenstein, 1999; Puelles, 2014; Puelles et al., 2000). The current proposal is that based on distinct transcription factors the embryonic pallium can be subdivided in four areas: medial, dorsal, lateral and ventral (but see Abellán, Desfilis, & Medina, 2014; Desfilis et al., 2018; L. Medina, Abellán, Vicario, Castro-Robles, & Desfilis, 2016 for a proposal of the hexapartite model), which each give rise to distinct structures of the central nervous system. In summary, the medial pallium gives rise to the hippocampal complex in birds and mammals, and the medial and parts of the dorsal cortex in reptiles. The dorsal pallium generates the hyperpallium / Wulst in birds, parts of the dorsal cortex in reptiles, and the neocortex in mammals. From the lateral pallium emerges in sauropsids the mesopallium, and in mammals the dorsal claustrum and basolateral amygdala. And the ventral pallium gives rise to the nidopallium in birds and reptiles, and the ventral claustrum and lateral anterior amygdala in mammals. Because of the genoarchitectural similarities, it was proposed that the sauropsid DVR was homologous to the claustrum and amygdala in mammals, and only the dorsal cortex / hyperpallium homologous to the neocortex (Puelles et al., 2017).

An analysis of the internal organization of the brain of the last common ancestor of amniotes would significantly aid to resolve this current debate. But since cerebral tissue does not fossilize, this is unfortunately not possible. Thus, the final word on the exact homologies between sauropsids and mammals is not out yet. It could be that a combination of two are at play, or that there is the need for a different thesis altogether (Güntürkün & Bugnyar, 2016; Jarvis, 2009). What the evidence does imply, is that the last common ancestor of amniotes was equipped with two distinct territories that possibly already had input cells of different sensory modalities, and output cells connected by interstitial neurons. Through amniote evolution, the dorsal and ventral part independently gave rise to the diverse spatial lay-out of the neural substrates of cognitive processing that are present in extant amniotes; the neocortex and the DVR (Briscoe & Ragsdale, 2018b; Montiel & Aboitiz, 2018; Striedter, 2005).

16

1.4.5 A brain for (complex) cognition Even though the exact homologies are yet to be resolved, the capacity for (complex) cognition evolved independently in the sauropsid and synapsid lineage and can be traced back to the DVR and neocortex respectively (Briscoe & Ragsdale, 2018b; Güntürkün, 2012). However, as could be shown in the behavioural overview, differences are evident between and also within the different classes. Are there neural correlates that can explain the observed variation in cognitive capacities?

The major diversification of birds and mammals occurred in parallel to an increase in the relative brain-to-body size of both lineages (Striedter, 2005). Compared to same-sized reptiles, birds and mammals have 10-fold larger brains (Dicke & Roth, 2016). Though a more recent analysis, incorporating phylogenetic relationships to rule out flight-related effects on avian body size, demonstrated ‘only’ a 5-6 times difference between birds and reptiles (Font, García-Roa, Pincheira-Donoso, & Carazo, 2019). Moreover, both corvids and primates have a comparable high encephalization quotient, which describes the increase of brain size in relation to an expected brain to body ratio (Jerison, 1973). Though these parameters seem to fit nicely to our behavioural overview of inter- and intra-class cognitive variation, caution is warranted when employing comparisons of brain-body ratios (Dicke & Roth, 2016; Willemet, 2013). For one, the idea of a universal scaling law implies that the brain would enlarge or decrease in size as a uniform structure. Instead, what is more often observed, and something I touched upon before, is that the brain actually follows a mosaic/concerted-type evolution. This idea postulates that brain regions scale independently of each other, though some in concert when regions are highly interconnected (Barton & Harvey, 2000; Iwaniuk, Dean, & Nelson, 2004; Willemet, 2013). Evidence for ‘cerebrotypes’, which are brain profiles consisting of a distinct assimilation of subregion volume fractions, was found in both mammals and birds (Clark, Mitra, & Wang, 2001; Iwaniuk & Hurd, 2005).

The enlargement of the avian and mammalian brain compared to reptiles is predominantly the consequence of an expansion of the telencephalon, and the pallium in particular. More specifically, in the avian lineage corvids have a cerebrotype of a relatively enlarged nidopallium and mesopallium in relation to the other pallial areas, the cerebellum, and brainstem (Iwaniuk & Hurd, 2005; Mehlhorn, Hunt, Gray, Rehkämper, et al., 2010; Rehkämper, Frahm, & Zilles, 1991). Likewise, in mammals primates have a relatively larger neocortex in proportion to the scaling of other structures compared to other taxa (Clark et al., 2001). However, the meso- and nidopallium and neocortex are not uniform structures, but instead consist of primary, secondary and associative sensory areas, regions involved in motor control, and areas that are involved in memory, learning and emotional processes. Again, it could be that these areas scale independent to each other depending on socio-ecological needs. For example, species

17

that are experts in tool-use have relatively enlarged brain structures that are important for motor learning (Mehlhorn, Hunt, Gray, Rehkämper, et al., 2010), and nocturnal primates show a hypertrophied olfactory cortex, whereas diurnal species have expanded visual areas (DeCasien & Higham, 2019). Thus, if the topic of interest is cognition and especially complex cognition, it is important to focus on even more finely delineated areas.

In mammals and birds, the seat for complex cognition are the prefrontal cortex (PFC) and the nidopallium caudolaterale (NCL) respectively. The PFC is situated at the anterior pole of the neocortex, and the NCL occupies the caudal nidopallium, thus these structures came about by convergent evolution. The PFC and NCL are in each lineage strongly involved in the three core capacities of complex cognition: working memory, inhibition, and cognitive flexibility (Emery & Clayton, 2004; Güntürkün, 2005, 2012; Güntürkün & Bugnyar, 2016). These structures accomplish this as integration centres that receive sensory input from all modalities, reciprocally connect to visceral, limbic and memory-related structures, and send efferent projection to basal ganglia and premotor areas (Kröner & Güntürkün, 1999; Sporns & Zwi, 2004). In addition, the PFC and NCL are strongly innervated by dopamine projections from the ventral tegmental area and the substantia nigra (Gaspar, Stepniewska, & Kaas, 1992; Waldmann & Güntürkün, 1993). Dopamine is the key that is involved in the process of gating, maintaining and manipulating incoming information to subsequently initiate the appropriate action (Ott & Nieder, 2019). Whereas the PFC has been researched extensively (Fuster, 2015), especially regarding the evolution and anatomy in monkeys, great apes and humans (Passingham & Wise, 2012), the NCL has been delineated in pigeons only (Herold et al., 2011; Kröner & Güntürkün, 1999), and scarcely researched in some other species (Braun, Bock, Metzger, Jiang, & Schnabel, 1999; Nieder, 2017). Close to nothing is known about the ancestral origin of this structure, or possible changes in size and trajectory along the extensive radiation of modern birds. This is a crucial component in our understanding of avian complex cognition

An alternative to comparative volumetrics to explain what kind of brain facilitates varying cognitive capacities, is the quantification of neuron numbers. The idea being that since a neuron is the basic processing unit of the brain, the number of processing units must be an indicator of processing power. Suzanne Herculano-Houzel and collaborators demonstrated that in contrast to the prevailing idea that neuron numbers in all brains scale in the same manner, each mammalian order can be characterized by its own neuronal scaling rule. Namely, with each increase of brain size in Primata, the number of neurons will increase isometrically (Herculano-Houzel, Collins, Wong, & Kaas, 2007). This is contrast to Rodentia, where a brain that doubles in size, will contain less than double the number of neurons, thus scaling hypometrically (Herculano-Houzel, Mota, & Lent, 2006). Neuron numbers have not

18

been studied extensively in reptiles yet, but data from the Nile crocodile showed much lower neuron numbers and densities. Namely, whereas it has a slightly larger brain than a squirrel, it has six times fewer neurons (Ngwenya, Patzke, Manger, & Herculano-Houzel, 2016). Recently, it was shown that all birds have higher neuron densities compared to all similar-sized mammals, when body size is taken into account. This was especially the case for parrots and songbirds, to which corvids belong, that pack more than twice the number of neurons compared to a similarly-sized mammalian brain. Interestingly, on average more than 50% of the neurons are located in the pallium, whereas in mammals the cerebellum contains most of the neurons (Olkowicz et al., 2016). It has been proposed that this is an important feature of the bird brain that allows them to display such remarkable feats with such small brains. They can accomplish to generate the required computational power by densely packing neurons togethers that can build complex networks with short inter-neuronal distances. An open question remains, namely, how are birds able to sustain such high numbers of neurons? Neural tissue is one of the most expensive tissues of the vertebrate body (Mink, Blumenschine, & Adams, 1981), and total cerebral glucose consumption is a linear product of the absolute number of neurons (Herculano-Houzel, 2011a). Moreover, in mammals it was shown that metabolic limits restrict the extend of brain expansion (Aiello & Wheeler, 1995; Isler & van Schaik, 2006b). In order to understand how birds bolster many more neurons compared to similarly-sized mammals, it is crucial to know the neuronal energy budget of the avian brain. This will help to understand another building block of the generation of cognition by the small avian brain.

1.5 Aims and hypotheses of this thesis

At the beginning of this introduction, I asked how the small avian brain is able to churn out such cognitive prowess. The answer to this question lies in a dual-approach where it is crucial to both consider socio-ecological challenges that necessitated the development of certain structures and circuits, and place this within the framework of the ancestral bauplan that informs of both the possibilities but also the limitations. With regards to the socio-ecological challenges, it is useful to take a comparative approach; in both the mammalian and avian lineage (complex) cognition evolved as a successful strategy. In the ~300 million years of separate evolution, birds and mammals developed different brains to facilitate this. In the extant species, there are striking parallels such as the presence of highly comparable structures like the PFC and NCL, but there are also stark differences, such as incredibly high neuron densities in birds. The aim of this thesis is to investigate these two points more closely to get a better understanding of how the birdbrain accomplishes these remarkable cognitive feats.

19

It is currently not known if an executive structure such as the NCL exists in reptiles. The presence of such an area in a reptile will inform when exactly cognition emerged in the sauropsid lineage. In Chapter 2 of this thesis I will investigate if a reptilian species, the Nile crocodile, possesses a higher order associative area as shown in birds and mammals. Because of the phylogenetic proximity and cerebral similarity to birds, I expect that if an NCL- like structure is present it to be visible in the caudo-lateral part of the crocodilian nidopallium.

Moreover, even though the NCL is better researched in birds than in reptiles, so far it has been identified and delineated in the pigeon only. This is problematic because we observe a wide range of variation in both behaviour and cerebral lay-out across different avian species. In mammals, the PFC varies both in size and number of parcellation depending on the species of analysis. Thus, it is highly likely that a similar diversity can be observed across the avian lineage. In order to investigate whether the NCL is present as a uniform structure or requires a species-specific approach, I will analyse the patterns of dopaminergic innervation in four different bird taxa in Chapter 3 of this thesis. I expect a variation of the NCL both in size as well as in number of subdivisions that could be explained by the behavioural capacities of the species.

Lastly, it has been suggested that one reason the avian brain generates such remarkable behaviour is due to high neuronal densities. This raises the question, how are birds able to sustain all these neurons? Across mammals it was shown that there seems to be a fixed neuronal energy budget, and that the total brain consumption is merely determined by the absolute numbers of neurons. If the same hold true for birds, it means that the pigeon brain consumes twice the amount of glucose compared to a rat of a similar brain and body size. Thus, Chapter 4 of this thesis will investigate whether birds have a similar neuronal energy budget as mammals. Since nothing is known about the cerebral energy consumption of birds, this will be an exploratory study.

20

21

22

Chapter 2 Dopaminergic innervation reveals a possible primordial NCL-like structure in the Nile crocodile (Crocodylus

niloticus)

23

24

This study was part of the recently published “Nuclear organization and morphology of catecholaminergic neurons and certain pallial terminal networks in the brain of the Nile crocodile, Crocodylus niloticus”, Billings, B.K., Bhagwandin, A., Patzke, N., Ngwena, A., Rook, N., von Eugen, K., Tabrik, S., Güntürkün, O., & Manger, P., in Journal of Comparative Neurochemistry. This chapter is based on a segment of the published article.

2.1 Introduction

As mentioned in the introduction, the best-studied brain structures underlying complex cognition are the prefrontal cortex (PFC) in mammals, and, more recently, the nidopallium caudolaterale (NCL) in birds. It is known from topological (Medina & Reiner, 2000) and genetic (Puelles et al., 2000; Tosches & Laurent, 2019) research that the two structures are not homologous. Namely, the PFC is situated at the frontal pole of the mammalian cortex as a six- layered structure that develops from the embryonic dorsal pallium (Puelles et al., 2000). In contrast, the NCL of birds is located at the caudal pole of the nidopallium, a subfield of the dorsal ventricular ridge (DVR), and originates from the ventral pallium (Güntürkün et al., 2017). Instead, these two structures represent a case of convergent evolution (Güntürkün, 2005, 2012), demonstrated by comparability of the PFC and NCL in function (Güntürkün & Bugnyar, 2016; Mogensen & Divac, 1982; Nieder, 2017), hodology (Kröner & Güntürkün, 1999; Leutgeb, Husband, Riters, Shimizu, & Bingman, 1996), and chemoarchitecture (Herold et al., 2011). This implies that the last common ancestor of birds and mammals did not have a single structure that gave rise to the PFC and NCL in the mammalian and avian lineage respectively, but instead, that there were (at least) two areas in the ancestral brain with the inherent capacity to evolve into such a complex multimodal structure.

25

In mammals, the evolutionary origin of the PFC can be dated to early mammals around 166 million years ago (mya). This is a time-point after the emergence of therapsids following the split from stem amniotes (Kemp, 2006; Rowe, Macrini, & Luo, 2011), and around the divergence of monotremes from marsupial and placental mammals (Bininda-Emonds et al., 2007). Though it is not possible to study the brains of these species, an analysis across all major clades of the mammalian lineage identified a subset of common neocortical features. These early mammals had a small neocortex, with approximately 20 areas of differentiation, among which a medial frontal and lateral orbitofrontal subdivision. This is considered evidence for an early emergence of the PFC, since it is more parsimonious to assume homology instead of convergent evolution of a PFC-like structure in each separate lineage (Northcutt & Kaas, 1995; Kaas, 2011).

In contrast, we know very little about the antecessor of the NCL. This is surprising, because the avian lineage has several more distantly related sister branches that would allow for an extensive comparative analysis. Namely, different to the mammalian lineage of synapsids, birds are not the only living descendants of the sauropsid line. Other extant species include the squamates (lizards and snakes), the tuatara, turtles and crocodiles (Blair Hedges & Poling, 1999; Crawford et al., 2012). Moreover, the split of these different reptilian lineages occurred relatively early in evolutionary history; e.g. the squamates branched off already in Permian (~278 mya), turtles at the Permian-Triassic boundary (~255 mya), and crocodiles in Middle- Late Triassic (~245 mya, Benton et al., 2009; Chiari, Cahais, Galtier, & Delsuc, 2012; Donoghue & Benton, 2007; Mulcahy et al., 2012). In comparison, the sister branch to placental and marsupial mammals are the monotremes, and this split occurred only in Early (~166 mya, Luo, 2007). Thus, if an analysis across the different reptilian branches reveals a primordial NCL-like structure, it can possibly be dated back to the last common ancestor of the sauropsids, 278 mya.

In birds, the NCL can be characterized by a high density of TH-fibres (Waldmann & Güntürkün, 1993) that show as pericellular baskets with multiple axonal swellings around large unstained perikarya (Wynne & Güntürkün, 1995). In the avian DVR, the NCL is situated above a second structure with a high density of catecholaminergic fibres; the arcopallium dorsale (AD). This is an area with a mixture of limbic and premotor features (Herold, Paulitschek, Palomero- Gallagher, Güntürkün, & Zilles, 2018). Thus, the avian NCL can be identified by the density and morphology of catecholaminergic innervation, and by its location directly dorsal to the AD at the caudolateral edge of the DVR. In order to investigate the primordial origin of the NCL, the dopaminergic innervation of the DVR of the Nile crocodile (Crocodylus niloticus) was mapped. As a member of the Crocodylia, it is the closest living relative of birds nested within archosaurs (Benton & Clark, 1988). Moreover, the crocodilian DVR is large for a non-avian

26

reptile, shows clear subdivisions and a highly differentiated cellular pattern comparable to bird DVR (Northcutt, 2013). Lastly, it was found that the rates of change of the crocodilian genome are very low, which implies higher similarity to the last common ancestor of archosaurs (Green et al., 2014).

2.2 Material & Methods

2.2.1 A note on nomenclature

2.2.1.1 Elaboration on the use of ‘reptiles’ The term ‘reptiles’ in its conventional use refers to a paraphyletic group which includes all reptilian branches except for the monophyletic group of birds. However, birds are in fact a sister group to Crocodylia, in a branch known as archosaurs. The other branches of the reptilian tree include turtles, nested with the archosaurs, and the squamates (lizards and snakes) nested with the single extant species tuatara (Benton & Clark, 1988; Crawford et al., 2012, 2015). Thus, the correct term for the collective of turtles, squamates, tuatara and crocodiles, while excluding birds, would be ‘non-avian reptiles’. However, in this paper the term ‘reptiles’ was adopted for ease of reading.

2.2.1.2 Neuroanatomy At the time of writing, the reptilian brain nomenclature is at a turning point. Traditionally, the literature on snakes, lizards, turtles and crocodiles employed a nomenclature different from avian neuroanatomy. Whereas the DVR of birds has been subdivided into mesopallium, nidopallium and arcopallium (and currently numerous subfields within these major divisions), across reptiles there has been a consensus only on anterior DVR (ADVR) and posterior DVR (PDVR, Ulinski, 1983). The exact distinction between ADVR and PDVR was based on cytoarchitectural features, and often the boundaries were difficult to determine, especially across different reptilian species (Senn & Northcutt, 1973). Indeed, it appears that what is labelled PDVR in lizards corresponds to caudal ADVR in snakes and crocodiles (compare Lanuza, 1998 and Clark & Ulinski, 1984; Ulinski, 1978)). This was more recently confirmed with gene molecular markers that showed the similarity of two DVR territories in lizards and crocodiles that were traditionally known as PDVR and caudal ADVR respectively (Briscoe & Ragsdale, 2018a; Desfilis et al., 2018).

Briscoe & Ragsdale (2018a) furthermore demonstrated clear homologies of alligator DVR with different parts of the avian DVR. Specifically, they found that the alligator DVR to consist of a mesopallium, nidopallium and arcopallium, highly comparable in genetic markers to the name- sakes in the avian DVR. These findings are supported by established literature demonstrating

27

similarities in topology and connectivity, and a recent structural MRI analysis could delineate all identified subfields (Billings, Behroozi, et al., 2020). Following Briscoe & Ragsdale (2018a), this thesis will employ part of the avian terminology where an explicit proposal for homology is present.

2.2.2 Immunohistochemistry The histology was executed by the laboratory of Paul Manger at the University of Witwatersrand (Johannesburg, South Africa). Specifically, the brain was immunohistochemically stained against tyrosine hydroxylase (TH). This is the rate-limiting enzyme in the production of dopamine and the standardized method to delineate the NCL in birds (von Eugen, Tabrik, Güntürkün, & Ströckens, 2020; Waldmann & Güntürkün, 1993; Wynne & Güntürkün, 1995). For a detailed description of the staining procedure, see Billings et al. (2020). In short, paraformaldehyde-fixated brain section were stained against TH (mono/poly rabbit anti-TH, AB151, Chemicon Millipore), and visualized with diaminobenzidine enhanced with the avidin/biotin complex. The brain sections were mounted, dehydrated, rehydrated and cover-slipped.

2.2.3 TH-fibre density analysis With a Zeiss Axio Imager M1 Microscope (Carl Zeiss MicroImaging) I took images of the right hemisphere of the entire anterior DVR (ADVR) at 20x magnification. For the automatic fibre analysis I employed a custom written quantification program following Sathyanesan et al. (Sathyanesan, Ogura, & Lin, 2012), see von Eugen et al. (2020). In short, I pre-processed the images in ImageJ (version 1.48) with the FeatureJ: plugin. This is a Hessian-based feature extraction that extracts curvilinear structures. Next, a baseline adjustment in MATLAB (version R2016a) increased the signal to noise ratio, and stained fibres were identified as high- intensity pixels on a grid. I established the threshold for peak detection manually by comparing counts of 6 random squares by researcher and program, and adjusting the threshold until a difference of less than 10% was reached. This resulted in a detailed approximation of TH+ fibre density per 200 µm2. Next, I closely evaluated zones of high innervation for the occurrence of pericellular baskets (Wynne & Güntürkün, 1995). Specifically, I selected 3 slides representative of the anterior, middle and posterior brain, approximately 5600 µm apart, and placed counting squares measuring 200 x 200 µm. On each section, I evaluated 10 regions of interest (ROI) in the area of highest fibre-density and 10 windows in the remaining mesopallium/nidopallium. I also quantified the number of axonal swellings that contacted the unstained cells within a basket. Findings are reported as mean ± standard deviation.

28

2.3 Results

In order to investigate whether crocodiles have an NCL-like structure, I assessed TH- innervation of the entire DVR in the Nile crocodile in combination with a detailed analysis of high-density areas. Here, I quantified pericellular baskets and the number of axonal swellings.

TH+ fibres were evident throughout the DVR, although I did not observe an area with a fibre density profile similar to the pigeon NCL (Waldmann & Güntürkün, 1993); however, I did observe a small zone with a high TH+ fibre density in the dorsal roof of the arcopallium, resembling pigeon AD (Waldmann & Güntürkün, 1993; Wynne & Güntürkün, 1995) (Figure 3A, B, E). This area demonstrated such a dense network of TH+ fibres that the distinct morphological features were obstructed. Within the caudolateral nidopallium dorsal to this area, I located a small area of a slightly higher TH+ fibre density Figure 3D) compared to the adjacent nidopallium (Figure 3C). Here, I observed a curvilinear fibre plexus containing varicosities that progressed in a non-preferential direction. Occasionally, fibres formed pericellular baskets with axons that coiled several times around an unstained perikaryon, contacting the cell with several axonal swellings (Figure 3D, black arrow, Figure 4B).

This region in the caudolateral nidopallium (Ncl, Figure 4A) contained 4.8 ± 1.3 baskets / 200 µm2 with 20.8 ± 7.5 boutons / basket. In the bordering nidopallium at this caudal level of the brain, I counted 0.23 ± 0.67 baskets / 200 µm2 with 20.7 ± 3.2 boutons / basket (Figure 4C). In anterior sections, the remainder of the mesopallium and nidopallium was devoid of baskets, apart from a small area in the ventrolateral nidopallium. This area seemed to be continuous with the Ncl, but exhibited a decreasing fibre and basket density. In the nidopallium approximately 5600 µm rostral to the Ncl, I quantified 1.1 ± 1.1 baskets / 200 µm2 with 18.1 ± 5.9 boutons / basket, and another 5600 µm more anterior, I counted 0.6 ± 0.6 baskets / 200 µm2 contacted by 15.4 ± 4.6 boutons / basket.

29

Figure 3 Catecholaminergic innervation of the region of highest fibre density in the DVR of the Nile crocodile. (A, B) At this caudal level of the nidopallium (Nc, see figurine), I observed different densities of fibres as depicted in the heat map. Cold (blue) colours represent low and warm (red) colours representing high densities. The densities corresponded to several different types of TH+ innervation. Most of the DVR exhibited low to medium fibre densities (C), but a region in the caudolateral nidopallium (Ncl) contained very high fibre densities (D). The Ncl demonstrated a curvilinear fibre plexus, and I observed numerous axonal swellings. Occasionally, the fibres formed characteristic baskets and coiled several times around unstained perikaryal, contacting with boutons-en-passant (marked by black arrows in D). The most fibre-dense area at this caudal section was situated in the dorsal roof of the arcopallium (A) without clearly discernible pericellular baskets (E). Scale bar in E = 500 µm, applies to C – E.

30

Figure 4 Quantification of baskets and boutons-en-passant in the dorsal ventricular ridge (DVR) of the Nile crocodile. (A) The caudolateral nidopallium (Ncl) contained the area of highest fibre density within the DVR. The figurine represents a sagittal view of the crocodile brain depicting the coronal level where the Ncl was located. (B) The fibres occasionally coiled several times around an unstained perikarya contacting with multiple axonal swellings, forming baskets. (C) The Ncl contained 4.8 ± 1.3 baskets / 200 µm2 with 20.8 ± 7.5 boutons / basket. The rest of the nidopallium at this caudal level contained 0.23 ± 0.67 baskets / 200 µm2 contacted by 20.7 ± 3.2 boutons / basket. The graph bars represent means ± standard deviation. Scale bar represents 1000 µm in (A), and 20 µm in (B).

31

2.4 Discussion

To study the primordial structure that could have given rise to the NCL, I investigated the dopaminergic innervation of the crocodile DVR with an immunohistochemical stain against TH. The NCL in birds can be identified both by the distinct dopaminergic profile and by the topological location situated dorsally to the AD in the caudolateral DVR. The NCL has a high TH+ terminal network with distinct pericellular baskets that coil around unstained perikarya while making multiple synaptic contacts (Durstewitz, Kröner, Hemmings, & Güntürkün, 1998; Durstewitz, Kröner, & Güntürkün, 1999; von Eugen et al., 2020; Waldmann & Güntürkün, 1993; Wynne & Güntürkün, 1995).

Across the entire crocodilian DVR, I could not identify a TH+ innervation profile comparably dense as observed in the avian NCL (von Eugen et al., 2020); however, I did identify a zone in the caudolateral part of the nidopallium (Ncl) with a higher fibre density compared to neighbouring areas. I refrain from immediately using the label ‘NCL’, because this is such a well-established term in the field that will immediately evoke assumptions about a certain connectivity and functionality for which we currently do not have sufficient evidence in reptiles (yet). Specifically, the Ncl contained the highest number of sparse pericellular baskets. I use the term ‘sparse’ here to indicate that in the crocodile Ncl, I observed only a small numbers of TH+ fibres that coiled a few times around a large unstained perikaryon (Figure 3D, black arrows). These fibres also displayed multiple swellings possibly representing presynaptic endings (Figure 4B). As in the avian DVR (von Eugen et al., 2020; Wynne & Güntürkün, 1995), I could also identify TH+ baskets in other parts of the DVR, but they were rare and constituted of only few fibres.

An additional feature that is shared by the crocodilian Ncl and the avian NCL, is the presence of a structure with an even higher TH+ fibre density situated immediately ventral to the caudal nidopallium. The fibre plexus is dense, and without discernible baskets. In birds, this area is termed the arcopallium dorsale (AD), and it is implicated in both premotor and limbic functions (Herold et al., 2018). The current analysis identified a small zone of high TH+ fibre density, resembling pigeon AD, in the dorsal roof of the arcopallium. This type of dopaminergic labelling was also demonstrated in a lizard species, where the field was labelled dorsolateral amygdala (DLA, Andreu, Dávila, de la Calle, & Guirado, 1994). Again, I opt for the term arcopallium because of clear homologies of the labelled area identified here plus surrounding ventrocaudal territory, and the DLA plus surrounding ventral areas in lizards, to the avian arcopallium (Briscoe & Ragsdale, 2018a; Desfilis et al., 2018). There are some differences between the crocodilian area and the avian AD; such as that it appears smaller and less differentiated. Moreover, the connectivity of this area is poorly understood in crocodilians, which has

32

complicated one-to-one evaluations of different arcopallial fields between reptiles and birds. The current finding that the dorsal subfield of the reptilian arcopallium shares a dopaminergic innervation pattern with the avian AD strengthens the idea of an arcopallial homology within at least the sauropsids.

Maybe not surprisingly, just like the avian NCL, the putative Ncl in crocodiles is located in the nidopallium. In birds, the nidopallium is known as a heterogenous territory that contains primary and secondary input fields for auditory, visual and somatosensory information (Reiner, Perkel, Bruce, et al., 2004). Across reptiles, this anterior part of the DVR (traditionally ADVR, now nidopallium; Briscoe & Ragsdale, 2018a) was shown to receive unimodal input from different sensory modalities via sensory specific thalamic nuclei. Visual information flows from the nucleus rotundus to lateral DVR (alligator; Pritz, 1975; turtle: Balaban & Ulinski, 1981; lizard: Bruce & Butler, 1984; snake: Ulinski, 1978), auditory input is relayed from the nucleus reuniens to the medial DVR (alligator: Pritz, 1974 turtle: Balaban & Ulinski, 1981; lizard: Bruce & Butler, 1984; snake: Ulinski, 1978), and somatosensory information is projected to central DVR from the nucleus medialis (alligator: Pritz & Stritzel, 1994; turtle: Cosans & Ulinski, 1990; lizard: Bruce & Butler, 1984; snake: Ulinski, 1978). This type of connectivity mirrors the sensory pathways present in birds (Kröner & Güntürkün, 1999), and there are clear homologies at the thalamic level for each sensory-specific nucleus (Pritz, 2016). Moreover, some of the terminal fields in the crocodile nidopallium express similar genetic markers to the nidopallial sensory areas in birds (Briscoe & Ragsdale, 2018a; Desfilis et al., 2018). Specifically, the lateral division corresponds to the entopallium and the medial area to Field L, which in birds are established primary visual and auditory fields respectively (Reiner, Perkel, Bruce, et al., 2004). A recent fMRI study in the Nile crocodile supported these findings where they demonstrated an increased BOLD response in the nucleus rotundus and entopallium following different visual stimuli. Moreover, following simple and complex auditory stimuli, they observed activity in Field L, and even a secondary auditory structure that could correspond to avian nidopallium caudomediale (NCM), a higher order auditory region (Behroozi et al., 2018). Thus, comparable to birds, and even in a similar topological order, the anterior nidopallium of reptiles shows segregated unimodal sensory streams. Interestingly, these segregated streams in birds have recently been demonstrated to be highly comparable to functional columns in the mammalian neocortex (Stacho et al., 2020), and it is possible that the initial lay-out was already in place in a sauropsid ancestor.

As a higher-order associative area, the defining feature of avian NCL is sensory overlap. Indeed, the NCL receives a reciprocal connection from all sensory modalities situated in anterior parts of the forebrain, and a multimodal projection from the posterior dorsolateral thalamic nucleus (Kröner & Güntürkün, 1999). For a long time, it was thought that this type of

33

sensory convergence was absent in the reptilian forebrain (Bruce & Butler, 1984; Lohman & Smeets, 1991; Ulinski, 1983); however, with the advancement of modern tracing techniques, sensory overlap has been shown in both lizard (Andreu, Dávila, Real, & Guirado, 1996; Lanuza, 1998) and turtle (Belekhova & Chkheidze, 1992). In both species, the caudal pole of the DVR in particular receives projections from all anterior unimodal sensory areas described before. Moreover, a multimodal projection from the medial posterior thalamic nucleus to caudal DVR was shown in lizard (Lanuza, 1998), caiman (Pritz & Stritzel, 1992), and turtle (Belekhova & Chkheidze, 1992). These findings highlight the possibility that a type of multimodal convergence does take place in the DVR. Specifically, this overlap has been found in the caudal nidopallium, which overlaps with the current findings of dopaminergic innervation.

The sensory overlap as observed in lizard and turtle, found to be stretching across the entire back of the brain, does extend beyond the smaller area delineated here. This discrepancy could be the consequence of species differences, where differences in brain size and/or cognitive necessities have influenced the location, size and subdivisions of the reptilian Ncl. This was also observed in different avian species (von Eugen et al., 2020; Chapter 3 of this thesis). Alternatively, it is caused by the employed tracing technique, since injections placed in the caudal DVR usually occupied the entire back of the brain (see for example Figure 2C-E in Lanuza, 1998). This problematizes fine delineation of smaller subareas. This open question will only be resolved by detailed hodological studies in a crocodilian brain.

The dopaminergic innervation of the NCL in birds has been shown to play a crucial part in executive functions such as working memory, extinction learning, rule learning, decision making, subjective value and context integration (Diekamp, Kalt, & Güntürkün, 2002; Herold et al., 2011; Kalenscher, Windmann, et al., 2005; Karakuyu, Diekamp, & Güntürkün, 2003; Lengersdorf, Pusch, Güntürkün, & Stüttgen, 2014; Starosta, Bartetzko, Stüttgen, & Güntürkün, 2017; Veit & Nieder, 2013). Is there behavioural evidence that indicates a reptile has the same need for a dopaminergically modulated associative structure? As discussed in the introduction, the research on reptilian cognition in scarce. But let us evaluate the example of reversal learning more closely, which was successfully demonstrated in lizards (Day et al., 1999; Szabo & Whiting, 2020), turtles (Holmes & Bitterman, 1966), and crocodiles (Gossette & Hombach, 1969). As mentioned, subjects are trained on a choice task in which one of two pictures or objects will be rewarded with food. Once a certain criterion is reached, the rewarded contingencies are reversed. To be capable of reversal learning, an organism needs to learn an association between a stimulus and a reward, and following the switch, it needs to update the reward contingencies, and inhibit responding to the first stimulus and instead respond to the second stimulus. This is facilitated by an entire brain circuitry, but there is a crucial role for a dopamine-modulated associative area be able to inhibit and make the decision to switch

34

(Clark, Cools, & Robbins, 2004; Izquierdo et al., 2017). To elaborate, in mammals a lesion in the orbitofrontal subdivision of the PFC impairs the capacity for reversal learning and shows as a perseverance for the previously rewarded stimulus (Butter, 1969; Daum, Schugens, Channon, Polkey, & Gray, 1991; Dias, Robbins, & Roberts, 1996). It has been proposed that the exact role of this structure is to signal outcome expectancies supported by the mesencephalic dopamine system (Schoenbaum, Roesch, Stalnaker, & Takahashi, 2009). Thus, there is a likelihood that the reversal learning capacity, and other related capacities that demand more than simple associative learning in reptiles is mediated by a similar circuit, for which an executive-like structure is crucial.

Clearly, we need additional research on the hodology, chemoarchitecture and function of the crocodilian Ncl before we can draw strong conclusions. Still, the presence of a high TH+ fibre innervation and pericellular baskets in a caudolateral nidopallial field that is situated to an arcopallial area similar to avian AD, present us with interesting parallels to the NCL in birds. These similarities possibly indicate the presence of a primordial NCL that was present at least in the last common ancestor of the archosaurs, 245 million years ago.

35

36

Chapter 3 The dopaminergic innervation of the executive caudal nidopallium; a comparative study of pigeon, chicken, zebra finch and carrion crow

37

38

This study has been published before as “A comparative analysis of the executive caudal nidopallium in pigeon, chicken, zebra finch and carrion crow” (von Eugen, K, Tabrik, S., Güntürkün, O., & Ströckens, F.) in Journal of Comparative Neurology.

3.1 Introduction

Birds and mammals are two of the most cognitively abled classes among modern fauna. The last common ancestor existed in the Carboniferous about 300 million years ago (Benton & Donoghue, 2006), and since then both groups developed differently organized forebrains that are dominated by a six layered cortex in mammals, and a dense nuclear structure in birds (Güntürkün, Stacho, & Ströckens, 2017). Despite these drastically differently structured cerebra, detailed studies revealed a comparable rich behavioural repertoire across the phylogenetic tree of both lineages. This regards basal mental faculties such as perception of biological motion (chicken; Mascalzoni, Regolin, & Vallortigara, 2010; jackdaw; Greggor, McIvor, Clayton, & Thornton, 2018; cats: Blake, 1993; common marmoset: Brown, Kaplan, Rogers, & Vallortigara, 2010), object permanence (pigeon: Zentall & Raley, 2019; parrot: Pepperberg & Funk, 1990; magpie: Pollok, Prior, & Güntürkün, 2000; cat/dog: Triana & Pasnak, 1981; primate: Blois, Novak, & Bond, 1998) and numerical competence (pigeon: Scarf, Hayne, & Colombo, 2011; chicken: Rugani, Vallortigara, Priftis, & Regolin, 2015; Rhesus macaque: Brannon & Terrace, 1998; chimpanzee: Boysen & Berntson, 1989). But also concerns complex capacities such as mental time travel (pigeon: Zentall, Clement, Bhatt, & Allen, 2001; Western scrub jay: Clayton & Dickinson, 1998; rat: Babb & Crystal, 2006; chimpanzee: Martin-Ordas, Haun, Colmenares, & Call, 2010), tool use (New Caledonian crow: Hunt & Gray, 2004; chimpanzee: McGrew, 2004), and even theory of mind (ravens: Bugnyar, Reber, & Buckner, 2016; chimpanzee: Krupenye, Kano, Hirata, Call, & Tomasello, 2016).

39

In mammals, the seat for complex cognition is the prefrontal cortex (PFC), located at the anterior pole of the neocortical frontal lobe (Fuster, 2015; Miller & Cohen, 2001). The functional equivalent to the PFC in birds is known as the nidopallium caudolaterale (NCL) and was first identified in pigeons (Mogensen & Divac, 1982). It is situated at the posterior pole of the avian pallium and shows striking parallels to the PFC in terms of function, neurochemical modulation, and connectivity (Güntürkün, 2005, 2012). Like the PFC (Fuster, 2015; Goldman-Rakic, 1987), the NCL is a higher order associative structure and reciprocally connected to secondary and tertiary areas from all sensory modalities, connects to visceral, limbic and memory-related structures, and projects to basal ganglia and (pre)-motor areas (Kröner & Güntürkün, 1999; Leutgeb et al., 1996; Shanahan et al., 2013).

A key function of an executive structure is to gate, memorize, and manipulate incoming information and subsequently prepare a goal-direction motor action. The crucial neurotransmitter implicated in these processes is dopamine (Ott & Nieder, 2019), and research has demonstrated several parallels in the dopaminergic architecture of the NCL and PFC. Both structures are densely innervated by dopaminergic fibres that arise from the midbrain substantia nigra (SN) and ventral tegmental area (VTA, rat; Fallon & Moore, 1978; Lindvall, Björklund, & Divac, 1978; monkey: Felten & Sladek, 1983; Gaspar, Stepniewska, & Kaas, 1992; pigeon: Kitt & Brauth, 1986; Waldmann & Güntürkün, 1993). The dopaminergic terminal network follows a common theme with small symmetric synapses that occasionally produce synaptic triads (Durstewitz, Kröner, & Güntürkün, 1999; Metzger, Jiang, & Braun, 2002; Schnabel et al., 1997). The fibres predominantly contact dendritic trees and spines rich in D1 receptors, with a relatively lower number of D2 receptors (Durstewitz, Kröner, Hemmings, & Güntürkün, 1998; Herold et al., 2011). Moreover, the reuptake of dopamine is slow and as a consequence it contacts extrasynaptic dopamine receptors situated outside the synaptic cleft via volume transmission mediated by diffusion (rat: Zoli et al., 1998; pigeon: Bast, Diekamp, Thiel, Schwarting, & Güntürkün, 2002).

On a functional level, the NCL and PFC have been implicated in an array of behaviours that involve working memory, self-control and cognitive flexibility - the fundamental concepts of executive functioning (Diamond, 2013; Fuster, 2015; Güntürkün, 2005, 2012; Nieder, 2017). Ablation of these structures with lesions or pharmacological blockage interferes with spatial and non-spatial memory performance (pigeon: Diekamp, Gagliardo, & Güntürkün, 2002; Gagliardo, Bonadonna, & Divac, 1996; Güntürkün, 1997; Lissek & Güntürkün, 2004; Mogensen & Divac, 1982; rat: Wikmark, Divac, & Weiss, 1973; cat: Divac, 1973; monkey: Rosvold & Szwarcbart, 1964; human: Müller & Knight, 2006), learning and memory consolidation (Hartmann & Güntürkün, 1998; Lengersdorf, Marks, Uengoer, Stüttgen, & Güntürkün, 2015; Lengersdorf, Stüttgen, Uengoer, & Güntürkün, 2014; Lissek, Diekamp, &

40

Güntürkün, 2002; Lissek & Güntürkün, 2003, 2005), and choice behaviour (Kalenscher, Diekamp, & Gunturkun, 2003). Moreover, the underlying neural code on both single neuron as well as population level shows a high degree of similarity between the NCL and PFC. That is, a subset of neurons specifically increase firing rate for the duration of the delay period in working memory tasks (pigeon: Diekamp, Kalt, & Güntürkün, 2002; Johnston, Anderson, & Colombo, 2017; Kalenscher, Güntürkün, et al., 2005; Rose & Colombo, 2005; crow: Veit, Hartmann, & Nieder, 2014; rat: Sakurai & Sugimoto, 1986; monkey: Fuster, 1973; Fuster & Alexander, 1971; Miller, Erickson, & Desimone, 1996; Procyk & Goldman-Rakic, 2006). Furthermore, neurons in the NCL and PFC are critically involved in multisensory learning processes (Moll & Nieder, 2015, 2017; Starosta, Stüttgen, & Güntürkün, 2014; Veit, Pidpruzhnykova, & Nieder, 2015), represent reward or value (Dykes, Klarer, Porter, Rose, & Colombo, 2018; Johnston et al., 2017; Kalenscher, Windmann, et al., 2005; Koenen, Millar, & Colombo, 2013; Scarf, Miles, et al., 2011; Starosta, Güntürkün, & Stüttgen, 2013) and encode abstract rules (crow: Veit & Nieder, 2013; monkey: Wallis, Anderson, & Miller, 2001). In conclusion, despite structural differences in their overall morphological layout, the NCL and PFC are the key structure implicated in cognitive control and realize this in a highly similar manner.

Besides the outlined differences and similarities in executive structure between these vertebrate classes, there is a considerable variation within the two lineages. The PFC is not a uniform structure across mammals. In line with the range of variation in behavioural and cognitive capacities, the PFC does not only vary in absolute and relative size, but also differs in the extent of parcellation and number of sub-divisions (Kaas, 2019). For example, the dorsolateral PFC is considered a specialization of the primate branch, and that it cannot be clearly identified in Rodentia (Carlén, 2017; Wise, 2008). As already suggested by Iwan Divac et al (1985): “Given the variation of topography of the PFC in different mammalian species, one might expect some such variability also in different species of birds”. However, in contrast to the PFC, which has been studied in numerous mammalian species (Passingham & Wise, 2012), in birds the NCL has been described in pigeons only. Besides some research on the NCL in chicken (K. Braun et al., 1999) and recent ground-breaking work in the carrion crow (Nieder, 2017), which only concerns electrophysiological recordings and lacks detailed neuroanatomy, we know very little about this structure in other members of the avian lineage. This is a problem, because there are significant differences in both behaviour as well as in neuroanatomy across the avian phylogenetic tree. For example, members of the corvids can readily acquire and transfer abstract rules, whereas pigeons either show no indication of transfer at all (B. Wilson, Mackintosh, & Boakes, 1985b), or require additional training (Wright et al., 2017). There is a similar pattern in self-control. Different members of the corvid family

41

demonstrate the capacity to delay gratification for up to 320 s for a higher quality reward (Dufour, Wascher, Braun, Miller, & Bugnyar, 2012), whereas pigeons will choose the smaller, immediate reward even with a delay of 4s (Green, Fisher, Perlow, & Sherman, 1981).

These behavioural differences are mirrored by differences in brain anatomy, both at a macroscopic and a microscopic level. Namely, bird species that with respect to complex behaviour are on par with non-human primates, such as some members of Passeriformes (songbirds) and Psittacines (parrots), have brains with an expanded mesopallial and nidopallial territory and high neuron densities. In contrast, basal birds, such as Galliformes (chicken) and Columbiformes (pigeon), have a brain with a relatively larger brainstem and a proportional cerebellum and forebrain of low neuron numbers (Iwaniuk & Hurd, 2005; Olkowicz et al., 2016). Moreover, Passeriformes and Psittacines are vocal learners, and thus the cerebral bauplan incorporates unique auditory and vocal nuclei that are absent in non-vocal learners (Chakraborty & Jarvis, 2015; Anton Reiner, Perkel, Mello, & Jarvis, 2004). Another prominent difference, particularly in the caudal pallium, is the relative arrangement of the striatum and arcopallium. In chickens (Puelles, 2007) and pigeons (Karten & Hodos, 1967) the arcopallium is positioned caudal and lateral to the striatum, whereas in Passeriformes (Izawa & Watanabe, 2007; Nixdorf-Bergweiler & Bischof, 2007) it is situated caudal and medial to the striatum. Given the differences in behaviour and pallial morphology between pigeons and other Aves, it is likely the NCL exhibits a similar degree of variance as well.

In order to test whether the NCL is a uniform structure across the avian phylogenetic tree, or whether we are facing a variation as observed in mammals, this study aimed to identify and delineate the borders of the NCL in several species of bird. Included were the scientifically relevant species pigeon, chicken, zebra finch and carrion crow to represent a spectrum of cognitive capacities, and brain and body size. The borders of the NCL were delineated based on an immunohistochemical stain against tyrosine hydroxylase (TH), the rate-limiting enzyme in the production of dopamine. This is a well-established method to visualize the dense dopaminergic input used to identify the NCL. The NCL was defined as the area in the caudal nidopallium with the highest TH+ fibre density combined with the presence of characteristic ‘baskets’. These are TH+ fibres coiled multiple times around a large unstained perikaryon (Waldmann & Güntürkün, 1993; Wynne & Güntürkün, 1995). In addition, we visualized the myelinated reciprocal connection between the NCL and the intermediate arcopallium (AI) with a Gallyas silver impregnation technique. This projection is termed the dorsal arcopallial tract (DA) and was employed here as a second criterion to identify NCL (Kröner & Güntürkün, 1999; Zeier & Karten, 1971). With this approach, the aim of this study was to describe the trajectory of the NCL in pigeon, chicken, zebra finch and carrion crow, and to verify the uniformity or diversity of the NCL across different species of bird.

42

3.2 Material & Methods

3.2.1 Animals and tissue preparation In this study, I analysed two pigeons (Columba livia), six chickens (Gallus gallus domesticus, White Leghorn), five zebra finches (Taeniopygia guttata) and four carrion crows (Corvus corone). All birds were adults, and except for the carrion crows obtained from a registered breeder. All procedures and care were in concordance with the German guidelines for use and care of animals for experimental purposes, and were approved by the ethics committee of the State of North Rhine-Westphalia, Germany, in agreement with the European Communities Council Directive 86/609/EEC regarding the use and care of animals in science. Since pigeon has been the model organism to study the NCL, it functioned here as a control for our analysis method. The other species were selected to represent diversity in phylogeny, variation of brain morphology and differences in cognitive performance. In terms of phylogeny, chickens represent the most basal group in the avian branch. As Galloanserae, they diverged approximately 72 mya from the other Neoaves used in this study. Pigeons, belonging to the Columbaves branch, are Neoaves that separated from the remaining two species in our study 64 mya. They thus represent an intermediate between the basal Galloanserae and the relatively modern zebra finch and carrion crow. These two species belong to the Passeriformes clade that emerged 56 mya (Prum et al., 2015). Regarding cognitive capacities, pigeon and chicken show rather low levels of complex cognitive control (Güntürkün, Ströckens, Scarf, & Colombo, 2017; Marino, 2017), while the carrion crow, being a member of the corvid family, is considered on par with non-human primates (Emery & Clayton, 2004; Güntürkün & Bugnyar, 2016). Unfortunately, since the main focus of research is on their song system, little is known about the cognitive capacities of zebra finches. Probably as a result of species specificities and cognitive capacities (see above), the morphology of the brain differs substantially between the selected species. In comparison to pigeon and chicken (Iwaniuk & Hurd, 2005), Passeriformes (zebra finch and carrion crow) have a brain with a relatively large forebrain. In addition, the Passeriformes’ brain shows a positional shift of the arcopallium, now located medial and caudal to the striatum, whereas in pigeon and chicken it is situated lateral to the striatum (Mello, Kaser, Buckner, Wirthlin, & Lovell, 2019).

Approximately 15 minutes before anaesthesia, pigeon, chicken and zebra finch were injected with 1,000IU heparin (Ratiopharm, Ulm, Germany). Next, animals were deeply anesthetized with an intramuscular injection of 0.45 ml / 100 g body weight 1% pentobarbital- 4.5% chloral hydrate solution (pigeon and zebra finch) or 0.03 ml / 100 g body weight 10% pentobarbital solution (chicken). After, animals were transcardially perfused with saline at body temperature (0.9% NaCl, 40°C) followed by cold 4% paraformaldehyde (PFA) in 0.12M sodium phosphate buffered saline (PBS, pH 7.4, 4°C). After perfusion, brains were dissected, postfixated in 30%

43

sucrose in PFA for 1 (chicken and pigeon) or 12 (zebra finch), and cryoprotected in 30% sucrose in PBS until sunken (approx. 12-48 hours). The carrion crow brains were obtained from a licensed hunter during regular pest control in the Netherlands. After a crow was shot, it was rapidly fetched by a trained hunting dog, and the brain was dissected on the spot within 5-10 minutes and immersed in 4% PFA. The total length of immersion fixation was based on the diffusion speed of formaldehyde in brain tissue (1 mm / hour), followed by an additional fixation period of minimum 24-48 hours to ascertain sufficient cross-linking and covalent bonding (Howat & Wilson, 2014). After 3 days in 4% PFA, the brains were postfixated in 30% sucrose in PFA for an additional 2 days. Next, they were transferred to 30% sucrose in PBS for 24-48 hours for cryoprotection. Brains were cut in the frontal plane into 10 series of 40µm sections with a microtome (Leica Microsystems, Wetzlar, Germany), and stored in 0.1% sodium azide in PBS at 4°C.

3.2.2 Immunohistochemistry

3.2.2.1 Tyrosine Hydroxylase (TH) Preliminary experiments revealed that in the two Passeriformes species antibody binding was rather low when compared to pigeons and chickens. Therefore, brain slices of zebra finches and carrion crows were pre-treated with heat-induced epitope retrieval (HIER, Jiao et al., 1999). The low levels of antibody binding are a possible consequence of prolonged fixation times, in which the epitope of interest is masked by a higher number of cross-links. Heating the tissue cleaves these bridges and unmasks the antigen binding site (Yamashita, 2007). In order to verify this procedure did not generate any false-positives, we performed a control of the HIER-procedure in immersion-fixated pigeon tissue. This resulted in a highly comparable labelling pattern to perfused untreated brain tissue. For the HIER procedure, sections were washed in 0.12M PBS and incubated in preheated 10mM sodium citrate buffer (pH 9.0) at 80°C for 20min. The final immunohistochemical protocol was modelled after the work of Dmitry Kobylkov from the University of Oldenburg. For all species, immunohistochemical labelling of TH was performed in one series of free-floating brain sections and conducted at room temperature unless stated otherwise. All washing steps consisted of three 10 min rinses in tris- buffered saline (TBS, pH 7.6) on a slow 5° rotator. After an initial rinse, sections were incubated in 0.6% H2O2 in distilled water to deactivate endogenous peroxidases. The sections were rinsed, and placed in 10% Normal Goat Serum (NGS) in 0.3% Triton X-100 in TBS to block nonspecific binding. After another washing step, overnight incubation in the primary antibody (polyclonal rabbit anti-TH, AB152, Merck Millipore, Darmstadt, Germany), 1:500 in 1% NGS in TBST took place at 4°C on a slow 7° rotator. The next day, the sections were rinsed and placed in the secondary antibody (polyclonal goat-anti-rabbit, Vectastain Elite ABC Kit, Rabbit IgG, Vectorlabs, Burlingame, USA) 1:1000 in TBST for 2 hours. Next, the sections were washed in

44

TBST and incubated in solution of the avidin/biotin complex (Vectastain Elite ABC Kit) 1:100 in TBST for 1 hour. The complex was visualized with a nickel enhanced 3,3’-diaminobenzidine (DAB) reaction, using the commercially available kit (DAB Peroxidase Substrate Kit, Vectastain, Vectorlabs, Burlingame, USA). Lastly, the sections were mounted on gelatine- covered slides, dried for 30 min, dehydrated in an alcohol series, rehydrated with xylene and cover slipped with DePex (Sigma-Aldrich, Darmstadt, Germany). The specificity of the antibody was verified with a Western blot.

3.2.2.2 Gallyas silver impregnation The second criterion to delineate the borders of the NCL is the myelinated reciprocal connection to the AI. This projection is also known as the dorsal arcopallial tract (DA, Kröner & Güntürkün, 1999) and the strong myelination of the tract can be visualized with the Gallyas silver impregnation technique. The silver stain makes use of the natural argyrophilic characteristic of myelin to bind silver particles and thereby labels myelinated structures black (Gallyas, 1971; Merker, 1983).

Of each species, one series of brain sections was mounted on gelatine coated slides and dried at 40°C for 1 hour. After, slides were incubated in 70% ethanol for 1 hour and rinsed in distilled water for 3 x 5 min. Next, sections were incubated in an infiltration solution (0,1% silver nitrate, 0,1% ammonium nitrate, 0,012% sodium hydroxide solution in distilled water) for 1 hour, and consecutively rinsed in 0.5% acetic acid for 3 x 10 min in the dark. Following, slides were incubated in a developing solution (a mixture of the 3:2:1ratio of A) 5% sodium carbonate, B) 0,2% silver nitrate, 0,2% ammonium nitrate, 1% silicotungstic acid, and C) 0,2% silver nitrate, 0,2% ammonium nitrate, 0,4% 37% Formalin in distilled water) for 10min, and rinsed in 0.5% acetic acid for 3 x 10 min followed by one rinse in distilled water of 5 min. To enhance contrast, sections were incubated in 1% gold chloride solution for 5 min and washed in distilled water for 4 x 10 min. Sections were fixed in 1% sodium thiosulfate and rinsed in distilled water for 5 min. Lastly, the slides were dried, dehydrated in an alcohol series, rehydrated in xylene and cover slipped with DePex.

3.2.3 Data acquisition Pictures of the caudal forebrain were taken at 20x magnification of a Zeiss Axio Imager M1 microscope (Carl Zeiss MicroImaging, Göttingen, Germany) for each species and stain. The range of images included the posterior lateral striatum up to the caudal back of the forebrain, including the entire arcopallium. Previous description of the trajectory of the pigeon NCL (Waldmann & Güntürkün, 1993; Wynne & Güntürkün, 1995) were guiding for this selection. This was considered a conservative topological approach, and translated into an analysis of the caudal telencephalon up to the disappearance of the medial striatum. As described by

45

previous studies (Kröner & Güntürkün, 1999; Waldmann & Güntürkün, 1993; Wynne & Güntürkün, 1995), the stereotaxic coordinates for the NCL in the pigeon ranged from A7.00 to A3.75 (Karten et al., 1967), and finch from A1.35 to P0.36 in the zebra (Nixdorf-Bergweiler & Bischof, 2007). There are currently no brain atlases available for adult chicken and carrion crow. To guide identification of different areas we employed the brain atlas of a chick (LPuelles, 2007) and of the Japanese jungle crow (Izawa & Watanabe, 2007). We labelled the analysed sections from anterior to posterior with ascending Roman numbers to facilitate referencing in the result section.

3.2.4 TH-fibre density estimation I employed a custom made automatic counting program that identifies fibre-like structures from images of a high magnification overlaid with a Hessian filter for the quantification of TH-fibre densities. The program is based on work from Sathyanesan, Ogura, & Lin (2012) and employs the Hessian-based curvilinear feature extraction of ImageJ (version 1.48, U.S. National Institutes of Health, Bethesda, MD, USA). The filtered images were processed and analysed with a custom made program in MATLAB (version R2016a, MathWorks, Natwick, MA, USA), written by Sepideh Tabrik. First, the signal to noise ratio is increased with a baseline adjustment, executed with the ‘msbackadj’ command of the bioinformatics toolbox from MATLAB. This function utilizes an Expectation-Maximization algorithm and groups the data points into “peak” (e.g. stained fibres) or “background”. Both groups follow a normal distribution. The average value of the “background” group generates the final baseline. Next, utilizing the ‘mspeaks’ function from the MATLAB bioinformatics toolbox, the number of fibres were quantified from high-intensity pixels (the stained fibres) above a certain threshold on a projected gridline. The grid overlaid the entire brain slice and measured squares of 100 µm x 100 µm. The threshold was established by manual quantification of fibres in parallel by an experienced researcher in three brain slices per series. To this end, the program randomly generated four squares of 100 µm x 100 µm, and the threshold was adjusted to obtain a difference of less than 10% between automatic and manual fibre quantification. After the threshold was determined, the program generated heat maps of detailed fibre counts per 100 µm2 over the whole brain slice.

3.2.5 Close-up analysis The second criterion for identification of the NCL, is the presence of distinctive TH+ baskets. Guided by the heat maps, the areas of highest fibre density were closely analysed in a qualitative manner. This entailed a verification of the presence of TH+ baskets, and assessment of basket morphology differences. This included the size of the innervated unstained cell body, and the strength of innervation (e.g. how many fibres). Moreover, the morphology of the TH+ fibres was inspected with respect to the directionality of the fibres

46

(disperse or linear), and presence of varicosities. The third criterion to verify the location of the NCL concerned the trajectory of the DA tract that runs between the intermediate arcopallium and the NCL in pigeon (Kröner & Güntürkün, 1999). Observations were made of both the morphology of the myelinated tracts and the progression of fibres. These three parameters, fibre network, basket morphology and trajectory of the DA, aided in separating the different areas of the caudal pallium. It should be noted the areas were primarily subdivided based on TH+ fibre density, and the main aim of the close up analysis was to validate the presence of baskets and further explore possible subdivisions of the NCL.

3.3 Results

3.3.1 TH+ fibre innervation In all four species, the general staining pattern of TH immunohistochemistry was in line with previous reports. I observed strongly labelled cells in mesencephalic VTA and SN (Figure 5), representing the well documented dopaminergic neurons of the mesocortical and nigrostriatal dopamine system respectively (Smeets & González, 2000). Both the LSt and AD in the caudal telencephalon were densely innervated by fibres and showed as almost uniformly labelled fields. As has been described before (Mello et al., 2019), in pigeon and chicken the AD emerged lateral to the LSt, whereas in the zebra finch and carrion crow the AD was appeared medial to the LSt,. In contrast, the auditory thalamo-recipient Field L, known to receive no dopaminergic input at all, was void of any labelled fibres (Kröner & Güntürkün, 1999; Waldmann & Güntürkün, 1993; Wynne & Güntürkün, 1995). Within the caudal nidopallium, I could identify several fields of higher TH+ fibre density in each of the four species. These were further differentiated based on absolute fibre density, fibre network organization and basket morphology if applicable.

47

Figure 5 TH labelling in frontal sections of the mesencephalon (one hemisphere, shown without forebrain) of pigeon (a), chicken (b), carrion crow (c), and zebra finch (d). All species showed a highly comparable labelling with TH+ soma, axons and neuropil in the well-documented dopaminergic cell groups SN and VTA. Stereotaxic coordinates are shown only for pigeon (A3.25) and zebra finch (A1.71), since there is currently no brain atlas available for adult chicken and carrion crow. Scale bar in (a)-(d) depicts 1000 µm.

48

3.3.1.1 Pigeon The field of highest caudal nidopallial TH+-fibre innervation appeared ventro-lateral directly adjacent to the lateral ventricle around A7.00. Moving posterior, this area extended along the dorso-medial and ventro-lateral borders of the nidopallium in a semi-lunar shape. From A5.50 on, the ventral border stretched further into the surrounding central nidopallium and the fibre innervation in the whole area intensified (Figure 6). The observed pattern of innervation was in concordance with previous studies of the pigeon NCL (Durstewitz et al., 1999; Kröner & Güntürkün, 1999; Waldmann & Güntürkün, 1993; Wynne & Güntürkün, 1995). In addition, I corroborated previous descriptions of a medial and a lateral subdivision of the NCL based on fibre density differences (Herold et al., 2011; Waldmann & Güntürkün, 1993). Medial NCL was visible as a smaller area of lower fibre density from A7.00 – A4.50. Lateral NCL curved around the dorsal edge of the medial subdivision, separated by a thin band of low TH+ innervation.

The innervation network morphology of both NCL subdivisions (Figure 7b, c) appeared as curvilinear (as opposed to straight) fibres. The fibre network did not have a clear direction of progression, but was primarily dispersed. Between A7.00 – A5.00, a subset of fibres did demonstrate a clear directionality and progressed in parallel to the dorsal border of the nidopallium. This stream of fibres became less pronounced from A5.00 onwards, and transformed into a disperse network. As already observed in the heat map, the whole of the NCL showed a much denser fibre network compared to surrounding caudal nidopallium (NC, Figure 7d). Within the NCL, I identified numerous baskets that consisted of multiple TH+ fibres coiled densely around an unlabelled perikaryon. These coiling fibres most likely form multiple axonal contacts with the some and probably the initial dendrites (Durstewitz et al., 1998, 1999). I observed chiefly an equal distribution of singular baskets, and sometimes a conglomerations of 2-4 baskets in close proximity. I also identified multiple en-passant contacts, visible as an axonal swelling in a part of a fibre not coiled around a cell body. These possibly synapsed on other unstained structures (Durstewitz et al., 1999). In concordance with previous reports (Waldmann & Güntürkün, 1993), the fibre density in lateral NCL (Figure 7b) was of much higher density compared to medial NCL (Figure 7c), and the higher number of baskets I observed received a more dense innervation.

49

Figure 6 description on next page

50

Figure 6 Series of exemplary heat maps representing TH+ fibres densities of one pigeon (left ), and schematic outlines of areas based on all analysed pigeons (right) in the caudal telencephalon (A7.00 – A4.00, Karten & Hodos, 1967). The heat maps were based on an immunohistochemical staining of frontal brain sections against TH, where TH+ fibre densities were quantified in 150µm2 squares with a custom-written software. High TH+ fibre innervation was observed in LSt and AD, whereas Field L was void of TH+ labelling. Based on a relative higher TH+ fibre density and a qualitative assessment of morphology of fibres and baskets (compare Figure 7), I was able to delineate the trajectory of the previously defined pigeon NCL including two distinct subdivisions (lat NCL and med NCL) within the caudal nidopallium. The dorsolateral corticoid area (CDL) and Hippocampal complex (Hp) were lost during the immunohistochemical process between A5.50-A4.00. Scale bar depicts 1000 µm.

Figure 7 Morphology of TH+ positive fibres and baskets of different structures in the caudal nidopallium in pigeon. (a) Overview of a frontal brain section at A5.50 (Karten & Hodos, 1967) stained against TH. (b)-(d) Close- up of the insets shown in (a). The pigeon NCL (b)+(c) showed a much higher TH+ fibre and basket density compared to surrounding nidopallium (d). The NCL has two subdivisions, where lateral NCL (lat NCL, (b)) showed a stronger TH+ innervation in terms of fibres and baskets than the medial one (med NCL, (c)). In both subareas, numerous singular baskets (bold arrows) were visible, that would occasionally conglomerate (double arrow). Furthermore, the baskets were contacted by axonal boutons from stained coiling fibres (unfilled arrows heads), and these varicosities were also observed in passing fibres, possibly indicating en-passant contacts (filled arrow heads). (d) The surrounding nidopallium showed low levels of TH+ fibre innervation, and only few sparsely innervated baskets. Scale bar in (a) depicts 1000 µm, scale bar in (d) 50 µm (representative for (b)-(d)).

51

3.3.1.2 Chicken In chicken I identified two separate fields of high TH+ fibre density (Figure 8). The first was visible as a small patch at the border between nidopallium and AD at section III. Moving caudal, this patch expanded below the lateral ventricle dorso-medially and intensified in fibre density reaching its maximum size and highest density at section VI. At this level, the area was visible as a semilunar shaped band located ventro-medial to the lateral ventricle and stretching approximately to the centre of the slide. Posterior to section VI, the band decreased in size and fibre innervation and vanished just before section VIII. The trajectory of this field is in close correspondence to the lateral NCL in pigeon. We therefore labelled this area NCL. I could not detect any subdivisions based on TH+ fibre innervation.

The second area of high fibre density first appeared at section III, but as a circular field in the dorsal section of the central caudal nidopallium. Moving posterior, it shaped as a sphere-like structure with an increasing fibre density and diameter. The maximum in both density and size was reached at section VI, where it almost extended up to NCL. Here, it appeared as if the dorso-medial tip of the NCL arched around the dorsal border of this central patch. Following the trajectory caudal, the labelled area decreased in size and remained in the centre of the tapering nidopallium. This second field has a high degree of overlap in its shape and trajectory to an area labelled by Puelles et al. (2007) as island fields of the caudal nidopallium (NCIF). I decided to adopt this nomenclature.

NCL and NCIF differed strongly in TH+ fibre network morphology. The fibre innervation of putative chicken NCL (Figure 9b) was highly similar to the NCL in pigeon. The network of NCIF (Figure 9c) was less dense, more dispersed, and with a lower number of more strongly innervated baskets when compared to NCL. In addition, the baskets formed more conglomerations of spots of multiple strongly interconnected baskets. These spots appear to correspond to what Puelles et al. (2007) labelled ‘island fields’. As was visible in the heat map, both putative NCL and NCIF had higher fibre densities compared to the surrounding NC (Figure 9d).

Figure 8 Series of exemplary heat maps representing TH+ fibre densities of one chicken (left side) and schematic outlines of areas based on all analysed chickens (right side) in the caudal telencephalon in the frontal plane (Puelles et al., 2007). Since there is currently no brain atlas available for adult chicken the ascending Roman numbers represent the brain sections from anterior to posterior. Data acquisition details are described in the figure captions of figure 6 and the main text. Like in pigeons, LSt and AD showed by far the highest innervation of TH+ fibres, while field L was empty. There are two areas of higher TH+ fibre innervation within the caudal nidopallium. The first area was highly comparable in trajectory and fibre/basket morphology to the pigeon NCL (compare Figure 6Figure 7) and was thus labelled NCL. The second field, situated ventral medial to NCL fit the description of previously identified NCIF and was thus labelled as such. CDL and Hp were lost in the staining process between section III – VIII. Scale bar depicts 1000 µm.

52

Figure 8 description on previous page

53

Figure 9 Morphology TH+ fibres and baskets in different structures of the caudal nidopallium in chicken. The Roman number refers to a brain section number of figure 4. (a) Overview of a frontal brain slice at section VI stained against TH. (b)-(d) Close-up of the insets shown in (a). TH+ fibre innervation in NCL (b) was higher compared to NCIF (c) and both were more densely innervated in comparison to surrounding NC (d). Fibres coiled around unstained soma formed numerous baskets (bold arrows) that were equally distributed over the NCL, with sporadic conglomerations of 2-3 baskets (double arrows). Varicosities were visible contacting the baskets (unfilled arrowheads), as well as in passing fibres possibly making en-passant types of contact (filled arrowheads). Compared to NCL, NCIF (c) TH+ fibre density was lower, the fibres were more disperse, and the area contained fewer but more pronounced baskets (bold arrows), more often organized in conglomerations of strongly interconnected baskets (double arrows). NCIF showed axonal swellings at baskets (unfilled arrowheads) and in progressing fibres (filled arrowheads). While the surrounding NC (d) contained TH+ fibres and baskets, density was much lower and basket morphology differed from NCIF and NCL. Scale bar in (a) depicts 1000 µm, scale bar in (d) 50 µm (representative for (b)-(d)).

54

3.3.1.3 Carrion crow Overall, TH+ fibre innervation of the caudal nidopallium in the carrion crow differed significantly in pattern, extent and diversification when compared to pigeon and chicken, but was highly similar to the zebra finch (see below). I could identify four distinct fields of high TH+ fibre innervation based on density, and fibre and basket morphology (Figure 10). The first spanned almost the entire field between AD and the dorsal border, situated in the medial corner of the caudal nidopallium. It extended from section I to the caudal border of the forebrain, and stretched between 1 - 3mm lateral. This area has a strong overlap with a territory labelled the caudal medial nidopallium (NCM), and thus I adopted the term. A second field of high TH+ fibre innervation showed as a 1-2mm wide band directly adjacent to the dorsal border of AD. It was first visible at section I and stretched from NCM across the whole length of the arcopallium into central nidopallium. Moving caudal, the area closely followed the arch of AD and expanded ventral and lateral. At section IV – section V it appeared to stretch ventrally beyond the ventral tip of the arcopallium. Since there are no previous description of this field, I labelled it medial part of the NCL (NCLm). The third area of higher TH+ fibre innervation first showed at section III as a small circular field at the dorso-lateral border of the nidopallium. Going caudal, it expanded into a band in parallel to the arch of the dorsal border of the nidopallium. The band reached its maximum extent at section VII, where it stretched medially towards NCM and ventro-lateral up to the ventral tip of AD. At later sections, the medial border retracted away from NCM and the entire area disappeared before section XII. I named this field the dorsal part of the NCL (NCLd). The fourth area that showed a high TH+ fibre innervation first appeared at section VIII and extended up to the caudal end of the forebrain. It stretched as a 2-3 mm wide band from the lateral to medial border of the ventral tip of the nidopallium. At more caudal sections, the band narrowed and curved, aligning in parallel to the ventro-lateral nidopallial border, leaving a small stripe of low fibre density at the ventro-lateral tip of nidopallium. After section XII, the area seemed to amalgamate with NCLm. I labelled this area as the ventral part of the NCL (NCLv).

All four subareas contained a dense TH+ fibre network with many baskets and axonal boutons both in passing fibres as well as in baskets. I observed variation in the details of the morphological structure (Figure 11). In NCM (Figure 11b) the fibre network was dispersed, and the baskets predominantly appeared singular. In contrast, in NCLm (Figure 11c) a subset of the fibres showed directionality and traversed in parallel to the lamina arcopallialis dorsalis (LAD), and occasionally I could identify interconnected conglomerations of baskets. These were so densely innervated by labeled fibres that they were visible as tiny black dots when observed with the naked eye. Another distinct feature of this area was that almost all fibres displayed axonal swellings along their whole length possibly representing many en-passant

55

types of contact. The TH+ fibre network in NCLd (Figure 11d) was mostly dispersed and showed a high degree of curvilinearity (i.e. as opposed to traversing straight, the fibres made many curves) when compared to NCM, NCLd and NCLm. Furthermore, the baskets were less densely innervated compared to the other areas. The NCLv (Figure 11e) was characterized by a relatively low TH+ fibre density compared to the other described areas and the fibre network was rather dispersed. The baskets predominantly conglomerated with 3-5 interconnected baskets made up by many TH+ fibres. Across all described areas, the innervation pattern was always higher than surrounding NC (Figure 11f), and became more intense moving from anterior to posterior.

Figure 10 Series of exemplary heat maps representing TH+ fibres of one carrion crow (left side) and schematic outlines of areas based on all analysed crows (right side) in the caudal telencephalon in the frontal plane (Izawa & Watanabe 2007). Since there is no brain atlas available for carrion crow ascending Roman numbers were employed to designate the brain section from anterior to posterior. Data acquisition details are described in the figure captions of figure 6 and the main text. In carrion crows, the TH+ innervation pattern differed significantly from pigeons and chickens. While again LSt and AD showed the highest innervation and field L was void of fibres (not shown), TH+ innervation in the rest of the caudal nidopallium was much more varied. In total, there were four distinct NCL-like areas with a high TH+ fibre density compared to surrounding NC. They were labeled in accordance with their established nomenclature and topology: NCM, NCLd, NCLm and NCLv. All four subareas had a distinct fibre network and basket morphology (compare Figure 11). The CDL and Hp were lost in the staining process between section IV – section XIII. Scale bar depicts 1000 µm.

56

Figure 10 description on previous page

57

Figure 11 description on next page

58

Figure 11 Morphology of TH+ fibres and baskets in the different NCL-like subareas of the caudal nidopallium in carrion crow. The Roman number refers to a brain section number of figure 6. (a) Overview of a frontal brain slice of the carrion crow at section X stained against TH. (b)-(f) Close-up of the insets shown in (a). TH+ fibre density in the four identified areas NCM (b), NCLm (c), NCLd (d), and NCLv (e) was much higher compared to adjacent NC (f). All subareas contain baskets of coiled fibres around an unstained soma (bold arrows), and varicosities at baskets (unfilled arrowheads) as well as in passing fibres (filled arrowheads). The baskets in NCM (b) had an equal distribution and only rarely conglomerated with other baskets. This was different to NCLm (c), where the baskets were much densely innervated, and thus more pronounced, and occasionally conglomerated (double arrow). The baskets of NCLd (d) also formed conglomerations, but TH+ fibre innervation was much lower compared to NCLm. Whereas the overall fibre density of NCLv (e) was lower, the baskets were very pronounced and often formed conglomerations. Scale bar in (a) depicts 1000 µm, scale bar in (d) 50 µm (representative for (b)-(d)).

3.3.1.4 Zebra finch The TH+ fibre distribution of the caudal nidopallium in the zebra finch was highly comparable to the observations in the carrion crow. Likewise, I could identify the four areas NCM, NCLm, NCLd, and NCLv (Figure 12), but with an additional rostral and caudal subdivision of NCLd based on TH+ fibre innervation and projection of the DA.

NCM first showed at A1.08, situated caudo-ventral to Field L in between the dorsal roof of the nidopallium and the arcopallium. The area was visible up to the caudal back of the forebrain and never stretched beyond 1 mm laterally from the medial nidopallial border. NCLm was visible as a narrow band that arched in parallel to the dorsal arcopallium. It expanded from A1.08 to the caudal end of the forebrain. At P0.09 it stretched beyond the ventral tip of the AD and appeared to fuse with NCLv (see below). NCLd was located in the dorso-lateral roof of the nidopallium and demonstrated a rostral and caudal subdivision that were separated by a field of lower fibre density. Rostral NCLd (NCLdr) stretched as a wide band caudo-lateral to Field L in parallel to the dorsal roof of the nidopallium from A1.35 – A0.45. The area was not situated immediately adjacent to the lateral ventricle, but instead separated by a thin band low in TH+ fibres. The caudal subdivision of NCLd (NCLdc) was a smaller area first visible at A0.18 and expanded up to the caudal end of the forebrain. It was situated in dorsal rood, halfway the medial-lateral plane, and moving caudal, it stretched towards the medial border. NCLv first appeared at A0.45, showing as a diffuse patch of TH+ fibres next to the ventral tip of the arcopallium. After A0.18, the patch shaped into a thin band in parallel to the ventral nidopallial border.

The morphological profile of each subarea consisted of a dense fibre network with baskets, and axonal swellings were visible both on passing fibres as well as within baskets. I observed minor differences in basket and fibre morphology (Figure 13). The basket and fibre profile of NCM (Figure 13c) in the zebra finch was highly similar to carrion crow NCM. The fibres were primarily dispersed and I observed mostly singular baskets. Compared to the other high density

59

nidopallial areas, it showed the highest density of fibres, especially in its caudo-medial range. The innervation network of NCLm (Figure 13d) was also similar the NCLm in the carrion crow, with a subset of fibres traversing in parallel to the LAD. The baskets were made up of many TH+ fibres and especially at posterior sections conglomerated with 2-4 baskets. The fibre network of NCLdr (Figure 13e) had a dispersed character, and the fibres seemed wrapped more loosely around unstained cell bodies giving the baskets a less pronounced appearance compared to the baskets of other NCL-subareas. They were equally distributed over NCLdr and made only few conglomerations. NCLdc (Figure 13f) had an innervation profile comparable to carrion crow NCLd, with as distinctive feature a high degree of curvilinearity and a dispersed fibre network. In addition, the baskets appeared less pronounced and I observed only few conglomerations. The fibre network of the dorsal half of NCLv (Figure 13g) was predominantly disperse, but in the ventral half the fibres curved in parallel to the ventral tip of the caudal nidopallium. The baskets were densely innervated by TH+ fibres and primarily in singular arrangements. As was observed in the heatmaps, all NCL-subareas had higher TH+ fibres densities compared to surrounding NC (Figure 13h).

Figure 12 Series of exemplary heat maps representing TH+ fibre densities of one zebra finch (left side) and schematic outlines of areas based on all analysed zebra finches (right side) in the caudal telencephalon in the frontal plane (A1.35-P0.36, Nixdorf-Bergweiler & Bischof, 2007). Data acquisition details are described in the figure captions of figure 6 and the main text. Distribution of TH+ fibres in zebra finches was highly similar to carrion crow (compare figure 10), and different from pigeons and chickens (compare figure 7,8). Like in the other species, LSt and AD showed the highest TH fibre density while field L was void of TH+ fibres. As in the carrion crow, the NCL- like subareas NCM, NCLm, NCLd, and NCLv showed a higher TH+ density than the surrounding NC and the subdivisions differed in fibre network and basket morphology. In contrast to the carrion crow, NCLd in zebra finches was subdivided into a rostral and caudal aspect, labeled NCLdr and NCLdc, respectively, separated by an area of low TH+ fibre density. The CDL and Hp were lost in the staining process. Scale bar depicts 1000 µm.

60

Figure 12 description on previous page

61

Figure 13 description on next page

62

Figure 13 Morphology of TH+ positive fibres and baskets in different structures in the caudal nidopallium in zebra finches. (a) Overview of a frontal brain section at A0.70 (Nixdorf-Bergweiler & Bischof, 2007) stained against TH. (c, e, g) Close-up of the insets shown in (a). (b) Overview of a more caudal brain section A0.18 stained against TH. (d, e, h) Close- up of the insets shown in (b). The four identified NCL-like subareas showed a higher TH + fibre density compared to the surrounding caudal nidopallium (h), contained baskets (bold arrows), as well as clearly visible axonal boutons at baskets (filled arrowheads) as well as on progressing fibres (unfilled arrowheads). TH+ fibre density in NCM (c) was higher compared to other subareas and showed evenly distributed, mostly singular baskets. As in carrion crow, the baskets in zebra finch NCLm (d) formed occasional conglomerations (double arrow) and were much more densely innervated in comparison to the other subareas. The high fibre density in the bottom right corner stems from the adjacent AD. The rostral (e) and caudal (f) subdivision of NCLd were similar in appearance characterized by a lower basket count compared to the other areas. The baskets were predominantly singular, and less pronounced in comparison to NCM and especially NCLm. This was in contrast with NCLv (g), which showed densely innervated baskets that occasionally formed conglomerations. The shown pictures are from a female zebra finch, hence song related nuclei (e.g. RA) are less pronounced compared to males. Scale bar in (a) and (b) depict 1000 µm, scale bar in (h) 50 µm (representative for (c)-(h)).

3.3.2 Myelinated trajectory of the dorsal arcopallial tract In all species analysed, the DA stretched between the arcopallium and the surrounding caudal nidopallium (Figure 14). In pigeons (Figure 14a), it extended between AI and NCL, and arched in parallel to the lateral ventricle to the dorsal nidopallial roof. The DA showed a high degree of overlap with the NCL, as delineated by TH+ fibre innervation. In the chicken (Figure 14b), it appeared more broad stretching slightly more medial, thereby encompassing a larger territory than what I designated putative NCL. As in pigeons, it arched dorsally in parallel to the lateral ventricle, but did not expand further than halfway the slice in the medial-lateral plane. The tract seemed to arch around NCIF, thereby not targeting it directly. The DA in the two Passeriformes species showed a significantly different pattern to pigeon and chicken, and was in concordance with the shifted location of the arcopallium and NCL in the carrion crow and zebra finch (Figure 14c-f). Specifically, the arcopallium was situated midway the medial-lateral axis instead of at the lateral-most border. The DA extended almost across the entire arcopallium in the frontal plane, and projected to the NCLd of the carrion crow (Figure 14c) and NCLdr of zebra finches (Figure 14e). At more posterior sections, DA arched towards the ventral tip of the arcopallium (Figure 14e, f). In the zebra finch it targeted the NCLdc progressing past NCLv. This latter projection was not observed in the carrion crow, and in this species, DA appeared to only target the NCLd.

Figure 14 Representative frontal brain slide of the caudal nidopallium stained with the Gallyas silver impregnation technique that labels myelinated fibres black. In both pigeon (a) and chicken (b), the DA progresses along the dorsal lateral nidopallial border, and connects the arcopallium with the area labeled as NCL. Note that in chicken, the DA does not target the NCIF directly, but instead arches around it. In Passeriformes, the arcopallium shifted medial and caudal (see main text), and the DA in carrion crow ((c), (e)) and zebra finch ((d), (f)) shifted in concordance (see main text). It connect the arcopallium with the NCL-like subareas defined here as NCLd (NCLdr and NCLdc in zebra finch). Scale bar depicts 1000 µm.

63

Figure 14 description on previous page

64

Figure 15 Close-up of DA in all species, visualized with the Gallyas silver impregnation technique that labels myelinated structures black. In pigeon (a) and chicken (b), the DA runs in between the arcopallium and NCL. Pigeon DA comprises of thin, mostly singular fibres, whereas in chicken, the thin fibres are intermixed with thicker processes likely representing fibre bundles. In the carrion crow (c) and zebra finch (d, e), DA runs in between NCLd and the arcopallium, passing through NC and NCLm. Similar to chicken, the fibre plexus in the two Passeriformes are a mix of thin singular fibres and thicker fibre bundles. In the zebra finch, the projection targeting NCLdr consisted more of fibre bundles (d), whereas DA targeting NCLdc comprised of more singular fibres (e). Scale bar in (c) depicts 1000 µm (representative for (a)-(c)). Scale bar in (e) depicts 500 µm (representative for (d), (e)).

65

The myelinated fibre morphology of DA differed between species (Figure 15). In the pigeon (Figure 15a), DA comprised almost entirely of thin, likely singular, fibres. I observed the same in chicken (Figure 15b), with additionally some thicker fibre bundles visible closer to the lateral border. In the carrion crows (Figure 15c), the DA consisted of a mix of both fibre types at more rostral sections, and singular fibres dominated when moving posterior. This could also be observed in the zebra finch, with NCLdr and NCLdc being targeted by thick fibre bundles and thin singular fibres, respectively (Figure 15d, e).

3.4 Discussion

The present study investigated the trajectory of the nidopallium caudolaterale (NCL) in pigeon, chicken, carrion crow and zebra finch based on density and innervation pattern of tyrosine hydroxylase (TH) positive fibres. Though many orders of the avian branch are not represented in this comparative analysis and thus is far from exhaustive, the study is a sound first step to explore the possible diversity of the NCL. Indeed, this research showed that the trajectory of the NCL requires a species-specific approach, in particular for the Passeriformes. In short, based on the TH+ fibre network, the putative NCL in chicken is highly comparable to pigeons, with the addition of one area known as island fields of the caudal nidopallium (NCIF, Puelles, 2007). In contrast, the two Passeriformes species show a strikingly different pattern. In both carrion crow and zebra finch, I identified four separate areas of a distinctively high TH+ fibre density that span across the entire caudal nidopallium. Next to the caudo-medial nidopallium (NCM), none of the three additional areas have been described before and based on their topography, I labeled them dorsal (NCLd), medial (NCLm), and ventral NCL (NCLv).

Corroborated by a large body of research, the NCL is regarded the avian analogue to the mammalian prefrontal cortex. Anatomically, this is supported by three lines of evidence: 1) a strong dopaminergic innervation from mesencephalic VTA/SN, 2) input from the secondary sensory areas of each modality, and 3) a downstream projection to premotor and motor structures. In concordance, as a dopamine-modulated area integrating sensory input and motor output, the NCL is involved in complex mental faculties such as reward coding, rule learning and working memory (Güntürkün & Bugnyar, 2016). In order to elucidate what comprises the NCL in different species of bird, I will discuss whether each area as delineated by high dopaminergic innervation receives multimodal sensory input, sends projections to (pre-)motor areas, and is involved in complex cognitive capacities.

66

3.4.1 Pigeon In accordance with previous delineations (Herold et al., 2011; Kröner & Güntürkün, 1999; Waldmann & Güntürkün, 1993; Wynne & Güntürkün, 1995), the pigeon NCL is best described as a semi-lunar structure located in the dorsal lateral roof of the caudal nidopallium. My analysis uncovered a lateral and medial subdivision of the NCL, separated by a narrow band of low fibre and basket density. This finding corroborates a previous subdivision based on differences in receptor density (Herold et al., 2011). The data is thus in full concordance with previous descriptions of the NCL, and this proves the reliability and effectiveness of the method of analysis.

It has been well established that pigeon NCL, and the lateral subdivision in particular, receives input from all sensory modalities. The input is arranged in a dorsal medial to ventral lateral manner of auditory, visual (tectofugal and thalamofugal), somatosensory, and trigeminal. The direct pathway for each modality to reach the NCL flows via a common sequence with a sensory specific thalamic nucleus that projects onto the primary pallial recipient that relays information to a adjacent secondary area, which projects onto the NCL (Kröner & Güntürkün, 1999; Leutgeb et al., 1996). In short, the primary pallial area of auditory information is field L2 that projects to the flanking fields L1 and L3, which innervate the dorsal medial most part of the NCL (Wild, Karten, & Frost, 1993). Curiously, this section of the NCL is the only part that receives an additional direct thalamic input (Kröner & Güntürkün, 1999). This projection arises from the shell region of the nucleus ovoidalis (Ov) and is thought to constitute a parallel auditory projection. The central part of the NCL is overlappingly innervated by two visual streams and a somatosensory pathway (Kröner & Güntürkün, 1999; Leutgeb et al., 1996). The thalamofugal visual pathway first projects to the caudal lateral densocellular (HD) and interstitial part of the hyperpallium apicale (IHA, Hodos, Karten, & Bonbright, 1973), which then targets the caudal hyperpallium apicale (HA, Shimizu, Cox, & Karten, 1995) that relays to the central NCL. In the tectofugal visual pathway, the primary pallial recipient is the entopallium (Benowitz & Karten, 1976), which projects to the entopallial belt (M. Watanabe, Ito, & Ikushima, 1985a) that transfers information to central NCL. Somatosensory input first arrives at the rostral aspects of HD and IHA (Funke, 1989), is then relayed to rostral parts of HA (Wild, 1987) from where it projects to central NCL. The primary pallial field of the trigeminal pathway is the nucleus basalis (Bas, Schall, Güntürkün, & Delius, 1986), which targets the frontal nidopallium (NF, Wild, Arends, & Zeigler, 1985), where the input is transferred to the ventral lateral part of NCL (Kröner & Güntürkün, 1999; Leutgeb et al., 1996).

Besides being a main recipient of multimodal sensory information, the NCL is heavily implicated in motor functions. The medial subdivision of the NCL primarily gives rise to efferents to the sensorimotor division of the lateral and medial striatum (Kröner & Güntürkün,

67

1999; Veenman, Wild, & Reiner, 1995), while the lateral NCL is reciprocally connected to anterior, dorsal, and intermediate arcopallium (AA, AD, AI, Kröner & Güntürkün, 1999) that are considered to be somatomotor (Herold et al., 2018; Zeier & Karten, 1971). Furthermore, the projection between NCL and AI in part relays information from the contralateral hemisphere via the anterior commissure (AC, Letzner, Simon, & Güntürkün, 2016). The DA runs between the AI (Zeier & Karten, 1971) and NCL, and, as demonstrated here, DA and NCL display a high degree of overlap, corroborating previous descriptions (Kröner & Güntürkün, 1999). In conclusion, the NCL of the pigeon is densely innervated by dopaminergic fibres that form distinctive baskets. It is in addition an integration area that constitutes the of multimodal sensory input and descending motor output (Kröner & Güntürkün, 1999; Leutgeb et al., 1996; Shanahan et al., 2013).

3.4.2 Chicken The study found that the trajectory of the putative NCL in chicken is very similar to the NCL in pigeons. Likewise, it is located in the dorsal lateral roof of the caudal nidopallium. I identified a second area of high TH+ fibre innervation situated ventral medial to the putative NCL, this area is known as island fields of the caudal nidopallium (NCIF, Puelles, 2007). In concordance with these findings, in chicken, high densities of TH+ fibres in the dorsal lateral caudal nidopallium have been described before (Schnabel et al., 1997), though others reported a homogenous distribution of TH+ fibres in the caudal nidopallium, with the exception of an empty field L (Martin Metzger, Jiang, Wang, & Braun, 1996; Moons, van Gils, Ghijsels, & Vandesande, 1994). These different findings could be the result of variations in the used immunohistochemical staining protocol, or there is the possibility the computerized quantification program was able to identify even subtle innervation pattern differences.

Similar to pigeon, the putative chicken NCL is a multimodal integration centre that is structured in a similar dorsal lateral to ventral medial fashion of sensory modalities (Martin Metzger, Jiang, & Braun, 1998; Martin Metzger et al., 1996). The dorsal medial NCL is the main projection field of auditory input relayed from field L1/L3 and the shell region of Ov (Wang, Zorio, & Karten, 2017). The central part of the putative NCL is the main target of two visual streams from the Ep and the caudal division of HA, and somatosensory information is transferred via the rostral part of HA. The ventral lateral area of the putative NCL receives trigeminal information from NF (Metzger et al., 1998, 1996). Different from pigeons, there seems to be a less apparent overlap of sensory information. It is plausible that in the chicken NCL the sensory modalities are more segregated. Alternatively, the sensory overlap would only be fully revealed when anterograde tracers would be placed in the secondary sensory areas, as has been performed in pigeons (Kröner & Güntürkün, 1999).

68

The downstream projections of the dorsal lateral caudal nidopallium in chicken are highly comparable to pigeon. The putative NCL projects to the basal ganglia, and is reciprocally and homotopically connected to AD, AId, and AIv (Metzger et al., 1998, 1996). Antero- and retrograde pathway tracing into NCL also revealed an interhemispheric connection to the contralateral arcopallium (Metzger et al., 1998, 1996). As in pigeon, AI is the principal terminal field of DA (Davies, Csillag, Székely, & Kabai, 1997), and this study revealed a high degree of overlap between the projection field of DA and NCL. The tract appears to not target NCIF, but instead to arch around the area to connect with the medial subdivision of NCL. One significant difference between pigeon and chicken DA concerns the structural morphology of the myelinated processes. In pigeon, the DA only comprises of thin fibres, whereas in chicken the thin fibre plexus is intermixed with thick processes of different lengths, most probably constituting bundles of fibres.

Little is known about the connectivity pattern of NCIF. Redies et al. (2001) introduced the term island fields to designate groups of cells with an expression profile of a specific type of cadherin, surrounded by a nidopallial matrix expressing cadherins of a different subtypes (Redies et al., 2002). Moreover, these fields have a similar neural birth date different from the surrounding nidopallium (Heyers, Kovjanic, & Redies, 2003; G F Striedter & Keefer, 2000). And, as corroborated by this study, are densely innervated by TH+ fibres (Puelles, 2007). When comparing with pigeons, the topology of NCIF corresponds to the caudo-central nidopallium (NCC). The NCC receives a strong projection from the dorsal intermediate mesopallium (dMI), sends output fibres to AI and AM, and has an elaborate intrinsic circuitry. Because of this connectivity pattern, it has been suggested to be associated to the limbic system and to be a key player in autonomic and neuroendocrine functions (Atoji & Wild, 2009). As the NCIF, the NCC in pigeons is targeted by a dopaminergic innervation (Waldmann & Güntürkün, 1993; Wynne & Güntürkün, 1995), but to a much lesser extent.

In conclusion, based on topology and the highly comparable circuitry of sensory input and motor output, I propose that the area of high dopaminergic innervation in the dorsal lateral roof of the caudal nidopallium is the NCL in chicken. There is currently not enough evidence to include or exclude NCIF as part of NCL.

3.4.3 Carrion crow and zebra finch As mentioned, compared to pigeon and chicken, the carrion crow and zebra finch show a significantly different dopaminergic innervation pattern of the caudal nidopallium. Both species belong to the oscine branch of the Passeriformes order, and their last common ancestor existed around 28 mya (Prum et al., 2015). Though their brains differ significantly in size, they are highly similar in architecture and organization (Izawa & Watanabe, 2007; Nixdorf-

69

Bergweiler & Bischof, 2007). This comparability was confirmed by the analysis of the TH+ innervation pattern. Therefore, I will discuss these species jointly. In both Passeriformes, I could identify four distinct NCL-like subareas. NCM is a well-defined auditory area (Maney & Pinaud, 2011) situated at the caudal medial border of the nidopallium. Immediately adjacent, NCLd stretches laterally along the dorsal nidopallial roof and parallels the arch and extent of the arcopallium. The zebra finch seemed to have a rostral (NCLdr) and a caudal (NCLdc) subdivision of NCLd that were not observed in the carrion crow. The NCLm in both species is located dorsally adjacent to the arcopallium, separated only by the LAD. Lastly, the NCLv is visible in the ventral lateral tip of the caudal nidopallium.

In the carrion crow, the findings corroborate previous descriptions of the pattern of dopaminergic innervation of the caudal nidopallium of the house crow (Corvus splendens); a different crow species (Sen, Parishar, Pundir, Reiner, & Iyengar, 2019). However, the interpretation of subdivisions and nomenclature made here diverge from the conclusions of Sen et al. (2019). What is considered NCM here, Sen et al. (2019) termed DNC. The current labeled NCLd compares to their mNCL, and what the label NCLv corresponds to their lNCL. NCLm is not explicitly described by Sen et al. (2019), but the area can be delineated from their stained slices (compare Figure 7 in Sen et al., 2019). There are two reasons for this change in nomenclature presented here. Firstly, it is important to promote consistency and thus to adhere to the existing songbird literature, which is the case for NCM (Maney & Pinaud, 2011). Secondly, due to the addition of NCLm, the proposed topological denominations of Sen et al. (2019) do not hold anymore, and thus needed to be revised. There is the possibility that the analysis and interpretation here differ due to general physiological differences between the two crow species. The carrion crow is significantly larger in body (470 gr.) and brain size (8.5 gr.) compared to the house crow (295 gr., 5.7 gr.; Jønsson, Fabre, & Irestedt, 2012), which could translate into differences in the neuroanatomy. Previous descriptions of the dopaminergic innervation of the caudal nidopallium of zebra finches do not refer to any of the subareas identified here, except for a narrow band along the medial border of the caudal nidopallium (Bottjer, 1993). This region would correspond to NCM, for which the strong dopaminergic innervation is been well-documented in other Passeriformes (canary (Serinus canaria): Appeltants, Ball, & Balthazart, 2001; White-throated sparrow (Zonotrichia albicollis): LeBlanc, Goode, MacDougall-Shackleton, & Maney, 2007; Matragrano, Sanford, Salvante, Sockman, & Maney, 2011).

The caudal nidopallium in the zebra finch receives input from all sensory modalities that largely overlaps with the delineated NCL-subareas here, but also projects to the surrounding caudal nidopallium. Similar to pigeon and chicken, the sensory modalities are partially overlapping and organized in a comparable dorsal medial to ventral lateral manner of auditory,

70

thalamofugal and tectofugal visual, somatosensory, and trigeminal input, respectively. That is, the dorsal medial NCM plays a pronounced part in auditory processing. It can be subdivided in a rostral and a caudal subdivision based on its connectivity pattern. Rostral NCM exclusively receives a projection from field L2 and input from the thalamic shell region Ov, whereas caudal NCM is predominantly targeted by field L1/3 (Mello, Vates, Okuhata, & Nottebohm, 1998; Vates, Broome, Mello, & Nottebohm, 1996). Situated medial to NCM, NCLd is targeted by the thalamofugal and tectofugal visual stream relayed by the caudal part of HA and Ep respectively (Sadananda, Korte, & Bischof, 2007), while somatosensory input to NCLd is transferred via the rostral part of HA (Wild & Williams, 1999). NCLv is the main projection field for trigeminal input from NF (Wild & Farabaugh, 1996). There is currently no data available on connectivity of the NCLm.

Downstream projections onto the arcopallium arise from different parts of the caudal nidopallium. NCM sends efferents to the medial dorsal and medial ventral part of AI (Mandelblat-Cerf, Las, Denisenko, & Fee, 2014) that in part overlaps with the RA-cup, which projects to subtelencephalic auditory nuclei (Mello et al., 1998). The area of the caudal nidopallium that includes NCLd targets AD and AIv in a homotopic manner (Mandelblat-Cerf et al., 2014; Paterson & Bottjer, 2017). In the ventral lateral tip, NCLv sends efferents onto the lateral section of AD and AIv. AIv relays the input via the anterior commissure to contralateral AV, which in turn projects homotopically onto the ipsilateral caudal nidopallium (Paterson & Bottjer, 2017). Similar to pigeon and chicken, AD and AI in songbirds are most probably key structures for general motor generation (Bottjer, Brady, & Cribbs, 2000; Dugas-Ford, Rowell, & Ragsdale, 2012b; Feenders et al., 2008; Mandelblat-Cerf & Fee, 2014; Stetner & Fee, 2017).

As mentioned before, compared to pigeon and chicken, the arcopallium of the Passeriformes appears to have shifted medially within the caudal pallium. The results presented here show that the DA moved in concordance with the arcopallium, and primarily connects to NCLd in both zebra finch and carrion crow. Similar to chicken, the DA of the zebra finch has two distinct types of innervation. The predominant type that project to NCLdr consists of bundles of fibres that emerge from the full medial lateral extent of the arcopallium. Moving caudal, the tract extends to the ventral arcopallial tip, with an increase of singular myelinated processes in the ventral most section in particular. I observed a comparable intermixed plexus of thick and thin myelinated fibres in the DA of the carrion crow.

In conclusion, the densely TH+ innervated subareas in the zebra finch and carrion crow receive a sensory input from almost all modalities and send efferent projections to sensorimotor related structures. I therefore propose that at least NCM, NCLd and NCLv comprise the NCL of the

71

Passeriformes examined in this study. There is currently not enough hodological data on NCLm to decide whether this is a possible fourth subdivision of the Passeriformes NCL or not.

3.4.4 Functions of the NCL modulated by dopamine The well-established method of delineation of the NCL is based on a strong dopaminergic innervation that arises from mesencephalic VTA and SN (Divac et al., 1985; Waldmann & Güntürkün, 1993; Wynne & Güntürkün, 1995). This approach originates from research in mammals, in which the dense dopaminergic innervation is considered a defining feature of the PFC (Güntürkün, 2005). The dopamine system is highly conserved across vertebrates (Smeets & González, 2000), and likewise, the NCL and PFC have a comparable dopaminergic architecture (Puig, Rose, Schmidt, & Freund, 2014). A wealth of literature supports the idea that dopamine is of crucial importance for the facilitation of executive functions, and goal- directed behaviour in particular (Ott & Nieder, 2019). A large body of research has demonstrated the executive involvement of the NCL as defined in the current study in pigeon, chicken, zebra finch and carrion crow.

The best-studied phenomenon in pigeons is working memory, conceptualized as the capability to hold and manipulate relevant information that is no longer perceptually available (Baddeley & Hitch, 1974; Diamond, 2013; Miller & Cohen, 2001). In fact, the first functional evidence supporting the analogy of the NCL and PFC was presented by the group of Ivan Divac who demonstrated that ablation of the NCL impaired choice alternation between two pecking-keys with an introduced delay in pigeons. Importantly, visual discrimination remained unimpaired (Mogensen & Divac, 1982). The crucial role of the NCL in working memory was furthermore substantiated with single unit recordings that revealed increased firing rates specifically linked to the delay phase of a working memory task (Diekamp et al., 2002; Johnston et al., 2017; Kalenscher, Windmann, et al., 2005; Rose & Colombo, 2005). Furthermore, this capacity is strongly dependent on dopamine, as demonstrated by an increase of dopamine levels in the NCL specifically during the delay phase (Karakuyu et al., 2003).

In chicken, the NCL was shown to be involved in imprinting. This is a robust and important learning mechanism that occurs during a critical period in the early life of a chick and facilitates social attachment (Bateson, 1966). While some brain areas are selectively active following either auditory of visual imprinting stimuli, the centre part of the NCL shows an increase in metabolic activity only following the simultaneous presentation of both (Bock, Schnabel, & Braun, 1997). This can be seen as a confirmation of the associative character of the chicken NCL, and refers to a possible role in memory recall and certain emotional processes (Braun et al., 1999).

72

Zebra finches are a model organism to study the avian song system, and as a results, data on their cognitive capacities is scarce. In the songbird literature, NCM is characterized as a higher order auditory area that is silent during the presentation of simple tones, but selectively involved in behaviourally relevant songs (Pinaud & Terleph, 2008). In the NCM, dopamine plays a key role in the modulation of the incentive salience of a particular song; the infusion of a dopamine antagonist or agonists in the NCM of female zebra finches decreased or increased their preference for that song, respectively (Barr, Wall, & Woolley, 2019). Functional studies on other parts of the caudal nidopallium demonstrate that the NCLdc shows an increase in metabolic activity while being chased around the cage (Sadananda et al., 2007) and following the first courtship display (Bischof & Herrmann, 1988). According to the authors, this implies that this area, termed NCLdc in the current study, is involved in the regulation of arousal.

In the carrion crow, most of what is known comes from single unit recordings from an area that corresponds to the rostral lateral parts of NCLv (Veit & Nieder, 2013). Based on a set of experiments, NCLv appears to be primarily involved in prospective coding of visual and multimodal stimuli during a delay period. Specifically, this translates into an array of capacities in which neurons in the rostral lateral NCLv encode visual or spatial working memory (Rinnert, Kirschhock, & Nieder, 2019; Veit et al., 2014), cross modal associations (Moll & Nieder, 2015, 2017), and abstract rules (Veit & Nieder, 2013). As an interesting side-note, whereas this area does specifically encode different visual stimuli and associations across sensory modalities, they did not identify auditory selective neurons (Moll & Nieder, 2015). To summarize, in concordance with the body of evidence for the NCL in pigeons (Güntürkün, 2005, 2012; Güntürkün & Bugnyar, 2016), the initial findings indicate the involvement of the NCL subareas as delineated in the current analysis, in complex associative mechanisms facilitating goal- directed behaviour.

3.4.5 An evolutionary perspective of the caudal nidopallium In concordance with previous descriptions (Mello et al., 2019), the findings presented here demonstrate a significantly different organization of the caudal telencephalon of songbirds when compared to pigeons and chickens. It has been proposed that the reorganization occurred within the songbird branch after the emergence of the Passeriformes 56 mya (Mello et al., 2019; Prum et al., 2015). Two hypotheses try to explain how the caudal forebrain exactly shifted. The rotational axis hypothesis (Mello et al., 2019) postulates that the medial lateral axis rotated into an anterior posterior axis. This is thought to be an explanation for why the arcopallium in the Passeriformes is arranged medial and caudal to the striatum, whereas in pigeon and chicken it is located mostly lateral to the striatum (Mello et al., 2019). The second

73

hypothesis proposes that the presence of additional nidopallium laterally adjacent to the arcopallium, make the arcopallium in Passeriformes appear more medial (Wang et al., 2015). A distinctive feature of the Passeriformes when compared to other birds, except for the Psittaciformes, is indeed an expanded meso- and nidopallial territory (Iwaniuk & Hurd, 2005; Mehlhorn, Hunt, Gray, Rehkämper, et al., 2010), which perhaps has played a role in the arcopallial shift. To continue on the rotational axis hypothesis, there is the possibility that in Passeriformes the entire ancestral NCL-like area rotated such that it now extends across the entire back of the forebrain. This idea does not reject the nidopallial expansion hypothesis, since this rotation alone does not account for the expanded protrusion of NCLv. Therefore, it might be that a combination of both mechanisms have been at play. What does become clear is that this reorganization has had a significant impact on the topography of the caudal nidopallium with an expected effect on the trajectory of the NCL in the Passeriformes analysed here (Figure 16).

Figure 16 description see next page

74

Figure 16 Schematic representation of the NCL in pigeon (a), chicken (b), carrion crow (c), and zebra finch (d). The NCL in pigeon and chicken is best described as a semi-lunar structure in the dorso-lateral part of the caudal nidopallium. The NCL in carrion crow and zebra finch consists of at least three subareas (NCM, NCLd, and NCLv) that span across the entire back of the caudal forebrain. NCLd in zebra finch consists of a rostral and caudal subdivision, of which only the latter is depicted here. There is currently not enough data on NCIF in chicken, and NCLm in carrion crow and zebra finch to include or exclude them as part of the NCL. Scale bar in (a)-(d) depicts 1000 µm. For abbreviations, see list.

As mentioned in the introduction, the PFC is not a uniform structure across mammals. Compared to mouse and rat, the primate lineage can be characterized by an expanded and more gyrified frontal lobe that appears to be more parcellated (Passingham & Wise, 2012), with an extended and intensified dopaminergic innervation pattern (Berger, Gaspar, & Verney, 1991). In addition, the PFC of monkeys and apes have distinctive granular areas that rodents seem to lack completely (Brodmann, 1909). These observations are the starting point for an unresolved debate whether primates are equipped with unique prefrontal territories, or whether they evolved from existing areas already present in early mammals. The latter scenario would imply that all subdivisions of the PFC are shared between primates and rodents, but in an expanded or condensed form (Preuss, 1995; Uylings, Groenewegen, & Kolb, 2003). The findings of this study show striking parallels to the mammalian research. In contrast to pigeon and chicken, the Passeriformes have an expanded meso- and nidopallium (Iwaniuk & Hurd, 2005; Mehlhorn, Hunt, Gray, Rehkämper, et al., 2010). Moreover, the expanded caudal nidopallium is, at least in zebra finch and carrion crow, more diversely and densely innervated by dopaminergic fibres. Concurrently, the NCL of both the zebra finch and carrion crow comprise of at least three parcellated subareas that span from medial to lateral across the caudal nidopallium. It is also important to note the parallels in cognitive capacities, since corvids are thought behaviourally on par with chimpanzees (Emery & Clayton, 2004; Güntürkün & Bugnyar, 2016). In conclusion, this study discloses yet another instance of the remarkable convergent evolution of the executive structure in birds and mammals.

75

76

Chapter 4 Neurons in the avian brain consume three times less glucose compared to mammals

77

78

4.1 Introduction

Neurons make up one of the most energetically costly tissues of the mammalian body, utilizing predominantly glucose as energy substrate (Dienel, 2019). Even at rest, the average neuron consumes approximately 5.79 * 10-9 µmol glucose per minute. Across mammals, these high costs have been shown to be relatively invariant across species, and in concordance it is the absolute number of neurons that drives the high metabolic demands of mammalian brains (Herculano-Houzel, 2011a; Hyder, Rothman, & Bennett, 2013). For example, the brain of a rat makes up 0.55% of the body weight and with a total of 189 million neurons it consumes 4.6% of total energy intake. In humans, even though the brain makes up only 2% of total body weight, the 86 billion neurons jointly consume 20% of the whole body energy budget (Herculano- Houzel, Catania, Manger, & Kaas, 2015; Mink, Blumenschine, & Adams, 1981b).

It is currently unclear if this fixed neuronal budget also applies to other vertebrate clades. The avian lineage diverged from mammals approximately 300 million years ago (Benton & Donoghue, 2006). In terms of neuronal metabolism, akin to mammals the avian brain utilizes glucose as the predominant energy source (Braun & Sweazea, 2008). As mentioned in the general introduction, a distinctive feature of the avian brain is, however, that they have much higher neuron numbers compared to similarly sized mammalian brains (Olkowicz et al., 2016). Pigeons, who have a comparable body and brain size to rats, are equipped with 310 million neurons. This is nearly twice as many neurons as rats who have an absolute number of 189 million neurons. This raises the question, how are birds able to sustain such high numbers of neurons?

The answer might challenge some long-standing ideas on the metabolic limits of brain growth. Namely, in mammals it has been demonstrated that the energetic costs of neurons pose a constraint upon brain scaling. An increase in brain tissue, and especially in the numbers of

79

neurons, either regionally or globally, is a trade-off between the metabolic costs that will be incurred and the benefits obtained from more sensory or information processing capacity (Laughlin, 2001). For example, in the primate lineage we observe that the increase of brain size came at the cost of higher foraging demands to increase of high-quality food intake (Isler & van Schaik, 2009), and a size-reduction of other energetically costly tissues (Aiello & Wheeler, 1995).

It is true that compared to mammals, birds have higher resting metabolic rates per gram of body weight (Clarke & Pörtner, 2010; White, Phillips, & Seymour, 2006), and require on average almost twice the amount of food per day (Nagy, Girard, & Brown, 1999). However, it is unlikely that these additional costs are the sole result of demands of their higher numbers of neurons; on average birds need to sustain a higher body temperature compared to mammals (Brian K. McNab, 1966; Prinzinger, Preßmar, & Schleucher, 1991) and facilitate powered flight, which is one of the most energetically costly forms of locomotion (Schmidt-Nielsen, 1972). Thus, it is more probable the neuronal energy budget in birds deviates from mammals, which would explain how birds have been able to raise the ceiling on neuron numbers.

This study set out to establish the energetic costs of brain tissue in a bird species. Glucose uptake was visualized in the awake and anesthetized pigeon (Columba livia) with positron emission tomography (PET) and fluorodeoxyglucose F-18 (18F-FDG) as radiotracer. Combined with kinetic modelling, it is possible to quantify the exact cerebral metabolic rate of glucose consumption (CMRglc, Alf et al., 2013; Kuntner, 2014). Lastly, we established the neuronal energy budget by combining the found CMRglc with known neuron numbers of the pigeon brain (Olkowicz et al., 2016) and compared our findings to different species of the mammalian clade.

4.2 Material and methods

4.2.1 Animals & housing The research proposal was reviewed and approved by the local animal ethics committee of Nordrhein-Westfalen Germany, and in accordance with the German Animal Welfare law following the recommendations of the EU directive 86/609/EEC. For this study, 15 adult homing pigeons (Columba livia) of unknown sex were obtained from a local breeder. All pigeons were of a normal weight (466.44 ± 37.38 gr). Pigeons were housed individually on a 12/12h light/dark cycle. Food and water was provided ad libitum, except for an overnight fasting to prevent aspirating food under anaesthesia. To facilitate awake scanning, four individuals were implanted with a plastic head pedestal (Behroozi, 2018; De Groof et al., 2013). In short, under deep anaesthesia (ketamine/xylazine 0.12 ml/100g body weight) an incision was made in the

80

skin to lay bare the surface of the skull. Next, a small plastic head block was secured directly on the skull with dental cement, and the skin was sutured. The head block was compatible with a custom-made holding tube to prevent any head motion in the scanner. In addition, one week prior to scanning, pigeons were habituated to the apparatus and procedure in a mock-PET environment with the head restrained for increasing time-lengths of 5, 10, 15, 30, and 45 minutes over 5 consecutive days. This procedure reduces movement to a negligible minimum and eliminated all visible stress indicators (Behroozi et al., 2020).

4.2.2 PET measurements Guided by Heike Endepols and with the assistance of Felix Ströckens, all PET scans were performed in a microPET Focus 220 PET scanner for small animals (CTI-Siemens, U.S.A.), see Figure 17A for experimental setup. Scanning took place in four separate scanning sessions. A session consisted of one acclimatisation day, followed by one or two scanning days. The subject was scanned in either awake or anesthetized state. For the anaesthetized scan, sedation was induced with 3 % isoflurane in 3:7 O2/N2O and maintained at 30 breaths per minute with 1-2 % isoflurane. Body temperature was kept at 39 ºC via a feedback-controlled warm water bed (medres, Germany). For scanning during the awake state, subjects were head fixated in the holding tube. At the start of the scan, subjects received an injection of 18F-FDG (~70 MBq in 500 µl sodium phosphate buffer) either intraperitoneal (i.p.) or intravenous (i.v.) via a brachial vein catheter (Vasofix Safety G24) fixed with tape. Emission data were acquired for 60 min. Venal plasma glucose level was measured after the scan (On-Call GK Dual, ACON Laboratories Inc.). At the end of each scanning day we performed a 10 min transmission scan using a 57Co point source for attenuation correction. In total, we obtained 16 scans from 15 pigeons with 1 pigeon scanned twice (Table 2). After scanning, pigeons were sacrificed and the heads were scanned by Mehdi Behroozi for a structural MRI in a 7-T horizontal-bore small- animal scanner (Bruker BioSpec, 70/30 USR, Germany).

81

Figure 17 Overview of experimental set-up. (A) Schematic representation of awake scanning set-up following intravenous (i.v.) 18F-FDG injection. The pigeon was fixated in a holding tube secured via an implanted headblock. 18F-FDG was injected via a catheter in the brachial vein, and arterial blood was sampled via a catheter in the brachial artery. The first two minutes, blood was sampled automatically with a syringe pump set to a pulling speed of 1 ml/min. and plasma radioactivity was measured online via a LSO/ADP detector unit. (B) Following an i.v. injection, 18F-FDG kinetics were determined with a two-tissue compartment model (top). It assumes two compartments that

18 represent blood and tissue, and 4 kinetic rate constants. The F-FDG first accumulates in the plasma (CP) and is

18 then transferred into the tissue (K1) where it resides as free F-FDG (CF). Next, it will either be transported back

18 into the blood stream (k2), or it will enter the metabolic cycle and be phosphorylated (k3) into metabolized F-FDG

18 18 (CM). From here, it can still be de-phosphorylated (k4) back into free F-FDG. The concentration of F-FDG in the tissue (CT) can be described as the fractional blood volume (vB) multiplied by the concentration in blood (CB, calculated from plasma activity times plasma fraction), added with fractional tissue volume (1 – vB) multiplied by the concentration of free 18F-FDG plus metabolized 18F-FDG. The exact concentrations of free and metabolized 18F-

FDG cannot be determined directly. Instead, final glucose metabolism (CMRglc) can be calculated from established kinetic rate constants, the measured 18F-FDG plasma levels, and the difference in kinetics of 18F-FDG and glucose metabolism as described by the lumped constant (LC).

Table 2 Overview experimental subjects. For this study in total 16 pigeons were scanned, with one pigeon scanned twice (*). There were three experimental groups, depending on route of injection (i.v. or i.p.) and state (awake or anesthetized).

Route of injection Awake Anesthetized Total i.v. 4 6 10 i.p. 0 6* 6 Total 4 12 16

82

4.2.3 Input function The input function was established from plasma values, which is the golden standard method to estimate the CMRglc (Backes et al., 2011). Arterial blood was sampled from the brachial artery via a fixed catheter (Vasofix Safety G24, B. Braun, Germany) inserted just before the start of the scan. The type of sampling was contingent on the route of injection. Following an i.v. injection, plasma radioactivity concentration immediately peaks within seconds. The dynamics during this early period are crucial for modelling and are best captured with a blood pump with an online detector (Alf, Martić-Kehl, Schibli, & Krämer, 2013). Following an i.p. injection, manual sampling at later time-points suffices because the increase in radioactivity concentration is much slower (Salerno et al., 2019). Thus, in the case of an i.v. injection, blood was sampled continuously via the catheter elongated with a heparinised tube during the first 2 minutes and radioactivity was directly measured with a custom-made blood sampler with a pulling rate set at 1 ml/min. (Backes et al., 2011; Breuer, Grazioso, Zhang, Schmand, & Wienhard, 2010). The delay in measurement between PET scanner and online blood detector was corrected for by dividing tube length by pulling velocity. The tube was then cut, and after discarding 200 µl catheter dead volume, discrete blood samples of 100 µl at 2, 5, 10, 20, 30, 45, and 60 min were taken. Following an i.p. injection, discrete samples were taken from the fixed catheter at 1, 5, 10, 15, 20, 25, 30, 35, 40, 45, 50, 55, and 60 min. The manual blood samples were immediately put on ice. This reduces the inner blood transport of 18F-FDG from plasma to red blood cells to negligible rates (Abumrad, Briscoe, Beth, & Whitesell, 1988; Hankin & Stein, 1972). Blood radioactivity measurements were done by Lukas Vieth. Plasma was separated by centrifuging full blood for 12 min at 1500 g in a microcentrifuge (Centrifuge 5415 R, Eppendorf), and 40 µl plasma was used for analysis. Radioactivity was measured both in the full blood and the plasma sample in a CompuGamma CS γ-counter (LKB / Wallac, Australia), and corrected for deadtime and radioactive decay. The ratio of plasma to full blood of the radioactivity rates (counts/min) at each sampling time was calculated. PET scanner, blood sampler, and γ-counter were cross-calibrated.

4.2.4 Image analysis. The emission scans were histogrammed into time frames (6 x 30, 3 x 60, 3 x 120, and 12 x 240 s for scans following an i.v. injection, and 20 x 180 s for scans following an i.p. injection) and Fourier rebinned, images were reconstructed with the 2-dimensional filtered backprojection employing OSEM3D/MAP reconstruction (Qi et al., 1998). The resulting voxel size measured 0.38 x 0.38 x 0.82 mm. Next to the attenuation correction, scans were corrected for deadtime with a global estimate based on running deadtime average and radioactive decay. Images were analysed with the VINCI Software Tool (Vollmar et al 2007). The images were co-registered to a template structural MRI pigeon brain to facilitate comparison.

83

4.2.5 Kinetic modelling The code for the model was written by Heiko Backes, and he supervised the modelling of the current data. Depending on the route of injection, a different kinetic model was applied to

18 determine the rate constants and CMRglc. Following an i.v. injection, the F-FDG kinetics were determined with the two-tissue compartment model (Sokoloff et al., 1977, see Figure 17B) that estimates 4 rate constants (K1, transport from blood to brain; k2, transport from the brain to blood; k3, phosphorylation; k4, dephosphorylation). Following Backes et al. (2011), parametric images of the rate constants were determined using a Powell algorithm with a voxel-wise fitting of

푡 ′ 푡 ′ ( ) ( ) ( ) ′ ( ′) −푟−(푡−푡 ) ′ −푟+(푡−푡 ) 퐶푇 푡 = 푣퐵퐶퐵 푡 + 1 − 푣퐵 (퐴− ∫0 푑푡 퐶푃 푡 푒 − 퐴+ ∫0 푑푡′퐶푃(푡 )푒 ) (Eq. 1) with CT(t) representing tissue (in this case brain) radioactivity concentration, and

퐶퐵(푡) = 휑푡 × 퐶푃(푡) (Eq. 2)

푘 +푘 +푘 1 푟 = 2 3 4 ± √(푘 + 푘 + 푘 )2 − 4푘 푘 (Eq. 3) ± 2 2 2 3 4 2 4

퐾1 퐴± = (푘3 + 푘4 − 푟±) (Eq. 4) 푟−−푟+ where t is time in minutes after injection and vB is the fractional blood volume, fixed to 0.05. This value of 5% is commonly used in rats, and since the blood volume per unit brain weight of was found the same in pigeon and rat (Heisey, 1968), it was employed here also. CB(t) expresses whole-blood radioactivity concentration calculated from plasma radioactivity

18 concentration CP(t) with Eq. 2. The net influx constant (Ki) for F-FDG results from

퐾1×푘3 퐾푖 = (Eq. 5) (푘2+푘3) which translates into CMRglc via

1 퐶푀푅 = 퐾 × × 퐶 (Eq. 6) 푔푙푐 푖 퐿퐶 푝,푔푙푐 where CP,glc is the post-scan venal plasma glucose level, and LC is the lumped constant that describes the difference of glucose versus 18F-FDG metabolism kinetics. Most commonly, a fixed value is taken for LC. It can be determined by model-independent and model-dependent methods, and somewhat variation is present in the mammalian literature (Alf et al., 2014). This variation has been ascribed to methodological differences, or specificities of species or disease state (Cunningham & Cremer, 1981; Krohn, Muzi, & Spence, 2007; Kuwabara, Evans, & Gjedde, 1990). Following Backes et al. (2011), here we employed a model-dependent method

84

that estimates a variable LC. It elaborates on the idea that the different rates that make up the LC are more robust to variation. Namely, the LC describes two separate kinetic processes, transport and phosphorylation, and the difference between 18F-FDG and glucose in these processes can be broken up into: L1=K1,FDG/K1,glc, L2=k2,FDG/k2,glc, and L3=k3,FDG/k3,glc. The values for the separate constants are: L1=1.48, L2=1.48 & L3=0.37, and were found similar in both human and rat (Hasselbalch et al., 1996; Hasselbalch, Madsen, Knudsen, Holm, & Paulson, 1998). The variable LC, can then be calculated from:

퐿3 퐿3 퐾푖 퐿퐶 = 퐿1 ( + (1 − ) ) (Eq. 7) 퐿2 퐿2 퐾1

This approach does assume that the kinetic rate differences of glucose and 18F-FDG for transport (L1, L2) and phosphorylation (L3) are comparable between pigeons, and rats / humans. This assumption is supported by the literature. In the mammalian nervous system, glucose is primarily transported by glucose transporter proteins (GLUT) 1 and 3. Likewise, high levels of mRNA and protein expression of GLUT1 and GLUT3 were found in the brains of different bird species (Braun & Sweazea, 2008; Kono et al., 2005; Sweazea & Braun, 2006; Welch, Allalou, Sehgal, Cheng, & Ashok, 2013), with a high degree of similarity in genetic and amino acid sequence (Welch, Allalou, Sehgal, Cheng, & Ashok, 2013). In the mammalian brain, glucose is phosphorylated to glucose-6-phosphate by hexokinase 1 (Wilson, 2003). A comparison of pigeon, chicken and rat demonstrated similar activity and specificity levels of hexokinase (Kerly & Leaback, 1957), and a sequence analysis with UniProt revealed a 88% similarity between the human and avian hexokinase 1.

The i.p. data were modelled with the Patlak graphical analysis (Patlak & Blasberg, 1985). This approach assumes at least one irreversible compartment, which translates into a negligible activity of dephosphorylation (k4 = 0). This analysis plots the ratio of TAC over arterial activity to the normalized integral of plasma activity

푡 퐶푇(푡) ∫0 퐶푃(푡)푑푡 = 퐾푖 ∗ + 푐 (Eq. 7) 퐶푃(푡) 퐶푃(푡)

Over time, the reversible compartment will reach equilibrium, and thus the ratio stays the same. Any change occurring after this point must be uptake from the irreversible compartment. This point of equilibrium shows as the curve approaching linearity to which a linear function can be fitted. The slope of the fitted line is the net uptake rate Ki. The average LC of 0.745 from the fitting of the i.v. data was taken to calculate CMRglc via Ki/LC * CP.

85

4.2.6 Statistics Statistical analysis was executed with IBM Statistics SPSS 21 (IBM Corp, Released 2012, Armonk, NY). For all tests, a value was considered an outlier when it lied outside of the following range: 3rd or 1st quartile + or – 1.5 * interquartile range. Values are reported as mean ± standard deviation, and results were significant if p < 0.05. Effect sizes were described with the partial eta squared (η2). To verify whether the LC could be jointly averaged from the awake and anesthetized pigeons, the LC values of both groups were assessed with an independent t-test. Both groups met the assumption of normality and homogeneity of variance as tested with a Shapiro-Wilk and Levene’s test respectively. For the LC values, one outlier was detected (case 645 of the i.v. anesthetized group), which was excluded from the analysis. Differences in CMRglc depending on brain region (forebrain or cerebellum) or group (i.v. awake / i.v. anesthetized / i.p. anesthetized) were assessed with a mixed ANOVA. The within-subject variable was repeated measure of “Brain Region”, and post hoc analysis was executed as a rerun of the repeated measures ANOVA with split files depending on group. The variable “Group” was assessed as between-subject variables, and post hoc analysis was run with a Bonferroni correction. Case 536 (of the i.v. anesthetized group) was considered an outlier and excluded from further analysis. All group values were normally distributed as assessed with a Shapiro-Wilk test, met the criterium of homogeneity of variance (Levene’s test), and homogeneity of covariance matrices (Box’s M test). The CMRglc per neuron analysis across species was done by calculating the CMRglc per neuron by dividing CMRglc/mg over neuron numbers/mg. Mammalian data of CMRglc per neuron come from Herculano-Houzel (2011) and sources therein. These data-points were fitted with a linear regression. The neuron numbers on the pigeon brain came from Olkowicz et al. (2016).

4.3 Results

The injected dose of 18F-FDG ranged from 53.9 – 70.4 MBq, with an average of 62.82 ± 5.47 MBq. Post-scan venal glucose levels (14.87 ± 2.42 mmol/L) were within normo-glycaemic range (Scanes, 2015). Raw parametric images (Figure 18) showed that 18F-FDG was taken up and distributed throughout the entire brain. Following i.v. injection, the distribution of average 18F-FDG uptake over 60 min differed between awake (Figure 18A) and anesthetized state (Figure 18B). In the awake state the relatively highest levels of activity were visible in the eye and in the visually related nucleus entopallium (see Figure 18C for schematic). In contrast, in anesthetized state, activity was more diffusely distributed, and next to the eye the relatively highest levels were found in brainstem and cerebellum.

86

Figure 18 Example PET scan. Two exemplary cases of cerebral 18F-FDG activity following an i.v. injection in awake (A) and anesthetized (B) state. Red indicates a high amount of uptake in Bq/ml, and blue represents low levels of uptake. Note that the activity-scale bar on the left differs in magnitude between (A) and (B), as 18F-FDG uptake is higher in awake compared to anesthetized state, see text. The distribution in awake state indicates relatively higher levels in the eye and entopallium, which is the primary sensory area in birds (see (C) for schematic overview of the brain and dotted lines indicate section planes). In contrast, relative 18F-FDG uptake in anesthetized state predominates, besides the eye, in the cerebellum.

To determine the sampled input function, arterial blood was continuously and manually sampled in awake (n=4) and anesthetized (n=6) state following an i.v. injection, and only manually following an i.p. injection in anesthetized state (n=6). The sampled input function of

87

one pigeon following i.v. is shown in Figure 19A (red line) for the awake state, and in Figure 19B (red line) for the anesthetized state. The activity in plasma samples could be characterised by an immediate sharp peak in activity (see insets Figure 19A,B) that would quickly drop and slope off. At approximately t=20 min a plateau was reached after which activity did not decrease further until the end of the scan. An exemplary arterial input function following an i.p. injection in anesthetized state is shown in Figure 19C (red line). In contrast to an i.v. injection, 18F-FDG plasma activity only increased slowly and reached its highest point at approximately t=30 min after which it plateaued until the end of the scan.

The time activity curve (TAC) for forebrain (black triangles) and cerebellum (grey triangles) of one exemplary pigeon per group are shown in Figure 19 A,B, and C for i.v. awake, i.v. anesthetized, and i.p. anesthetized respectively. Following an i.v. injection, the TAC showed the sharpest increase within the first 5 minutes following an i.v. injection, with only a slight increase after that time-point eventually plateauing until the end of the scan. Activity in the two brain regions following an i.p. showed a less steep slope that only appeared to decrease at approximately t=38 min, after which the increase continued until the end of the scan.

Figure 19 Arterial IF and TAC. Representative single cases of the arterial input function (AIF) and time activity curve (TAC) in each of the three groups. Following an i.v. injection in the awake (A) and anesthetized (B) state, the AIF shows in initial sharp peak captured by the automatic sampling device (AIF auto). The peak quickly drops and slopes off to a plateau (AIF manual). The insets depict the initial 5 min. The TAC of the forebrain (black triangles) and cerebellum (grey triangles) rise the steepest in the first 10 min. after which it slopes off into a plateau. (C) Following an i.p. injection in anesthetized state, the AIF has a much slower increase that only reaches a plateau after 20-30 min. A similar less steep slope also applies to the TAC that only seem to plateau towards the end. Note that the scale bars on the left differ per case depending on maximum measured plasma activity.

88

The two different modelling approaches showed a good fit (Figure 20). Both forebrain and cerebellum had a similar time activity curve (Figure 20A). The forebrain displayed slightly higher activity levels at later time-points, which will be discussed below. The residual differences of the two-tissue compartment model (inset Figure 20A) were at almost all points below ± 5%. Except for the first time-point where it showed an underestimation of ~20%. The pattern of difference was highly comparable for the two brain regions. With the Patlak graphical analysis (Figure 20B), it was possible to fit a linear function to all anesthetized pigeons following i.p. injection and estimate the slope (Ki). The curves are highly similar for forebrain and cerebellum, which represents comparability of ratios in the reversible compartments. For this pigeon, the linear function of the forebrain could be described with y = 0.0063x + 0.261, and of the cerebellum with y = 0.007x + 0.2339. This corresponds to a net uptake rate (Ki) of

0.0062 and 0.007 for forebrain and cerebellum respectively. This slightly higher Ki of the cerebellum reflects the relatively higher metabolic rate of the cerebellum in anesthetized state, which will be discussed further below.

Figure 20 Model fits. Representative cases of the two different modelling approaches. (A) The two-tissue compartment model fit of an awake pigeon following i.v. injection shows a good correspondence to both forebrain (black) and cerebellum (grey) kinetics. The inset demonstrates residual differences for both areas. Except for the first time-point (~20% deviation), all fits differ less than ± 5%. (B) The Patlak graphical analysis shows that data of the anesthetized pigeon following i.p. injection approaches linearity. From here the slopes can be determined that represent net uptake rate (Ki). In this case, the forebrain linear fit can be described with 0.0063x + 0.261, and thus Ki = 0.0063. The cerebellum can be described with 0.007x + 0.2339, and thus has a slightly higher Ki of 0.007.

89

Table 3 Lumped constant (LC) per group. The LC described the difference in kinetics of 18F-FDG and glucose metabolism. The outlier 645 (*) of the anesthetized group was excluded. The LC values between awake and anesthetized group did not differ significantly, thus, an overall mean of 0.745 was calculated and used to analyse the i.p. data.

State Pigeon ID LC

Awake 752 0.72534 886 0.73564 920 0.74301 986 0.80556 Anesthetized 146 0.65596 232 0.76294 536 0.6913 590 0.64832 645* 0.97955 877 0.77136 Mean ± SD 0.745 ± 0.0898

The variable lumped constant was estimated in all pigeons of the i.v. awake and anesthetized group (n = 10, Table 3). The LC values of the two groups did not differ t(7) = 1.512, p = 0.174, and averaged 0.745 ± 0.090. Therefore, a LC value of 0.745 was used to estimate the CMRglc values in the i.p. anesthetized group.

Kinetic modelling derived four whole brain kinetic rate constants for both groups that received

18 an i.v. injection (n = 10, Table 4, Table 5). The K1 described the transport of F-FDG from the plasma into the cell, and averaged 0.054 ± 0.0046 for awake and 0.044 ± 0.02 for anesthetized.

The k2 is the rate of transport back into the plasma and that was around 0.14 ± 0.013 for awake 18 and 0.12 ± 0.029 for anesthetized pigeons. The k3 refers to the phosphorylation of F- FDG/glucose with a mean of 0.068 ± 0.0029 for the awake and 0.059 ± 0.014 for the anesthetized group. The k4 is the rate of dephosphorylation and that was close to 0.0109 ± 0.00028 for awake and 0.015 ± 0.0047 for anesthetized pigeons. The metabolic rate constant

Ki was determined in all groups (n = 16,Table 6). For the i.v. awake group it averaged at 0.018 ± 0.00085, for the i.v. anesthetized group at 0.013 ± 0.0048, and for the i.p. anesthetized group at 0.0071 ± 0.00098 for the i.p. anesthetized group.

90

Table 4 Whole-brain kinetic rate constants of the i.v. awake group. The 2-compartment model generates four

18 kinetic rate constants that describe the metabolic process of F-FDG/glucose. K1 describes the rate of transport from plasma to tissue, and k2 the transport back into the plasma. k3 refers to the rate of phosphorylation and k4 the rate of dephosphorylation.

Pigeon ID K1 k2 k3 k4 752 0.059467 0.145314 0.065471 0.011181 886 0.052587 0.139686 0.065656 0.010692 920 0.053711 0.139874 0.067666 0.010586 986 0.048302 0.116201 0.071797 0.011052

Mean ± 0.053517 ± 0.135269 ± 0.067648 ± 0.010878 ± SD 0.004601 0.012977 0.002939 0.000284

Table 5 Whole-brain kinetic rate constants of the i.v. anesthetized group. The 2-compartment model generates

18 four kinetic rate constants that describe the metabolic process of F-FDG/glucose. K1 describes the rate of transport from plasma to tissue, and k2 the transport back into the plasma. k3 refers to the rate of phosphorylation and k4 the rate of dephosphorylation.

Pigeon ID K1 k2 k3 k4 146 0.03852 0.14062 0.045927 0.015046 232 0.034705 0.117774 0.06212 0.01051 536 0.081841 0.14606 0.056323 0.013523 590 0.051934 0.145407 0.044475 0.017242 645 0.025632 0.069231 0.082532 0.022591 877 0.031792 0.116077 0.062161 0.010143

0.044071 ± 0.122528 ± 0.058923 ± 0.014843 ± Mean ± SD 0.020483 0.029357 0.013874 0.004657

91

18 Table 6 Whole-brain rate constant Ki per group. Ki describes the net influx rate of F-FDG and is calculated from K1k3/(k2+k3).

Group Pigeon ID Ki i.v. awake 752 0.01842413 886 0.0167502 920 0.01720763 986 0.0183946 Mean ± SD 0.01769414 ± 0.000847 i.v. anesthetized 146 0.00936722 232 0.01154268 536 0.02260573 590 0.01178818 645 0.01303513 877 0.01068067 Mean ± SD 0.01316993 ± 0.004781 i.p. anesthetized 933 0.00759835 938 0.00539267 980 0.00659152 980_2 0.008066 981 0.00709664 987 0.00775446 Mean ± SD 0.00708327 ± 0.000978

The glucose consumption rate, expressed as CMRglc, was calculated from the metabolic rate constant Ki divided by the LC, and multiplied by the post-scan venal plasma glucose level (see Table 7). There was a significant difference between the three groups in glucose consumption, F(2,12) = 27.20, p < 0.01, η2 = 0.92 (see Figure 21). The cerebral tissue of awake pigeons that received an i.v. injection consumed more glucose (forebrain: 27.79 ± 2.11 µmol/100gr./min, cerebellum: 24.90 ± 2.88 µmol/100gr./min.) than both other groups (p < 0.01). In addition, the anesthetized group that received an i.v. injection had higher levels of CMRglc (forebrain: 18.33 ± 3.11 µmol/100gr./min., cerebellum: 20.78 ± 3.58 µmol/100gr./min.) than the anesthetized group that received an i.p. injection (forebrain: 12.66 ± 2.63 µmol/100gr./min., cerebellum: 13.84 ± 2.56 µmol/100gr./min., p < 0.01).

92

Table 7 CMRglc per group. The cerebral metabolic rate of glucose per 100 gr. of tissue (forebrain, cerebellum or whole brain) per minute per group. The outlier 536 from the i.v. anesthetized group was excluded from analysis.

State Injection Pigeon id Forebrain Cerebellum Whole brain

Awake i.v. 752 30.95 28.12 29.53 886 26.75 24.73 25.74 920 26.85 21.16 24.01 986 26.61 25.60 26.11 Mean ± SD 27.79 ± 2.11 24.90 ± 2.88 26.35 ± 2.31 Anesthetized i.v. 146 16.14 19.23 17.69 232 20.64 23.28 21.96 536* 46.95 52.97 49.96 590 21.99 24.80 23.39 645 18.46 20.93 19.69 877 14.43 15.65 15.04 18.33 ± 3.11* 20.78 ± 3.58* 19.55 ± 3.33* Mean ± SD 23.10 ± 12.01† 26.14 ± 13.53† 24.62 ± 12.77† Anesthetized i.p. 933 14.94 17.18 16.06 938 9.61 9.94 9.77 980 11.64 12.25 11.94 980_2 16.71 15.77 16.24 981 11.40 14.13 12.76 987 11.66 13.74 12.70 Mean ± SD 12.66 ± 2.63 13.84 ± 2.56 13.25 ± 2.50

* without outlier, † with outlier

Moreover, CMRglc differed between the forebrain and the cerebellum, depending on Group F(2,12) = 16.82, p < 0.01, η2 = 0.74, see Figure 21. The rerun analysis showed that only for the anesthetized group with an i.v. injection the cerebellum consumed more than the forebrain F(1,4) = 57.29, p < 0.01, η2 = 0.94. A similar trend was observed for the anesthetized group that received an i.p. injection, F(1,5) = 4.19, p = 0.096, η2 = 0.46. Lastly, in the awake group with i.v. injection the difference approached significance, F(1,3) = 8.25, p = 0.064, η2 = 0.73, where the forebrain consumed more glucose than the cerebellum.

93

Figure 21 Boxplots of the average CMRglc of forebrain and cerebellum per group. The three groups differed significantly in their glucose uptake (F(2,12) = 27.20, p < 0.01, η2 = 0.92), and depending on group, CMRglc also differed between forebrain and cerebellum (F(2,12) = 16.82, p < 0.01, η2 = 0.74). The i.v. awake group had a higher CMRglc than both other groups. In this group, the difference between brain regions approached significance (F(1,3) = 8.25, p = 0.064, η2 = 0.73). The i.v. anesthetized group had a higher glucose uptake than the i.p. anesthetized group, and the cerebellum significantly consumed more than the forebrain (F(1,4) = 57.29, p < 0.01, η2 = 0.94). A similar trend of difference in brain region consumption was observed for the i.p. anesthetized group, yet it was not significant (F(1,5) = 4.19, p = 0.096, η2 = 0.46). ** = p < 0.01.

In order to compare the found CMRglc of the pigeon with the mammalian neuronal energy budget, the awake data following i.v. injection is divided by the known number of neurons of the pigeon brain (Olkowicz et al., 2016). Since it was found that pigeon brain tissue requires 26.35 ± 2.31 µmol glucose per 100 gr. per min, this comes down to 1.78 * 10-9 µmol glucose per neuron per minute. This is 3.25x lower than the found mammalian neuronal energy budget (Herculano-Houzel, 2011).

94

4.4 Discussion

The aim of this study was to estimate the energetic costs of brain tissue of an avian species, the pigeon, and compare the derived cerebral metabolic rate of glucose (CMRglc) per neuron with the neuronal energy budget of mammals. In addition, since this is one of the first experimental approaches to assess the CMRglc in a bird species, some methodological variables were considered. Cerebral glucose consumption was assessed with positron- emission tomography (PET) with fluorine-18 fluorodeoxyglucose (18F-FDG) as radiotracer in either awake or anesthetized pigeons following an i.v. or i.p. 18F-FDG injection. The exact

CMRglc can then be estimated with a kinetic modelling approach, following Backes et al. (2011). It was found that pigeon brain tissue requires 26.35 ± 2.31 µmol glucose per 100 gr. per min. This translates into a neuronal energy budget of 1.78 * 10-9 µmol glucose per neuron per minute, which is 3.25 times lower compared to the average mammalian neuron. Such a striking magnitude of difference points towards structural dissimilarities possibly present within the avian cerebral bauplan. Before elaborating on possible explanations for the lower energy consumption, first some methodological considerations and validation of the approach are discussed.

4.4.1 Methodological considerations and validation Sokoloff et al (Sokoloff et al., 1977) were the first to develop a quantitative approach to measure local cerebral glucose consumption. They used the radioactively labelled glucose analogue 2-deoxy-D-14C-glucose (14C-DG) as a tracer and quantified its distribution in the brain with autoradiography. Here, they also developed the model to determine the kinetic rate constants from an arterial input function and brain tissue activity. The principles of this technique were combined with PET to facilitate dynamic in vivo estimations of CMRglc. In the PET studies, 14C-DG is replaced by 18F-FDG, but like its predecessor it is a glucose analogue that competes for the same transport carriers and hexokinase for phosphorylation. Due to the lack of a hydroxyl group at the second carbon atom, the glucose analogues cannot be metabolized beyond phosphorylation and are trapped in the cell (Phelps et al., 1979; Reivich et al., 1985). At the time, no significant differences were found between CMRglc estimated with 14C-DG autoradiography method or the 18F-FDG PET approach (Reivich et al., 1985), though a more recent analysis found that the latter approach results in slightly lower values (Toyama et al., 2004). Since the 18F-FDG PET method is (almost) non-invasive, and allows for real-time quantification of metabolic rates, it has become the preferred method. Especially combined with kinetic modelling with an arterial input function, it is the golden standard approach to quantify CMRglc values (Berti, Vanzi, Polito, & Pupi, 2013). Since this is one of the first times

18 the CMRglc of a bird has been determined with F-FDG PET, some discussion of different methodological parameters is required.

95

4.4.1.1 Lumped constant A crucial variable in the model is the lumped constant (LC), which describes difference in the

18 kinetics of glucose and F-FDG. The accuracy of the CMRglc strongly depends on an accurate LC, and the two variables are inversely correlated (Krohn et al., 2007). There is somewhat variation in the LC-values employed in mammalian studies, and these discrepancies are thought to be one of the sources of variation of determined CMRglc in the literature (Alf et al., 2014). Here, a model-based determination of the LC was employed that separates the LC into the underlying processes it describes; (1) the difference of 18F-FDG and glucose in rate of transport to and from the cell, and (2) the difference of 18F-FDG and glucose in the rate of phosphorylation (Backes et al., 2011). Based on Michaelis-Menten formalism, which describes enzyme kinetics, the ratio of difference between two substances that use the same system is a constant, because it describes ratios between constants (Gjedde & Diemer, 1983). Thus, there are fixed ratios of difference between glucose and 18F-FDG in transport and phosphorylation, which can be included in the model to obtain more physiologically meaningful values (Cunningham & Cremer, 1981; Kuwabara et al., 1990). Here, an LC of 0.745 ± 0.0898 was found for pigeons. This value is at the high end of the range employed in rat and mouse studies (0.45 – 0.7; Alf et al., 2014; Alf, Martić-Kehl, et al., 2013; Kuwabara et al., 1990). Since the obtained values did not differ between the awake and anesthetized group, it was employed to all groups. This corroborates the results of Sokoloff et al. (1977) in rats, and is grounded in the idea that since the LC describes ratios of kinetic constants, it is expected to be an inherent constant for a species and to not vary depending on awake or anesthetized state.

4.4.1.2 Route of injection Most dynamic 18F-FDG-PET studies inject 18F-FDG intravenously (i.v.) to facilitate a quick uptake of the tracer and a good distribution over the entire body. Less common is a intraperitoneal (i.p.) injection, since the uptake is much slower and there is a higher chance the tracer is injected into an organ and remains close to the place of injection (Gaines Das & North, 2007). Because the biodistribution of the tracer is comparable to an i.v. injection after 60 minutes, it is still considered a valid alternative (Fueger et al., 2006). The i.p. injection is more often used in static PET experiments, where the slow tracer uptake is a beneficial aspect. In static PET experiments, the tracer is injected outside the scanner and during the uptake the animal is subjected to a specific behavioural task or is allowed to move around freely. After 30- 45 minutes the animal is anesthetized and a short scan reveals the relative distribution of the tracer to specific brain regions (Alstrup & Smith, 2013). Another strong plus point of the i.p. injection is that it is less invasive for the animal and imposes fewer technical challenges, compared to a venal catheter placement (Hubrecht & Kirkwood, 2010).

96

In this study the two routes of injection were compared to investigate whether it had an effect on the found CMRglc. Indeed, a significant difference between the i.v. anesthetized and i.p. anesthetized group was found, where the latter had lower CMRglc values compared to the i.v. group (see Figure 21). This is probably a consequence of the different modelling approaches used for the two routes of injection. Namely, the i.v. data was analysed with a two-tissue compartment model with four kinetic rate constants, whereas the i.p. data was modelled with the Patlak plot. This modelling approach to i.p. data is necessary, because the tracer uptake is much slower and as a consequence the arterial input function does not provide enough information to disentangle the four different kinetic parameters. Instead, only the net 18F-FDG influx rate (Ki) is derived from the Patlak graphical analysis, and k4 is assumed to be zero

(Patlak & Blasberg, 1985). A study in mice showed, that when they compared CMRglc following either an i.v. or an i.p. injection, but analysed both data-sets with the Patlak plot, no difference depending on route of injection was found (Wong, Sha, Zhang, & Huang, 2011). Furthermore, Alf et al., (Alf, Wyss, et al., 2013) analysed the exact same data-set with either a two-tissue compartment model or the Patlak analysis, and found that the second approach produced a significantly lower CMRglc. They concluded that, especially for longer time-scans, the k4 cannot be assumed zero and that significant dephosphorylation does take place. The same can be concluded for the pigeon data-set. The results from the two i.v.-groups show that the k4 ranges from 0.0101 – 0.226 (see Table 4, Table 5) and leaving this parameter out will result in an underestimation of the CMRglc.

4.4.1.3 State and brain region The one significant benefit of the 14C-DG autoradiography approach was that it allowed for quantification during awake uptake, whereas almost all non-human 18F-FDG PET studies are done in anesthetized animals. In the 14C-DG autoradiography approach, the radioactive tracer is injected, and the animal can walk around freely until approximately 30-45 minutes after injection when the animal is sacrificed for analysis (Sokoloff et al., 1977). Indeed, besides humans, awake 18F-FDG PET experiments have been done only once in mouse (Toyama et al., 2004), rat (Huang et al., 2017), and macaque monkey (Noda, Takamatsu, Minoshima, Tsukada, & Nishimura, 2003). Here, following Behroozi et al. (2018), an awake scanning protocol was developed for pigeons. The animals were fixated in a holder with an implanted head-block, and habituated to test and procedure to reduce stress. With this procedure, it was possibly to scan the pigeons awake and compare CMRglc in different states. Isoflurane anaesthesia reduced whole brain energy consumption of the pigeon brain by 26%. This replicates the findings in the macaque monkey, where a reduction of 20% was found (Noda et al., 2003), both of which are lower than the rat who displays a CMRglc reduction of 64% (Huang et al., 2017). Interestingly, in the pigeon there was a significant difference between forebrain

97

and cerebellum CMRglc and anaesthesia. Namely, in the awake state, there was a trend that the forebrain consumed relatively more glucose compared to the cerebellum, whereas in the anesthetized state this pattern was reversed (Figure 21). Thus, it appears that anaesthesia has a relatively stronger decreasing effect on the forebrain than the cerebellum. This could be explained by the relative suppressing effects of the functions of the forebrain and cerebellum during restrained awake and anesthetized state. It can be expected that the biggest difference would apply to sensory processing areas, which are predominantly situated in the forebrain (Reiner, Perkel, Bruce, et al., 2004). Indeed, even though no such a difference between the cortex and cerebellum has been noted in mammals, the strongest decrease in activity following anaesthesia in the rat was found in the thalamic auditory medial geniculate nucleus and auditory cortex in the rat (Sokoloff et al., 1977), and the visual occipital cortex in the macaque (Noda et al., 2003). These are indeed the dominant sensory systems for these two species (Burn, 2008; Kaas, 2020). Like primates, pigeons are highly visual animals (Güntürkün, 2000), thus we would expect the largest decrease in the avian analogue to the primary visual cortex. This structure is known as the entopallium, situated in the centre of the forebrain, and it indeed shown the highest level of activity in the awake state (Figure 18). It was shown in the Japanese quail that compared to surrounding tissue, the entopallium is rich in mitochondria (Watanabe, Ito, & Ikushima, 1985b). This is another indicator for the high metabolic demand of this structure in the avian forebrain (Borowsky & Collins, 1989). Thus, it is likely that the significant effect between forebrain and cerebellum was mainly driven by this nucleus.

4.4.1.4 Comparison to other bird PET study The field of avian PET imaging is in its infancy. So far, there has been only one account that executed a preliminary modelling study to get an insight into kinetic dynamics and the CMRglc in the avian brain (Salerno et al., 2019). The study employed a comparable methodological approach to here, where a dynamic PET scan following an 18F-FDG i.v. injection was fitted to a two-tissue compartment model with four rate constants, and an input function was derived from blood samples. The study was executed in pigeons under isoflurane anaesthesia, and they estimated a significantly higher CMRglc of 39.6 µmol/100 gr/min, compared to the 19.55 µmol/100 gr./min. found here.(Table 5). There are two crucial differences in methodology that can explain the discrepancy in results. First, Salerno et al. (2019) set the LC to the value of 0.58, which is typically used in rats (Schiffer, Mirrione, & Dewey, 2007). The findings here, however, show that the LC in the pigeon is typically higher than the LC values found in rat and mice studies. Importantly, an underestimation of LC results in an overestimation of CMRglc (Krohn et al., 2007). Secondly, Salerno et al. (2019) derived the blood samples for the input function from the brachial vein, whereas we sampled from the brachial artery. The significant difference is that the brain will extract tracer from the available 18F-FDG within the arterial

98

plasma pool, whereas the venal tracer concentration represents the post-extraction scenario for the full body. As a consequence, the plasma concentration over time is underestimated, which again will result in an overestimation of the CMRglc.

4.4.2 Neuronal energy budget of mammals and birds As mentioned in the introduction, across mammals, the glucose consumption per neuron shows only a small degree of variation between different species. As such, the average mammalian brain neuron was estimated to utilize 5.79 * 10-9 µmol glucose per minute

(Herculano-Houzel, 2011). Dividing the estimated CMRglc in the awake pigeon by the known number of neurons (Olkowicz et al., 2016), gives the neuronal energy budget of the pigeon of 1.78 * 10-9 µmol glucose per minute. Compared to mammals, this is a difference with a factor of 3.25 (seeFigure 22). This finding explains how birds, or at least pigeons, are able to sustain almost twice as many neurons compared to the similar sized rat, without the associated metabolic costs. What could be the underlying mechanism that the neuronal costs in an avian species are so much lower compared to mammalian counterparts?

Figure 22 CMRglc per whole-brain neuron across species. Graph depicts neuronal densities per mg on x-axis and CMRglc per mg on the y-axis representing CMRglc per neuron across different mammalian species (human, macaque, baboon, rat and mouse) and the pigeon. All mammalian species fall on the same regression line that can be described by a linear function 5.76*10-6 + 5.62*10-9X (R2=0.949, p = 0.005). This corresponds to an average

-9 -9 -9 CMRglc per neuron of 5.79*10 ± 0.76*10 . In contrast, in the pigeon, a neuron consumes on average 1.78*10 , which is approximately 3.25x lower than the average mammalian neuron. Data from mammalian species comes from Herculano-Houzel, 2011, silhouettes are obtained from phylopic.org.

99

Before I can elaborate on possible energy efficient approaches in the nervous system of birds, it is important to understand how the energy budget of the nervous system is divided over the different components. The energy budget of the brain can be subdivided in signalling and nonsignaling costs, that differ for neurons and glial cells. Signalling costs are dependent on the number of cells and firing rate, whereas the nonsignaling costs are the sole product of cell numbers (Yu, Herman, Rothman, Agarwal, & Hyder, 2017). Independent of cell-type, it was estimated that signalling demands approximately 70-80% of the energy budget, and thus nonsignaling costs account for the remaining 20-30% (Howarth, Gleeson, & Attwell, 2012; Hyder et al., 2013; Yu et al., 2017). In neurons, synaptic transmission is responsible for almost three-quarters of the signalling related costs, and action potential generation accounts for approximately 15%, the remaining 10% is divided over glutamate and GABA recycling, and calcium activity . The nonsignaling costs are divided between housekeeping costs, which is basic upkeep of the cell, and maintenance of the resting membrane potential (Howarth, Gleeson, & Attwell, 2012; Hyder et al., 2013). For glial cells, there are additional spiking-related calcium activity costs that consume 30-50% of the glial cell budget (Verkhratsky & Kettenmann, 1996; Yu et al., 2017). This budget shows that neurons are indeed the biggest energy sink of the brain, and even energetically costly when not active.

Compared to primates, the brain tissue of rodents consumes three times more glucose per mg brain tissue (Figure 22). However, since mice and rat have triple the neuronal densities, the final neuronal energy budget is comparable (Herculano-Houzel, 2011). Theoretical calculations of the metabolic costs of different cerebral components in rats and humans identified that the higher metabolic costs per gram of brain tissue are primarily a consequence of the higher firing rate in rodents, and the associated action potential and synaptic transmission costs (Hyder et al., 2013; Yu et al., 2017). Though there is a wide variety of frequencies, the average firing rate of the human brain is estimated at 1.15 Hz, whereas in rats around 4.3 Hz was calculated (Hyder et al., 2013). It seems that rodents and primates have two different coding strategies that are strongly linked to cell size (Levy & Baxter, 1996; Niven, Anderson, & Laughlin, 2007; Sengupta, Faisal, Laughlin, & Niven, 2013; Sengupta, Stemmler, Laughlin, & Niven, 2010; Yu et al., 2017). In mice and rats, the brain consists of many small neurons that have lower maintenance and resting costs, and lower spiking costs due to a smaller axon and dendritic tree with fewer ion channels and synapses (Attwell & Laughlin, 2001; Braitenberg & Schüz, 1998; Vetter, Roth, & Häusser, 2001). On the downside, smaller neurons severely penalize the information coding capacities of one neuron (Faisal, White, & Laughlin, 2005; Simon B. Laughlin & Sejnowski, 2003; Sengupta et al., 2013). In contrast, in primates there are fewer larger neurons per mg, that each have a high signal-to-noise ratio, but concomitant high signalling and nonsignaling costs (Braitenberg & Schüz, 1998;

100

Herculano-Houzel, Manger, & Kaas, 2014; Niven et al., 2007; Sengupta et al., 2013). It was estimated that a single spike in the human cortex is 3.3 times more costly than in the rat cortex (Lennie, 2003). These different set-ups translate for the rodent into low amounts of information being encoded by many small neurons at high spiking frequencies, and in primates the large compartments code sparsely with a high information rate.

The pigeon brain contains almost twice the number of neurons of a rat brain (Herculano-Houzel et al., 2006; Olkowicz et al., 2016). Following model predictions, this would imply a small cell size (Mota & Herculano-Houzel, 2014), with a concomitant low signal to noise ratio (Sengupta et al., 2013), and thus a high firing rate (Niven & Laughlin, 2008a), fitting to the rodent bauplan. However, since the pigeon does not show the associated metabolic costs of the rat layout, there must be crucial differences in place. To test the model predictions, the obvious parameter to asses would be the firing rate, since this appears to be the driving factor of difference between the rat and the human brain. Unfortunately, the emerging field of electrophysiology in pigeons and other birds has not amounted to a large enough data set to determine an average firing rate. Based on empirical and modelling work that indicates that small neurons have high firing rates (Sengupta et al., 2013, 2010), it would be most parsimonious to expect pigeons to have a comparable spiking frequency to rats. This implies there are energy efficient strategies in place, related to bird-unique factors, that have reduced the costs of each spike.

The key mechanism to reduce the cost of a spike in a similarly sized neuron is a reduction of the overlap of Na+ and K+ currents that flow across the membrane during an action potential (Sengupta et al., 2010). Namely, an early onset of K+ channels will generate an outward K+ that will neutralize the inward Na+, and this will increase the amount of Na+ flow into the cell that is needed the propagate the action potential. The concomitant costs stack up after the action potential, when the Na+ K+-ATPase needs to pump Na+ and K+ across the membrane to restore the resting membrane potential at the expense of ATP (Hasenstaub, Otte, Callaway, & Sejnowski, 2010). One of the first reports of the extent of overlap was from the giant squid axon. In this neuron, it was found that the efficiency of the overlap was 4 times higher than the theoretical minimal charge required to depolarize a pure capacitor (Hodgkin & Huxley, 1952). This is much higher than the data from mammalian neurons; in hippocampal mossy fibres, the Na+ influx was only 1.3 times higher than the theoretical minimum (Alle, Roth, & Geiger, 2009), and a comparable efficiency of 1.25 was found in cortical pyramidal neurons (Carter & Bean, 2009). The increased efficiency is a consequence of fast and almost complete Na+ channel inactivation, before opening of K+ channels (Schmidt-Hieber & Bischofberger, 2010).

101

The change in time constants of ion channels in mammalian neurons compared to squid are a consequence of the warmer body temperature of mammals (Yu, Hill, & McCormick, 2012). Birds are, like mammals, endotherms, but are characterized by even higher body temperatures (Prinzinger et al., 1991). As a consequence, the core brain temperature of the pigeon is with 40-42ºC (Kilgore, Bernstein, & Hudson, 1976; Pinshow, Bernstein, Lopez, & Kleinhaus, 1982) much higher compared to 36-37ºC of the rat brain (Alföldi, Rubicsek, Cserni, & Obál, 1990). Yu et al. (2012) modelled the time constants of Na+ and K+ channel activation and inactivation, and found an exponential and strong decrease with increases in temperature. A significant difference of Na+ activation, Na+ inactivation, and K+ activation was even observed between 36 ºC and 42 ºC (see figure 4A-C in Yu, Hill, & McCormick, 2012). Thus, it could be that the higher body temperature of the pigeon reduces the overlap between Na+ and K+ currents, which minimizes ATP costs per action potential. However, an important secondary finding of the temperature-based modelling showed that an increased temperature also predicts higher firing rates since spike afterhyperpolarization is reduced, especially for temperatures above 36 ºC (Yu et al., 2012). It is currently unclear how firing rate and ion channel kinetics are organized in the bird brain, and exact physiological measurements are required before this can be disentangled. However, the temperature-dependent crucial role of Na+ and K+ overlap in neuronal energy efficiency, and higher brain temperatures in the bird brain could function as an important starting point.

Other possible cost reductive solutions relate to the bird-specific organization of the pallium, and myelination of axons. The bird brain has a drastically different lay-out compared to the mammalian cortex. Namely, instead of a six-layered cortex, the avian pallium is of a nuclear organization (Jarvis, 2009), and as a consequence, combined with the high neuronal density, the inter-neuronal distance is shorter compared to mammals. This fits with an important energy efficient strategy of ‘saving wire’, where there is a penalty on long-distance connections since these are metabolically costly to maintain, and metabolically inefficient since they dissipate energy (Laughlin & Sejnowski, 2003). Indeed, the layout of the cortex is designed in such a way to minimize wiring length (Cherniak, Mokhtarzada, Rodriguez-Esteban, & Changizi, 2004; Klyachko & Stevens, 2003). Despite this different gross morphological layout, the connectome of birds is highly similar to the mammalian cortex, with modular small-world networks that communicate via far-ranging connections, and rely on central connection hubs (Shanahan et al., 2013). However, due to the nuclear as opposed to layered organization, it is possible the small-worlds are connected by shorter path-lengths, and thus cut costs for inter-area communication (Olkowicz et al., 2016; Shanahan et al., 2013).

102

Another strategy to reduce wiring costs, is to myelinate the axons (Yu et al., 2012). Traditionally, it was predominantly observed in far-ranging axons, such as descending motor neurons, where it is known to increase speed of neuronal signalling up to a 100-fold (McDougall et al., 2018; Moore, Joyner, Brill, Waxman, & Najar-Joa, 1978; Waxman & Bennett, 1972). More recently, it has also been implied as a metabolically saving strategy (Perge, Niven, Mugnaini, Balasubramanian, & Sterling, 2012). Especially for the insulation of cortical neurons with a short action radius, the myelin cannot be explained by time-efficiency (Micheva et al., 2016). Because of the specific avian bauplan mentioned earlier, the forebrain of birds does not consist of the classical white matter – grey matter division. Instead, the entire pallium is an interplexus of neuropil and myelinated axons, with some larger tract and fibre bundles (Letzner, Simon, & Güntürkün, 2016; Rehkämper & Zilles, 1991). The short-range myelinated tracts could represent an energy-saving strategy.

It is important to note that most of these modelling studies are based on whole brain and organism cost-benefit analyses. This implies that the details of variation in cell size, axon length, number of synapses, myelination, spiking frequency, information rate, and ion channel kinetics are given less weight for the benefit of understanding the evolutionary pressures that give rise to certain systems. The modelling is grounded in empirical evidence, but for each general rule, it is possible to find an exception. Therefore, it is of crucial importance to gather more empirically quantified numbers on these different parameters of the avian brain before hard conclusions can be drawn on the metabolically efficient strategies of the bird brain. For now, the interim conclusion is a very exciting one. Since the nervous system is so metabolically expensive, there is a strong evolutionary pressures to reduce the costs while maintaining the required function (Hasenstaub et al., 2010; Simon B. Laughlin & Sejnowski, 2003). Indeed, across the nervous systems of vertebrates and invertebrates, we find energy efficient solutions that strike a balance between the required information that needs to be encoded to generate specific adaptive behaviour, and the energy expended (Niven & Laughlin, 2008b). This study demonstrated that pigeons are able to sustain almost twice the number of neurons compared to the similarly-sized rat due to a much lower metabolic demand per neuron. It seems that in the 300 million years of separate evolution the avian brain developed more energy efficient strategies to keep the costs per neuron low.

103

104

Chapter 5

General discussion

105

106

5.1 Summary

In the introduction, I asked how the small bird brain is able to generate such impressive cognitive behaviours. As I pointed out, absolute brain size has not been considered the appropriate proxy for cognitive capacities for a long time as indeed, size is not the only difference between avian and mammalian brains (Dicke & Roth, 2016). In the millions of years of separate evolution, the forebrain of both lineages organized in radically different ways. In mammals, a six-layered neocortex developed from the dorsal pallium, whereas in birds and reptiles a nuclear organization matured from the ventral pallium (Jarvis, 2009; Puelles et al., 2017). Compared to the neocortex, it is thought that the avian/reptilian bauplan brings about certain advantages such as the possibility to densely pack high numbers of neurons and maintain short interneuronal distances (Olkowicz et al., 2016). However, despite the sauropsid- specific cerebral organization, there are numerous examples of comparable cerebral processes that rely on analogous functional areas which operate within a similar circuitry as the neocortex (Kröner & Güntürkün, 1999; Shanahan et al., 2013; Stacho et al., 2020). These similarities underpin the comparability of cognition and behaviour in birds and mammals. What this suggests is that the parallel evolution of cognition in birds was pushed by equivalent socio- ecological challenges as those of mammals, but confined to the framework of the sauropsid cerebral bauplan (Striedter, 2005; Wylie et al., 2015).

In this thesis I researched two phenomena that discuss different aspects of the underlying building blocks of avian complex cognition. First, I looked at an executive structures in birds, reptiles and mammals that evolved in times of comparable socio-ecological conditions, namely the NCL and the PFC. Secondly, I researched the avian specificity of high neuron densities and discovered how the bird is able to sustain such high numbers of neurons. To summarize, in the first study, I investigated the existence of a primordial NCL-like structure in the Nile

107

crocodile and identified an area of higher dopaminergic innervation in a topologically equivalent location to the NCL in birds. This possibly demonstrates the presence of an NCL-like structure in at least the last common ancestor of birds and crocodiles. In the second study, I further explored the location and extent of the NCL in different bird species and found that it diversifies depending on order and even family. Namely, compared to pigeon and chicken, the Passeriformes can be characterized by an NCL that spans across the entire back of the caudal forebrain and can be partitioned in at least four subdivisions. This indicates that avian species with more complex capacities have an extended and more parcellated NCL. Lastly, in study three, I estimated the neuronal energy budget of the pigeon brain and found that avian neurons on average consume 3.25 times less sugar than mammalian neurons. Such a magnitude of difference must point towards a structural dissimilarity present within the basic bauplan of the avian brain. Even though no definite conclusions can be drawn, I suggested three explanations for the observed phenomenon: (1) a higher body temperature facilitates more metabolically efficient channel kinetics, (2) the specific nuclear lay-out of the avian brain facilitates more efficient wiring and (3) the pattern of myelination of especially short-distance myelin tract reduces wiring costs.

In this discussion, I will incorporate my findings into a bigger framework that places the capacity for flight into a prominent position as both a driving force as well as a limiting factor to brain expansion, development of the NCL, and energy-efficient high neuron densities. First, I will attempt to reconstruct the evolution of the NCL in the archosaurian lineage. To this end, I will elaborate on the possible origin of the NCL, and compare the current findings on size and subdivisions in the crocodile and the different avian species. Next, I will discuss three proxies that will aid in the reconstruction of the evolution of this executive structure, namely general brain expansion, endothermy and behavioural flexibility. Second, I will consider cognition as it evolved in a flying body by evaluating avian energetics, oxidative phosphorylation (a/k/a aerobic respiration), and the impact on the bird brain

5.2 The evolution of an executive structure in the archosaurian lineage

5.2.1 The NCL in reptiles, birds and dinosaurs As discussed extensively in Chapter 3, the NCL and PFC are the key brain areas that function as an executive structure to facilitate (complex) cognition. An executive structure operates at the highest stage of the perception–action cycle, where it processes and manipulates cognitive networks that represent associations, knowledge or memory, over time (Fuster, 2015). Importantly, based on the long separate evolution of mammals and birds, their pallial origin,

108

and morphological dissimilarity, it is widely accepted the NCL and PFC evolved convergently in each lineage (Güntürkün, 2005b, 2012; Güntürkün & Bugnyar, 2016; Kröner & Güntürkün, 1999). As described in Chapter 2, in each reptilian order, anterior nidopallial parts of the DVR process distinct sensory modalities that are relayed to a caudal nidopallial field. This caudal territory is furthermore interconnected with thalamic nuclei, basal ganglia, hippocampus, hypothalamus and amygdala (Lanuza, 1998; Ulinski, 1983). In this thesis I demonstrated that in the crocodile, a subpart of this caudal field is possibly more extensively modulated by dopamine, which is another hallmark of an executive structure (Kröner & Güntürkün, 1999; Waldmann & Güntürkün, 1993; Wynne & Güntürkün, 1995). To the best of my knowledge, the presence of this structure has not yet been demonstrated in other reptiles. Even though this type of connectivity and dopaminergic modulation of the Ncl could be an example of convergent evolution, a more parsimonious explanation would be that the stem archosaurian possessed an expanded ventral pallium that processed visual, somatosensory and auditory information and relayed this to a caudal field interconnected with motor, memory and limbic structures. Thus, a tentative origin of an NCL-like structure can be dated to the split of crocodiles and birds approximately 245 mya (Benton & Donoghue, 2006). This is an exciting finding because it dates back the origin of the executive structure in the avian lineage approximately 150 million years. Moreover, it will have intriguing consequences for how we view the behavioural capacities of other extinct archosaurs such as Tyrannosaurus rex, and Stegosaurus ungulates, since this would imply they were possibly equipped with at least a primordial Ncl-like structure such as demonstrated in the crocodile. The Ncl as discovered in the crocodile differs from the avian NCL especially in size of the structure, and the number of subdivisions.

5.2.2 Size and subdivisions The most striking difference is that in the crocodile, the Ncl appears to occupy a relatively smaller space compared to what has been observed in birds. I would argue this is in line with a trend observed in the four avian species, and which is also present in mammals, where fewer cognitive capacities require a smaller and less diversified executive structure (Passingham & Wise, 2012). In Chapter 3, I described that the non-Passeriformes had a smaller and less differentiated NCL compared to the songbirds, corroborating the limits of their cognitive capacities. Such a trend has also been observed for the PFC in mammals. As mentioned, the PFC is not a uniform structure across mammals. Compared to mouse and rat, the primate lineage can be characterized by an expanded and more gyrified frontal lobe that appears to be more parcellated (Passingham & Wise, 2012). These are clear examples of the Principle of Proper Mass, which details how the expansion of a brain region executing specific tasks is proportional to the importance of these processes for the execution of crucial behaviours for

109

the organism (Jerison, 1973). This is best exemplified with the prefrontal cortex in humans. Compared to non-ape primates, great apes and humans have an exceptionally larger PFC, indicated by a non-allometric scaling factor (Smaers, Gómez-Robles, Parks, & Sherwood, 2017).

Importantly, the expansion gives rise to more differentiated subdivisions of the PFC that vary in their connectivity, neurotransmitter modulation, and function. Namely, the primate PFC can be subdivided in dorsolateral, dorsomedial, ventromedial and orbital prefrontal section (Passingham & Wise, 2012). These subdivisions have a distinctive granulated layer 4, which has not been demonstrated in the PFC of rodent or any other mammal (Brodmann, 1909). This has problematized definite conclusions on one-to-one homologies between rodent and primate PFC (Carlén, 2017). The general consensus for now is that the rodent PFC is much less segregated and instead shows an intermingling of anatomical and functional characteristic (Fuster, 2015; Uylings et al., 2003). Again, this seem to replicate the observations made in Chapter 3. In the pigeon and chicken, the NCL is confined to specific area, with a high degree of overlap in connectivity and function. In the Passeriformes, the areas are more segregated. Even though much less is known about their functionality, the observed differences in connectivity hint at diverting functional roles. In the crocodile, no subdivisions could be observed on the basis of dopaminergic innervation, nor did any hodological analysis reveal subdivisions.

5.2.3 Three proxies for reconstruction The discovery of an Ncl in the crocodile dates the origin of the NCL to the basal archosaur that lived 245 mya. This finding allows for a reconstruction of the evolution of the NCL within the archosaurian lineage. However, since nothing is known about this structure in extinct species, I will discuss three proxies that substantiate both the origin of this executive structure 245 mya and allow for a reconstruction of how the NCL evolved from the small undifferentiated area in the crocodile, to the expanded and diversified structure as observed in the crow. The proxies I will discuss are relative brain expansion, the benefits of behavioural flexibility, and the evolution of endothermy.

Since neural tissue does not fossilize well, it will not be possible to detail the NCL of extinct species. Instead, I will use general forebrain expansions as a proxy for NCL enlargement. Though there are several issues with this approach, as outlined in the introduction, I would argue that here it is valid because part of the observed size difference between the avian NCL and the reptilian Ncl can be explained by differences in general forebrain expansion. As mentioned earlier, compared to reptiles, birds and mammals have a more encephalized brain (Jerison, 1973). Specifically, for Passeriformes and primates, this expansions has been

110

ascribed to an enlarged mesopallium and nidopallium, and neocortex respectively (Passingham & Wise, 2012; Rehkämper et al., 1991). Whereas this has not yet been demonstrated for birds, a clear allometric relationship between the size of the neocortex and the PFC has been described for most mammalian species (Smaers et al., 2017). This is evidence for concerted brain expansion, and it is not unlikely something similar occurred in the avian lineage. The time-points of brain expansion can be dated by investigating digitalized endocast of fossils within the ancestral lineage of birds, which can be used to estimate brain size, and even forebrain size (Walsh & Knoll, 2017; A. Watanabe et al., 2018). In addition, femur-length can be used as a proxy for body size (Seymour, Bennett-Stamper, Johnston, Carrier, & Grigg, 2004). This enables a reconstruction of changes in relative forebrain size, and thus the NCL, either as a consequence of a sole increase in absolute forebrain size, a decrease in body size, or a combination of both (Smaers, Dechmann, Goswami, Soligo, & Safi, 2012).

In addition to knowing when these brain expansions occurred, it is interesting to understand evolutionary selection processes around encephalization events. Understanding these processes will aid in a better conception of the modern avian brain, and how it accomplishes its cognitive power. Specifically, since neural tissue is metabolically expensive, there should be clear benefits for survival, otherwise it will be strongly selected against (Niven & Laughlin, 2008b; Sol, 2009). Again, since it is not possible to assess distinct behaviours in extinct species, such as working memory or flexible planning, I will analyse this with the proxy of behavioural flexibility. Behavioural flexibility facilitates an adaptive response to unpredicted or novel social, climatic, or environmental challenges (Ricklefs, 2004; Sol, 2009), and it relies on individual cognitive domains such as working memory, inhibition and cognitive flexibility (Aboitiz, 2001). Behavioural flexibility predicts the rate of success of invading new territory (Sayol, Downing, Iwaniuk, Maspons, & Sol, 2018; Sol, Timmermans, & Lefebvre, 2002), and was demonstrated to be associated to a reduced risk of extinction in birds (Ducatez, Sol, Sayol, & Lefebvre, 2020). Moreover, rates of innovation, invasion success, and adaptability to climate and environmental variation were positively correlated to relative brain size (Fristoe, Iwaniuk, & Botero, 2017; Lefebvre, Reader, & Sol, 2004; Sayol, 2016; Sayol et al., 2018; Wagnon & Brown, 2020). And it is not simply a larger brain that facilitates behavioural flexibility, but I would argue a brain that is enlarged also due to a specific circuitry that facilitates (complex) cognition with the NCL/PFC as crucial component. Namely, to recapitulate, the NCL/PFC is the crucial structure that facilitates rapid learning, fast inhibition of behaviours that are not beneficial, switching between behavioural strategies, an abstraction process that can be applied to new situations (Güntürkün & Bugnyar, 2016; Güntürkün et al., 2017b). Thus, it seems that larger brains with an executive structure are successful in unpredictable

111

environments via increased behavioural flexibility (Wagnon & Brown, 2020). This is known as the cognitive-buffer hypothesis, which proposes that larger brains evolved in order to adapt to environmental variability (Allman, McLaughlin, & Hakeem, 1993; Sol, 2009). Recently, it was also shown that the last pulse of brain expansion in the avian lineage could be linked to ecological catastrophe (Ksepka et al., 2020), and I will argue that the need for increased behavioural flexibility was also the main driver for earlier pulses in brain size increase, and thus NCL expansion, in the archosaurian lineage.

The last requirement to detail the origin and evolution of the NCL in archosaurs necessitates evidence for significant physiological changes that occurred in order to generate enough energy to support neural tissue (Hammond & Diamond, 1997). This is known as the expensive brain hypothesis, which describes that an increase in brain size should be predated, or occur in parallel to, either an increase in metabolic rate, or reduction of other metabolically costly tissues or functions (Aiello & Wheeler, 1995; Isler & van Schaik, 2009). I will argue that in the archosaurian lineage, the former occurred in the shape of the development of endothermy which resulted in increased metabolic rates. Below I will first elaborate on the concept of endothermy, and how it relates to behavioural flexibility and encephalization. Next, I will detail the evolution of the NCL in the archosaurs on the basis of pulses of brain expansion that occurred in parallel to the evolution of endothermy and the need for behavioural flexibility to survive in a time of high climatic and environmental variability.

5.2.4 Endothermy, behaviour and the brain Endothermy is the maintenance of a stable body temperature above 30ºC, independent of ambient temperatures. A constant body temperature is the result of controlled internal heat production. It is contrasted with ectothermy, in which body temperature fluctuates and is dependent on external sources for heat. Endothermy evolved twice in the vertebrate lineage, and is present in all birds and mammals. This does not mean all birds and mammals are homeotherms, since heterothermy is present in species that go into hibernation or torpor. The avian and mammalian endothermy differs from other variants present in living organisms due to the source of the internal heat production. Most forms of endothermy are attained by muscular contraction, whereas in birds and mammals internal heat is in addition generated by organs (e.g. gut, kidney, heart, and brain). This type of ‘visceral endothermy’ allows the organism to maintain a constant body temperature even at rest (Nespolo, Bacigalupe, Figueroa, Koteja, & Opazo, 2011).

The endothermic maintenance of a stable body temperature comes at a high prize; birds and mammals have 5-10 higher basal metabolic rate compared to ectothermic reptiles of equal size and body temperature (Bennett & Ruben, 1979). Selection for such a costly physiology

112

must come with great benefits, and so far many explanations have been put forward. For example, it allows species to inhabit a wider range of geographical and ecological niches, it increases the reproductive success by enabling parental care and lactation, and a high and stable temperature enables higher efficiency of enzymatic processes (Koteja, 2004; Nespolo et al., 2011). A very prominent hypothesis positioned endothermy as a by-product of selection for increased aerobic capacity to facilitate higher levels of sustained activity (Bennett & Ruben, 1979). They identified that the evolution of endothermy went hand in hand with enhancement of oxygen uptake and transport, and an increased capacity of tissues to utilize oxygen. This enlarged aerobic scope tremendously increased the action radius and behavioural pattern complexity of endotherms. In concert, this required the development of a complex and encephalized nervous system, which in turn was enabled by the larger aerobic scope and increased daily metabolic rate (Balanoff, 2011; Lovegrove, 2017; Rowe et al., 2011). This demonstrates the interconnectedness between endothermy, behavioural repertoire and complexity of the nervous system, which can now be employed to reconstruct the evolution of the NCL (see Figure 23).

5.2.5 The origin and evolution of the NCL Based on the presence of an Ncl in the crocodile and expanded and diversified NCL in birds, I proposed that the stem archosaur probably had an NCL-like structure. Is there evidence for the presence of sufficient metabolic rates to sustain the additional neural tissue, and indicators for the need of an associative structure? Indeed, the first evidence of endothermy within the sauropsid lineage was found in stem archosaurs at the Permian-Triassic border. This was a time of ecological catastrophe and biodiversity crisis, caused by a variety of events including vulcanism and wildfires (Erwin, 1994; Shen et al., 2011). It wiped out 90% of marine species and 70% of the terrestrial vertebrates (Benton, Tverdokhlebov, & Surkov, 2004). Stem archosaurs were one of the surviving lineages, and archosaurian diversification in early Triassic occurred in a climate with a sharp drop in oxygen levels, rapid increase of carbon dioxide, and fast global warming (Shen et al., 2011). Temperature shifts up to 10 – 15 ºC occurred in several waves (Benton, 2020). During this period, some species reaped the benefits of increased body temperature through external warming of increased global temperatures. Because of the large size of basal archosaurs, external heat could sustain a stable body temperature, known as homeothermy, through thermal inertia (Lovegrove, 2017). In parallel, evidence of endothermy was found in the physiology of early Triassic archosaurs that indicated enhanced cursoriality (Kubo & Kubo, 2013), a four chambered heart (Seymour et al., 2004), increased growth rates (Cubo, Le Roy, Martinez-Maza, & Montes, 2012), thermoregulation (Dawson et al., 2020) and an improved gas exchange system to cope with decreasing oxygen level of early Triassic (Brocklehurst, Schachner, Codd, & Sellers, 2020;

113

Farmer & Sanders, 2010). As a side-note, this type of endothermy was generated by heat from muscular locomotion activity or (non-)shivering thermogenesis, thus not the visceral endothermy that characterizes modern birds and mammals. However, this all points towards enhanced aerobic activity facilitated by an increased metabolic rate (Legendre, Gucrossed D Signnard, Botha-Brink, & Cubo, 2016; Lovegrove, 2017).

Thus, there is evidence for a type of endothermy present in the ancestor of birds, and of crocodiles. Indeed, in contrast to other reptiles, the extant crocodile has a four chambered heart (Seymour, 2016; Seymour et al., 2004), unidirectional airflow (Farmer & Sanders, 2010), and the possibility of semi-erect motion (Reilly & Elias, 1998), which is all evidence for an endothermic physiology (Lovegrove, 2017). Interestingly, extant crocodilian species are ectothermic, with average body temperatures of 25-30 ºC, which they adjust through behavioural thermoregulation (Downs, Greaver, & Taylor, 2008; Seebacher, 1999). It has been suggested the crocodilian lineage reverted from an active foraging to a sit-and-wait predator, with concomitant reversal from endothermic state (Seymour et al., 2004). The remnants of a physiology for enhanced aerobicity is a reminder of the ancestral endothermy, and this is perhaps also the case for the brain. Namely, crocodiles and alligators have the largest brains of all reptiles, and in relative terms compare to basal birds and mammals (Northcutt, 2013). Moreover, this expansions is primarily driven by an enlargement of the DVR (Northcutt, 2013), where I demonstrated the existence of the tentative Ncl. Lastly, this all occurred in a period following the Permian-Triassic catastrophe marked by environmental turbulence that drove many species to extinction (Benton et al., 2004), and a type of behavioural flexibility could have been advantageous. This all would indicate that the basal archosaur, with enhanced aerobic activity and higher metabolic rate, had a relatively larger brain compared to the predecessor, with expanded ventral pallium and this possibly marks the origin of an NCL-like structure.

The next pulses of brain expansion occurred in late Triassic, long after the split of crocodiles and Aves. This time was again marked by environmental and climatic changes that globally affected flora and fauna, though considerably less dramatic compared to the end-Permian extinction event (Lucas & Tanner, 2018; Rigo et al., 2020). The most successful species in this time were found in the avian stem lineage, evidenced by a high rate of niche innovation. This was revealed by an analysis of evolutionary rates in body size diversification in the avian stem lineage, which is indicative of the rate of niche innovation. Benson et al. (2014) showed that whereas declining rates marked most non-avian Dinosaur lineages, feathered Maniraptora showed continued rapid evolution, culminating into the Paraves. This indicated that these taxa continued to exploit ecological innovation. Moreover, it is described as an important phase in the evolution of endothermy, driven by size miniaturization, insulation with feathers, and enhanced regulation of body temperature (Lovegrove, 2017). The miniaturization of body size

114

in the avian lineage took place over a span of 50 million years in early Jurassic. They derived from large bipedal Triassic theropods such as the ~163 kg Tetanurae, and show a directional trend in reduction of body size along the subsequent nodes, such as the Maniraptora (~9.8 kg) and the Paraves (~3.3 kg), up to the Avialae of ~0.8 kg (Benson et al., 2014; Lee, Cau, Naish, & Dyke, 2014; Rezende, Bacigalupe, Nespolo, & Bozinovic, 2020). Because inertial homeothermy was no longer possible for smaller-sized species, and global temperatures had dropped, a type of insulation must have been present. Indeed, body size miniaturization went hand in hand with increased insulation with feathers (Lovegrove, 2017). In concordance, the origin of simple feathers has been dated to either the base of Ornithodira (~250 mya, Yang et al., 2019), which includes Pterosauria and Dinosauria, immediately after the split with Crocodylia, and at the base of Maniraptora there is evidence for vaned feathers (~170 mya, Benton, Dhouailly, Jiang, & McNamara, 2019; Godefroit et al., 2014). The combination of bipedalism, enhanced activity patterns, and the emergence of insulation all point towards greater thermoregulatory capacities and higher metabolic rates (Lovegrove, 2017).

This period is marked by two pulses of neural tissue expansion of predominantly the telencephalon. The first pulse was noted for the Maniraptora, and an even larger pulse was found in the Paraves (Balanoff, Bever, Rowe, & Norell, 2013; Ksepka et al., 2020). This increase in EQ was accompanied by a doming of the skull and brain, which is visible as a curvature in the cranial shape (Beyrand et al., 2019). To elaborate, the ancestral archosaurian condition was a lissencephalic or linear brain, where , pallium, diencephalon, mesencephalon, cerebellum and medulla are organized along an almost straight rostral – caudal axis, and the spinal cord emerges caudally to the brain. This bauplan is observed in all non-avian reptiles (Walsh & Knoll, 2017). With the encephalization of the brain, and especially the telencephalon, the cranial form took on a more bulbous shape. In this lay-out, the cerebellum is situated immediately caudal to the forebrain, which bulges over mesencephalon and the medulla and spinal cord emerge ventrally. The first evidence of doming can be found in some Maniraptora, and even more so in the Paraves, correlating to the extent of encephalization (Beyrand et al., 2019). This is evidence for a concomitant growth of the NCL- like area. Indeed, the larger relative brain size can in part be explained be the reduction of body size, but this does not apply to the increased doming of the brain within the skull. This is a sole consequence of enlarged neural tissue of especially a relatively inflated (Beyrand et al., 2019). Since the NCL-territory is situated in the caudal back of the cerebrum, I suggest the pulses in Maniraptora and Paraves caused a concomitant enlargement of the NCL.

115

The shift from Paraves to the crown birds was not marked by a pulse in brain expansion. This occurred in the Cretaceous (~150/160 mya), and features the emergence of muscle-powered flapping flight with concomitant increase of basal metabolic rate (Brusatte et al., 2015; Lovegrove, 2017). This is considered the last phase of endothermy in the avian lineage as a consequence of the higher metabolic demands of pectoral flight muscles (Lovegrove, 2017; McNab, 1994). Further evidence is found in modern-day flightless birds that are characterized both by lower metabolic rates and lower body temperature (Clarke & Rothery, 2007; McNab, 1994). Thus, here the increase in metabolic rate was not a direct consequence of environmental disruption, but to facilitate flight (Lovegrove, 2017). The new form of locomotion did facilitate radiation into a wealth of new ecological niches (Rayner, 1988). This did not immediately result in an increase in neural tissue, which demonstrates flight was not a main driver of brain expansion (Balanoff, Smaers, & Turner, 2016). The increase in metabolic rate was, however, an important prerequisite for the emergence of larger brains to support the additional neural tissue.

The last major pulse of brain enlargement occurred in the Neoaves after the large KPg mass extinction event. Within the crown birds, different patterns of brain and body size changes were noted (Ksepka et al., 2020). To link the present findings of the expansion and diversification of the NCL in different bird species to specific brain expansion events, I will discuss the pattern for the chicken, pigeon, zebra finch, and crow. As mentioned in the introduction, three avian lineages survived the ecological catastrophe following the impact of an ~66 mya (Renne et al., 2013). After the event, especially the Neoaves underwent an explosive radiation, and this occurred with concomitant relative brain expansions in parallel in different orders (Ksepka et al., 2020). Since the major brain expansion events were noted after the Cretaceous- Tertiary border, and thus predominantly for late-emerging branches of the Neoaves, for the chicken and pigeon I expect no significant body and brain size changes. Indeed, Ksepka et al. (2020) demonstrated that at the base of the crown birds both body and brain size decreased, with body size decrease outpacing brain size reduction. This results in a relative brain size of Galliformes and Columbiformes that is slightly higher compared to the Paraves, but still within the range of non-avian therapods (Balanoff, Smaers, & Turner, 2016). The increase in brain size is also apparent in the extent of pallial doming, since the forebrain has a higher degree of curvature in the dorsal roof (Beyrand et al., 2019). It is likely the brain, and position and extent of the NCL, in chicken and pigeon represent the ancestral state of the avian brain at the base of the crown birds. The passerines, to which both zebra finch and corvids belong, show an increase in both body and brain size, with a steeper slope compared to the ancestral condition. The corvids are in addition characterized by a secondary increase in both body and brain size, where the expansion of brain outpaces growth of body size (Ksepka et al., 2020). This brain

116

expansion is primarily a consequence of enlargement of the meso- and nidopallium (Cnotka, Güntürkün, Rehkämper, Gray, & Hunt, 2008; Mehlhorn, Hunt, Gray, Rehkämper, & Güntürkün, 2010). As shown in Chapter 3, this expansion was concomitant to an enlarged and diversified NCL in zebra finch and, especially, the crow.

Next to an expansion, the caudal forebrain of the songbirds shows a ‘rotation’, which caused a shift in the location of both the arcopallium and the NCL in passerines (Mello et al., 2019; Chapter 3). One possible explanation can be found in the ‘motor theory of origin’ (Feenders et al., 2008). Zebra finch and crows belong to the group of oscine passerines, also known as songbirds, and are characterized by the capacity to produce and modify song. They are one of three orders of birds that are capable of vocal learning, the other being parrots and hummingbirds (Nottebohm, 1972). In all three orders, vocal production and vocal learning is accomplished by distinct nuclei and pathways that came about by convergent evolution (Petkov & Jarvis, 2012). The motor theory of vocal learning proposed that the song circuitry stems from ancestral motor production pathways through brain pathway duplication. After duplication, the pathway connects to brainstem efferents that control respiration and vocalizations (Chakraborty & Jarvis, 2015; Feenders et al., 2008). This theory is furthermore supported by a similar genetic make-up in song and motor circuitry (Pfenning et al., 2014). Thus, in the songbird lineage, part of the expansion can be explained by pathway duplication of existing motor circuits to facilitate song learning and production. Importantly, these new song pathways are situated immediately adjacent to existing motor pathways (Feenders et al., 2008). Perhaps, this caused the lateral-caudal push of nido- and mesopallial territory to stay within reasonable brain and skull morphology, comparable to the doming effects of hypertrophied cerebrum (Beyrand et al., 2019). Indeed, like oscine passerines, the other vocal learners parrots are also characterized by relatively larger brains, of a similar cerebrotype, with high neuronal densities (Iwaniuk & Hurd, 2005; Ksepka et al., 2020; Olkowicz et al., 2016). Moreover, in oscine species where females produces less or no complex song patterns, males have relatively larger brains (Garamszegi, Eens, Erritzøe, & Møller, 2005). It is currently not known how the additional vocal pathways influenced the gross morphological lay-out of the forebrain, and especially the location of the NCL. As discussed in Chapter 3, one identified sub-area of the NCL in the zebra finch and crow is the NCM, a well-known higher order auditory area in songbirds (Pinaud & Terleph, 2008). This region has not been identified in pigeon or chicken (Martin Wild et al., 1993; Martin Metzger et al., 1998), which indicates this part of the NCL could be a songbird-specific higher order auditory area. Since the NCM is situated immediately adjacent to the medial border, it possibly pushed other caudal nidopallial territory more lateral.

117

Thus, the origin of the NCL can be placed in an active homeothermic basal archosaur, just after the PT mass extinction event around 250 mya. From here, the brain remained relatively unchanged, or at most decreased in size, within the crocodilian lineage. The extant crocodile has a small Ncl, which matches its sit-and-wait ectothermic lifestyle. From the basal archosaur, two major brain expansions occurred that were set off following the end-Triassic extinction event. One occurred within the Maniraptora, which showed a high degree of niche expansion, the second within the Paraves. This occurred in parallel to the second phase of endothermy characterized by miniaturization and insulation. During these brain expansions, the NCL enlarged concomitantly from the small area as present in the crocodile brain, up to what is found in chicken and pigeon. The last pulse occurred within the Neoaves, following the KPg- extinction that wiped out all non-avian dinosaurs. Especially the Passeriformes were characterized by a significant body and brain enlargement, with in the Corvidae brain size expansion outpacing body enlargement. This translated into an NCL that spans across the back of the forebrain, and diversified in several functional subdivisions.

Figure 23 Simplified overview of the origin and evolution of the NCL within the archosaurian lineage. The origin and evolution of the NCL can be reconstructed using three behavioural proxies: brain expansion, evolution of endothermy, and the need for behavioural flexibility. The origin of the NCL is dated to the stem archosaur ~245 mya, just after the mass extinction event at the Permian-Triassic border. These circumstances gave rise to enhanced metabolic rates and a concomitant larger lissencephalic brain than the predecessor with an expanded ventral pallium and NCL-like structure, comparable to what we observe in crocodiles today. After the split, Ornithodira show evidence of simple feathers, and bipedalism arose in the Dinosauria, both evidence of increasing endothermy and metabolic rates. The next two pulse of brain expansion occurred in the Late Triassic in Maniraptora (~178 mya) and the Paraves (~168 mya). This was again a period of environmental turbulence, characterized by size miniaturization, the emergence of feathers and a doming of the forebrain. The origin of the crown birds occurred around the emergence of flight ~150 mya. Within the crown birds main brain expansion occurred only after the KP-extinction event ~66 mya. The Galloanserae branched off before and this clade has relatively smaller brains with a small and undifferentiated NCL. Next, Columbaves branched off early in the Neoaves ~64 mya, and they have slight relatively larger brains, with an NCL that is highly comparable to the chicken. The Passeriformes are marked by a significant brain expansion, and this translates in the zebra finch and carrion crow into an expanded and more parcellated NCL. Phylogeny and emergence of flight based on Brusette et al (2015) and Nesbitt et al. (2011). Brain endocasts are from Beyrand et al (2019), and species silhouettes from phylopic.org.

118

119

5.3 The brain in a flying body

What becomes apparent from this evolutionary reconstruction of the NCL, is that pulses of brain expansion occurred concomitantly or shortly after increases in metabolic rate. As mentioned, this can be explained by the expensive brain hypothesis, which elaborates how an increase in neural tissue should occur in parallel to either an increase in metabolic rate or a decrease in other energetically costly organs or functions (Aiello & Wheeler, 1995; Isler & van Schaik, 2006a; Navarrete, Van Schaik, & Isler, 2011). I outlined this correlation in ancestral species, but this hypothesis is in part also supported in extant mammals and birds. In mammals, a significant relationship was found between relative brain size and basal metabolic rate (Isler & van Schaik, 2006b; Isler & Van Schaik, 2009). Interestingly, birds show higher basal metabolic rates than mammals (Clarke & Pörtner, 2010; White et al., 2006), and in addition, I demonstrated in the current research that avian neural tissue is per neuron more than three times more energy-efficient than in mammals. Yet, after more than 300 million years of separate evolution, no avian species currently exists that maintains a brain with three times more neurons than a human brain nor attained a volume larger than 46 mm3 (Herculano- Houzel, 2009; Olkowicz et al., 2016). Thus, there must be limiting factors that place a ceiling on neuron densities and brain growth within the avian lineage.

Before I can discuss some of these limiting factors, it is important to keep in mind that what is ‘cheap’ for a mammal, might still be costly for a bird. This is clearly apparent in songbirds, who demonstrate one of the most impressive types of neural plasticity among adult vertebrates. Namely, depending on the season, the volume of their song nuclei can vary drastically (Tramontin & Brenowitz, 2000). The most extreme example comes from the spotted towhee (Pipilo maculatus), whose HVC (a song nucleus important for learning and production of song) will increase 300% in size when the breeding season is approaching (Smith, 1996). Accordingly, during winter the HVC reduces both in neuron number as well as in neuron size (Smith, 1996), which are efficient strategies for saving energy (Sengupta et al., 2013, 2010). These strategies have also been demonstrated in brain areas that are not involved in song learning or production. The black-capped chickadee (Poecile atricapillus) shows seasonal variation in the size of the hippocampus driven by variation in neuron number (Smulders, Sasson, & DeVoogd, 1995; Smulders, Shiflett, Sperling, & Devoogd, 2000). The hippocampus is involved in spatial cognition, and in this species plays a specific role in seasonal food- hoarding behaviour. This is evidence that changes in size and neuron numbers of specific nuclei are a more avian-general strategy to reduce the costs of the brain (but see Pozner, Vistoropsky, Moaraf, Heiblum, & Barnea, 2018 for discussion of the different causes of ). What this shows is that neurons are not a ‘free for all’ commodity for birds. What could be the limiting factor on brain size and number of neurons?

120

There are multiple answers to this question; the evolutionary history of birds is long and complex, and for the wide variety of approximately 10,000 bird species that live today it is possible to point at different constraining, and driving, factors on brain size and neuron density (Willemet, 2013). It is unfortunately not possible to discuss all elements, and in this thesis I will highlight the capacity for flight. I think this bird-specific form of locomotion is of special significance to understand the avian brain, and it limits in at least two ways. First, flight poses a constraint on size and weight, since every additional gram will make it more energetically costly to fly (Vincze, 2016). The avian body has several weight-reducing strategies in place, such as postcranial skeletal pneumaticity, which is the presence of spaces filled with air within bones (Wedel, 2005). In the same vein, a cap on brain growth would limit overall weight, and this is additionally important for a flying species that cannot be too top-heavy. Second, flight is one of the most metabolically demanding forms of locomotion across all vertebrates. Compared to similarly-sized mammals who run at the maximum of their sustainable speed, forward flapping flight is 2.2 times more costly (Patrick J. Butler, 2016). Additional evidence for how flight poses a constraint on brain growth comes from the size of the flight muscles. Namely, relative brain size scales negatively with the size of pectoral muscles (Isler & van Schaik, 2006a). These muscles are most important for powered take-off during flight, and, depending on the species, make up 10 – 35% of body weight (Butler, 1991). There are several perspectives on how these types of correlations could be interpreted (see Morand-Ferron, Cole, & Quinn, 2016), but one suggestion is that this might indicate that avian species with an energetically costly, or energetically inefficient, flight style have a stronger limitation on brain size (Isler & van Schaik, 2006a). An example of an energetically costly flight style is long- distance migration, which can be up to 11,000 km in one stretch (Gill et al., 2009). Indeed, there is a robust negative correlation between relative brain size and flight distance during migration (Vincze, 2016). An example of an energetically inefficient flight style are the Galliformes, who fly in small bursts over short distances facilitated almost exclusively by fast glycolytic muscle fibres that produce ATP via anaerobic glycolysis (Butler, 2016). In accordance, the brain of Galliformes has a lower encephalization quotient compared to other birds (Boire & Baron, 1994), and consists of low neuron numbers (Olkowicz et al., 2016). For example, the red junglefowl (Gallus gallus) is 2.7 times bigger than the pigeon (Columba livia), but has a brain of only 1.4 times its size and still contains 30% fewer neurons in absolute terms (Olkowicz et al., 2016). This can in part be explained by a phylogenetic effect. Different orders adhere to a taxon specific brain to body scaling relationship (Nealen & Ricklefs, 2001), and this might also influence neuron densities, as observed in mammals (Herculano-Houzel et al., 2015). However, there is additional evidence that flight style influences brain size and neuron density, both as a constraining, as well as a driving factor.

121

Across the avian lineage, a wide variety of flight styles can be identified, which can be causally associated with the elaboration of different locomotor modules (sensu Gatesy & Dial, 1996). Flightless birds, such as the emu, and poor flyers, such as the chicken, have stronger developed hindlimb muscles that consist of oxidative muscle fibres. In contrast, their pectoral muscles are either drastically reduced, as in the emu, or large but consisting of fast glycolytic muscle fibres, as is the case for the chicken. The fast glycolytic fibres facilitate short-burst flight, but can also be found in large bodied migratory avian species with a poor power-to-mass ratio to enable take-off during flight (Dial, 2003; Gatesy & Dial, 1996). It is interesting that exactly flightless birds or energy inefficient flyers have relatively small brains with low neuron densities (Olkowicz et al., 2016). On the other side of the spectrum are the Passeriformes. These make up the largest group of the avian lineage and display a wide variety of flight styles, though most have strongly developed flight muscles and a low mass-to-power ratio to enable forward flapping flight (Dial, 2003). This is an extremely costly flight style, but powered by aerobic respiration in fast oxidative muscles (Butler, 2016; Welch & Altshuler, 2009). In concordance, they have significantly higher basal metabolic rates, and warmer body temperatures (McNab, 2009). Importantly, Passeriformes have strongly encephalized brains of higher neuron densities compared to non-passerines (Olkowicz et al., 2016). A similar combination of large pectoral muscles with oxidative fibres, high metabolic rates, warm body temperature, and large brains of high neuron densities also characterizes the Psittaciformes (McNab & Salisbury, 1995; Rosser & George, 1986). The crucial difference is that, in contrast to for example the chicken, passerines and parrots have converged on higher aerobicity, which is locomotion geared towards sustained activity (Meléndez-Morales, de Paz-Lugo, & Meléndez-Hevia, 2009; Nespolo et al., 2018). In order to support such an endurance physiology, major changes occurred within the passerine body, and in all birds that have sustained forward flapping flight. The physiology of all bird species, and in contrast to mammals, is characterised by important adaptations in the respiratory and cardiovascular system. To name a few, birds have more efficient ventilation due to unidirectional lungs with associated air sacs, which facilitates an exchange of oxygen and carbon dioxide both during inhalation and exhalation. Moreover, the specific arrangement of air and blood capillaries creates a large surface area, and the blood-gas barrier is extremely thin, both of which enhance diffusion. Compared to non-passerines, passerines have an even higher oxygen diffusing capacity and rate of oxygen consumption (Grubb, 1983; Lasiewski & Calder, 1971; Maina, 1984). In terms of the cardiovascular system, birds have relatively larger hearts with a larger stroke volume compared to similarly-sized mammals (Faraci, 1991). These physiological changes facilitate a high aerobic capacity which enables sustained flight (Bishop, 2005; Nespolo et al., 2018), and I would argue that this endurance physiology also influenced brain energetics. As discussed extensively in Chapter 4, in order to sustain neurons, even the ones

122

that are at rest, large amounts of glucose and oxygen are needed (Mink et al., 1981). Importantly, neural energy metabolism is predominantly aerobic with 95% of the glucose being metabolized oxidatively (Mrozek, Vardon, & Geeraerts, 2012). Therefore, adaptations within the cardio-respiratory physiology to obtain and transport larger amounts of oxygen will also have effect at the cerebral and neuronal level. I suggest this can partially explain how the Passeriformes and Psittaciformes can sustain such high neuron densities compared to non- Passeriformes. Moreover, a physiology towards endurance also plays out at a behavioural and cognitive level. Namely, an increased aerobic capacity will enlarge a species’ action radius, and influence the complexity of its behavioural repertoire (Bennett & Ruben, 1979). Thus again, this is not about ‘just’ a larger brain with more neurons, but a brain that includes an executive structure that can keep track and orchestrate behaviour over time (Güntürkün, 2012). There is also an interesting parallel to mammals here related to the relationship between large brains with a high neuron density and aerobic energetics. Namely, Homo sapiens have a significantly higher metabolic rate compared to other great apes (Pontzer et al., 2016). Moreover, the human neocortex shows a specific upregulation of gene expression linked to aerobic energy metabolism (Uddin et al., 2004), and these changes in the metabolome are particularly substantial in the prefrontal cortex (Bozek et al., 2014; Fu et al., 2011). Some of these findings have been explicitly linked to differences in locomotion between humans and great apes (Sockol, Raichlen, & Pontzer, 2007); with the advent of bipedalism, Homo sapiens developed a more economic form of locomotion, reliant on endurance, which provided a surplus of energy available for other tissues, such as the brain (Pontzer, 2017).

This type of evolutionary reconstructions on the basis of traits present in extant species are a debatable approach, and it will be a challenge to gather empirical support that flight was the main driver, and neural tissue came second. Moreover, it should be considered in the light of an intricate interplay of development of an aerobic physiology, emergence of flight, increasing body temperatures, and expansion of neural tissue that occurred in tandem over millions of years. Thus, exactly how differences in energetics metabolism, flight style, brain size and neuron density are interconnected within the avian clade requires additional research. Especially considering the wide variety of bird species that exist, who differ tremendously in morphology, physiology, ecology, and behaviour. Moreover, it is important to consider additional factors such as, but not limited to, diet and feeding behaviour, breeding style and parental investment, and phylogeny (Bennett & Harvey, 2009; Isler & Van Schaik, 2009; Ksepka et al., 2020; Willemet, 2013). However, these parallel lines of evidence in birds and mammals emphasize the role of sustained activity, and the development of a physiology for oxidative phosphorylation, in brain expansion.

123

Thus, even though the avian neuron uses 3.25 times less glucose compared to a neuron in a mammalian brain, it is still costly to expand neural tissue and especially increase the number of neurons. One of the main limiting factors appears to be flight, both in the need to remain small, and in the energy budget allocation to either locomotion or brain tissue. Simultaneously, the evolution of especially aerobic energetics to facilitate flight could enable additional brain expansion and high numbers of neurons, as evident in the Passeriformes and Psittaciformes. I would even suggest that without flight, the bird brain would not have been able to have attained the impressive size and neuron density that we observe in crows and parrots. This all shows that when one wants to understand how the bird brain can achieve such impressive cognitive feats, it is important to keep in mind it is a cerebral structure in a body that flies, that is warm, and that is small.

5.4 Conclusion and future directions

The aim of this thesis was to gain a better understanding of the underlying building blocks of avian complex cognition by specifically evaluating two phenomena: the executive caudal nidopallium and the neuronal energy budget. The research presented here demonstrated the existence of an executive NCL-like structure in the crocodile, and on the other end of the spectrum, a highly expanded and parcellated NCL in the Passeriformes. These findings allowed for a reconstruction of the origin and evolution of the NCL. This posits the NCL first emerged in the basal archosaur ~245 mya, and evolved in concert to several brain expansions that occurred in parallel to the evolution of endothermy and the benefits of enhanced behavioural flexibility. This reconstruction places the capacity for sustained flight in a key position in the development of a large and neuron-dense brain, as observed in crows and parrots. Possibly, flight might have been a driving factor in the development of an aerobic physiology, and facilitated an increased action radius and more complex behavioural repertoire. In parallel, this required a highly developed nervous system, which in turn could be sustained by increased metabolic rates. In addition, this thesis showed that neurons in the avian brain are three times more energy efficient compared to mammals, which further elucidates how birds have been able to pack such high numbers of neurons in their brains. The exact mechanism how this efficiency is attained is currently not fully understood. Thus, this thesis showed that complex cognition in the small avian brain is attained by an expanded and diversified executive caudal nidopallium that is high in energy-efficient neuron numbers, and it posits flight as one of the key factors.

124

5.4.1 The executive structure in sauropsids From the evolutionary reconstruction of the NCL in the archosaurian lineage, some glaring gaps remain to be filled. Most importantly would be to investigate the NCL in a member of the Paleognathae. This branch diverged from the Neognathae around 110 mya (Figure 23), and includes several extant branches such as the ostrich, tinamou, and kiwi (Prum et al., 2015). These species are situated in an exciting position from an evolutionary perspective, since they are the oldest avian species branching between Crocodylia and Neognathae. In addition, according to Ksepka et al. (2020), Paleognathae did not undergo a substantial brain expansion like the Neoaves. Therefore, I would expect a brain similar to chicken, but relatively smaller and with a small and undifferentiated NCL. On the other side of the evolutionary spectrum, it is important to get a better understanding of the different subdivisions of the NCL in the Passeriformes. The PFC of primates is parcellated in additional areas that each have a distinct functional role (Fuster, 2015), and it is likely something comparable is at play in the songbirds. My research identified at least four different subdivisions, of which only two are currently understood to a reasonable degree, namely NCM and NCLv. I would propose that the first step is an extensive tract-tracing analysis, comparable to Kröner & Güntürkün (1999), to disentangle the functional properties of the different areas. Next, based on the revealed connectivity, it is possible to design more targeted behavioural experiments for electrophysiological recordings. This would extend the work of Nieder and colleagues (Nieder, 2017; Nieder, Wagener, & Rinnert, 2020; Rinnert et al., 2019; Veit & Nieder, 2013; Wagener, Loconsole, Ditz, & Nieder, 2018) with the aim of going beyond the NCLv. I think a similar approach in the crocodilian brain would be extremely valuable. In addition, a crucial first step in this species would be to design a working memory task and test the limits and capacities of the crocodile in this domain. Next, if there are positive results, it would be interesting to uncover whether there is a subpopulation of neurons situated in the area delineated in this thesis, which sustains their activity specifically during the delay period (Bettina Diekamp, Kalt, & Güntürkün, 2002b). As this is one of the hallmark tests of an executive area (Hahn & Rose, 2020).

5.4.2 Avian neuronal energy budget First, is important to verify whether the low neuronal energy budget also applies to other bird species, and whether this can be described by a general rate of consumption or whether it diverges depending on the species. Ideally, the species of interest represent diverse branches of the avian tree, and vary in brain volume, body size, and flight mode. Thus, it would be important to include species of the Paleognathae, such as the flightless emu with a large body, relatively small brain and low neuron densities. Furthermore, chickens would be a suitable representative of another more basal species, with a large body, a relatively small brain, and low neuron density (Olkowicz et al., 2016; Rehkämper et al., 1991). This could be contrasted

125

with a type of goose, who like chicken belong to the Galliformes, but display long-distance migratory flights, and have large pectoral muscles with a mixture of fast oxidative and glycolytic muscle fibres (Gatesy & Dial, 1996; Prum et al., 2015; Rosser & George, 1986). From the other end of the spectrum it would be important to include members of the Passeriformes and Psittaciformes because of the more encephalized brains with high neuron densities and the particular forward flapping flight style facilitated by oxidative muscles fibres (Kilgore et al., 1976; Olkowicz et al., 2016; Rosser & George, 1986). From both orders it would be important to study both a small-bodied species such as the zebra finch or budgerigar (Melopsittacus undulatus), and a large-bodied species such as the carrion crow or grey parrot (Psittacus erithacus) to disentangle the effect of body size. Moreover, compared to the other species of interest, songbirds and parrots have higher body temperatures (McNab, 1966). If indeed temperature is a decisive factor that drives down the neuronal energy costs, the passerines and psittacines would have an even lower neuronal energy budget compared to the pigeon. This could be further corroborated by including an ectotherm, such as the crocodile, into the analysis. Here, I would expect that the costs per neuron are significantly higher. Moreover, since the crocodile is a endotherm-turned-ectotherm (Seymour et al., 2004) with an endothermic-like physiology, it would be interesting to also study another reptile such as a member of the Lepidosauria or Testudines, and get a more complete overview of the interplay between body temperature, metabolic rate, and neuronal energy budget.

This series of experiments will clarify in greater detail the extent to which neuronal energy budgets are the same or vary across the sauropsid lineage, but it will not explain how the energy costs of the neuron are kept so low in birds. In Chapter 4, I proposed three different possible mechanisms, one is related to temperature, and the other two are specifically related to the nuclear organization of the avian brain. The question of temperature might in part be answered by comparing avian species with a different body temperature. However, this will solely show a correlation, and we are most interested in the causal relationship. A modelling study showed that temperature reduces the overlap between outward Na+-current and inward K+-current by influencing channel activation times (Sengupta et al., 2010; Yu et al., 2012). This overlap can be deduced from the action potential waveform as measured by whole-cell patch clamp recordings (Shu, Yu, Yang, & McCormick, 2007). It would be important to analyse channel kinetics in, for example, both the chicken and the songbird in order to study the effect of temperature within the same phylogenetic class.

Investigating how nuclear organization influences energy efficiency in terms of saving wire and/or myelination of short path lengths requires diligent anatomical analyses. Ideally, one would quantify soma size, axon length, axon diameter, percentage of myelination, number of synapses, etc. However, this would be extremely time-costly and labour-intensive. This type

126

of full quantification was attempted for the cerebral cortex, and it would take an experienced tracer 15 minutes to segment 1 µm3 into separate quantifiable cell profiles (Kasthuri et al., 2015). This means it would take one experienced researcher 55 million years to segment one complete pigeon brain. Alternatively, a subpopulation of specifically axon length and axon diameter can be assessed following the method of Innocenti et al. (2014), where subjects are injected with anterograde and retrograde tracers in different areas of the brain. The two axonal parameters can be derived with light and electron microscopy. In the monkey brain, axon diameter and length differ substantially between both area of origin and target area in the monkey brain (Tomasi, Caminiti, & Innocenti, 2012), thus it would be important to analyse tracings from different areas of the avian brain. Since some differences between species have been noted across mammals (Caminiti, Ghaziri, Galuske, Hof, & Innocenti, 2009), it would be important to assess different bird species also. Alternatively, it is possible to measure the axon density and distribution of axon diameter, and even quantify neurite density, orientation and dispersion using diffusion MRI (Barazany, Basser, & Assaf, 2009; Genç et al., 2018). Noting the great strides MRI research in pigeons has made in the past couple of years (Behroozi, 2018), this might be a more promising avenue for assessing axon parameters such as wiring efficiency in the bird. Moreover, using diffusion MRI might also aid in quantifying the percentage of fibres that is myelinated, and assess whether this applies to a substantial amount of short-range fibres.

This still leaves two major questions concerning evolutionary causation. This is a controversial type of question that leaves little room for falsifiable hypotheses, but is nevertheless all the more interesting. The first question is related to the temperature of the bird brain: why did the bird brain get so much warmer compared to the mammalian brain? It would be important to understand what key events in the evolution of birds, in contrast to mammals, resulted in the end in these higher body and brain temperatures. For example, a significant difference exists in the respiratory systems of birds and mammals, producing the observed differences in metabolic rate and aerobic respiration. As mentioned, the archosaurs have unidirectional lungs that facilitate larger amounts and more efficient oxygen uptake. The question would be, what were possible environmental pressures around the time these lungs evolved that resulted in this particular design? A possible answer could be related to atmospheric oxygen concentrations that varied drastically across the Mesozoic (Hudgins, Uhen, & Hinnov, 2020). The evolution of an efficient respiratory system was beneficial in itself, but it might also have had far-reaching consequences for the later development of endothermy, aerobic respiration, and in the end the neuron-dense brain of birds.

127

The second question relates to the nuclear organization of the avian pallium. As detailed in the introduction, the largest part of the forebrain of birds is nucleated as opposed to the laminar structure that dominates the forebrain of mammals. These territories are functionally analogous, but came about by convergent evolution. Interestingly, the DVR of birds derived from the ventral pallium, whereas the neocortex in mammals developed from the dorsal pallium (Jarvis, 2009; Puelles et al., 2017). What intrigues me here is the question: why did the ventral pallium enlarge in the bird? Again, to answer this question, it would be necessary to reconstruct the circumstances under which the ventral pallium started to increase. This would entail getting an understanding of the basal species from which the avian lineage derived, and constructing an overview of the environmental factors that posed driving or constraining factors at the time. Of course, these changes occurred over a time-line of millions of years, which make it nearly impossible to fabricate such a reconstruction, but therefore all the more interesting. A possible explanation that has been put forward is that basal sauropsids and synapsids relied on different sensory systems that consequently organized in diverging ways within the forebrain (Balanoff et al., 2013). In short, there were so many steps in the evolutionary history of mammals and birds that evolved in parallel into something similar, but even more interesting are those steps where something evolved differently. Next, identifying this initial different step allows a possible reconstruction of a whole chain of events that in the end created a bird, and not a parallel mammal. To conclude, how does the birdbrain churn out such cognitive prowess? Because it is a brain in a bird body.

128

129

130

References

131

132

Abellán, A., Desfilis, E., & Medina, L. (2014). Combinatorial expression of Lef1, Lhx2, Lhx5, Lhx9, Lmo3, Lmo4, and Prox1 helps to identify comparable subdivisions in the developing hippocampal formation of mouse and chicken. Frontiers in Neuroanatomy, 8, 59. https://doi.org/10.3389/fnana.2014.00059

Aboitiz, F. (2001). What determines evolutionary brain growth? Behavioral and Brain Sciences, 24(2), 278–279.

Abumrad, N. A., Briscoe, P., Beth, A. H., & Whitesell, R. R. (1988). Temperature dependence of glucose transport in erythrocytes from normal and alloxan-diabetic rats. BBA - Biomembranes, 938(2), 222–230. https://doi.org/10.1016/0005-2736(88)90161-7

Agnolín, F. L., Egli, F. B., Chatterjee, S., Marsà, J. A. G., & Novas, F. E. (2017). Vegaviidae, a new clade of southern diving birds that survived the K/T boundary. The Science of Nature, 104(87). https://doi.org/10.1007/s00114-017-1508-y

Aiello, L. C., & Wheeler, P. (1995). The Expensive-Tissue Hypothesis: The Brain and the Digestive System in Human and Primate Evolution. Current Anthropology, 36(2), 199–221. https://doi.org/10.1086/204350

Alf, M. F., Duarte, J. M. N., Lei, H., Krämer, S. D., Mlynarik, V., Schibli, R., & Gruetter, R. (2014). MRS glucose mapping and PET joining forces: re-evaluation of the lumped constant in the rat brain under isoflurane anaesthesia. Journal of Neurochemistry, 129(4), 672–682. https://doi.org/10.1111/jnc.12667

Alf, M. F., Martić-Kehl, M. I., Schibli, R., & Krämer, S. D. (2013). FDG kinetic modeling in small rodent brain PET: Optimization of data acquisition and analysis. EJNMMI Research, 3(1), 61. https://doi.org/10.1186/2191- 219X-3-61

Alf, M. F., Wyss, M. T., Buck, A., Weber, B., Schibli, R., & Krämer, S. D. (2013). Quantification of brain glucose metabolism by18F-FDG PET with real-time arterial and image-derived input function in mice. Journal of Nuclear Medicine, 54(1), 132–138. https://doi.org/10.2967/jnumed.112.107474

Alföldi, P., Rubicsek, G., Cserni, G., & Obál, F. (1990). Brain and core temperatures and peripheral vasomotion during sleep and wakefulness at various ambient temperatures in the rat. Pflügers Archiv European Journal of Physiology, 417(3), 336–341. https://doi.org/10.1007/BF00371001

Alle, H., Roth, A., & Geiger, J. R. P. (2009). Energy-efficient action potentials in hippocampal mossy fibers. Science, 325(5946), 1405–1408. https://doi.org/10.1126/science.1174331

Allman, J., McLaughlin, T., & Hakeem, A. (1993). Brain weight and life-span in primate species. Proceedings of the National Academy of Sciences of the United States of America, 90(1), 118–122. https://doi.org/10.1073/pnas.90.1.118

Alstrup, A. K. O., & Smith, D. F. (2013). Anaesthesia for positron emission tomography scanning of animal brains. Laboratory Animals, 47(1), 12–18. https://doi.org/10.1258/la.2012.011173

Andreu, M., Dávila, J., Real, M., & Guirado, S. (1996). Intrinsic connections in the anterior dorsal ventricular ridge of the lizard Psammodromus algirus. The Journal of Comparative Neurology, 372(1). https://doi.org/10.1002/(SICI)1096-9861(19960812)372:1<49::AID-CNE5>3.0.CO;2-M

Andreu, M. J., Dávila, J. C., de la Calle, A., & Guirado, S. (1994). Monoaminergic Innervation Patterns in the Anterior Dorsal Ventricular Ridge of a Lacertid Lizard, Psammodromus algirus; Brain, Behavior and Evolution, 44(3), 175–186. https://doi.org/10.1159/000113589

133

Appeltants, D., Ball, G. F., & Balthazart, J. (2001). The distribution of tyrosine hydroxylase in the canary brain: Demonstration of a specific and sexually dimorphic catecholaminergic innervation of the telencephalic song control nuclei. Cell and Tissue Research, 304(2), 237–259. https://doi.org/10.1007/s004410100360

Atoji, Y., & Wild, J. M. (2009). Afferent and efferent projections of the central caudal nidopallium in the pigeon (Columba livia). Journal of Comparative Neurology, 517(3), 350–370. https://doi.org/10.1002/cne.22146

Attwell, D., & Laughlin, S. B. (2001). An energy budget for signaling in the grey matter of the brain. Journal of Cerebral Blood Flow and Metabolism, 21, 1133–1145. https://doi.org/10.1097/00004647-200110000-00001

Auersperg, A. M. I., von Bayern, A. M. P., Gajdon, G. K., Huber, L., & Kacelnik, A. (2011). Flexibility in problem solving and tool use of kea and new caledonian crows in a multi access box paradigm. PLoS ONE, 6(6). https://doi.org/10.1371/journal.pone.0020231

Babb, S. J., & Crystal, J. D. (2006a). Discrimination of what, when, and where is not based on time of day. Learning & Behavior, 34(2), 124–130. https://doi.org/10.3758/BF03193188

Babb, S. J., & Crystal, J. D. (2006b). Episodic-like Memory in the Rat. Current Biology, 16(13), 1317–1321. https://doi.org/10.1016/j.cub.2006.05.025

Backes, H., Walberer, M., Endepols, H., Neumaier, B., Graf, R., Wienhard, K., & Mies, G. (2011). Whiskers Area as Extracerebral Reference Tissue for Quantification of Rat Brain Metabolism Using 18F-FDG PET: Application to Focal Cerebral Ischemia. Journal of Nuclear Medicine, 52(8), 1252–1260. https://doi.org/10.2967/JNUMED.110.085266

Baddeley, A. D., & Hitch, G. (1974). Working Memory. Psychology of Learning and Motivation, 8, 47–89. https://doi.org/10.1016/S0079-7421(08)60452-1

Balaban, C. D., & Ulinski, P. S. (1981). Organization of thalamic afferents to anterior dorsal ventricular ridge in turtles. I. Projections of thalamic nuclei. The Journal of Comparative Neurology, 200(1), 95–129. https://doi.org/10.1002/cne.902000108

Balakhonov, D., & Rose, J. (2017). Crows Rival Monkeys in Cognitive Capacity. Scientific Reports, 7(1), 1–8. https://doi.org/10.1038/s41598-017-09400-0

Balanoff, A. (2011). Oviraptorosauria: Morphology, Phylogeny, and Endocranial Evolution. Columbia University, New York.

Balanoff, A. M., Bever, G. S., Rowe, T. B., & Norell, M. A. (2013). Evolutionary origins of the avian brain. Nature, 501(7465), 93–96. https://doi.org/10.1038/nature12424

Balanoff, A. M., Smaers, J. B., & Turner, A. H. (2016). Brain modularity across the theropod-bird transition: testing the influence of flight on neuroanatomical variation. Journal of Anatomy, 229(2), 204–214. https://doi.org/10.1111/joa.12403

Barazany, D., Basser, P. J., & Assaf, Y. (2009). In vivo measurement of axon diameter distribution in the corpus callosum of rat brain. Brain, 132(5), 1210–1220. https://doi.org/10.1093/brain/awp042

Barr, H. J., Wall, E. M., & Woolley, S. C. (2019). Dopamine in the songbird auditory cortex shapes auditory preference. BioRxiv, 761783. https://doi.org/10.1101/761783

134

Barton, R. A., & Harvey, P. H. (2000). Mosaic evolution of brain structure in mammals. Nature, 405(6790), 1055– 1058. https://doi.org/10.1038/35016580

Bast, T., Diekamp, B., Thiel, C., Schwarting, R. K. W., & Güntürkün, O. (2002). Functional aspects of dopamine metabolism in the putative prefrontal cortex analogue and striatum of pigeons ( Columba livia ). Journal of Comparative Neurology, 446(1), 58–67. https://doi.org/10.1002/cne.10187

Bateson, P. P. G. (1966). The characteristics and context of imprinting. Biological Reviews, 41(2), 177–217. https://doi.org/10.1111/j.1469-185X.1966.tb01489.x

Behroozi, M. (2018). Establishing a novel fMRI approach to investigate visual cognitive properties in pigeons. Ruhr-University Bochum, Bochum.

Behroozi, M., Billings, B. K., Helluy, X., Manger, P. R., Güntürkün, O., & Ströckens, F. (2018). Functional MRI in the Nile crocodile: a new avenue for evolutionary neurobiology. Proceedings of the Royal Society B: Biological Sciences, 285(1877), 20180178. https://doi.org/10.1098/rspb.2018.0178

Behroozi, M., Helluy, X., Ströckens, F., Gao, M., Pusch, R., Tabrik, S., … Güntürkün, O. (2020). Event-related functional MRI of awake behaving pigeons at 7T. Nature Communications, 11(1), 4715. https://doi.org/10.1038/s41467-020-18437-1

Belekhova, M., & Chkheidze, , DD. (1992). Connections of the amygdaloid complex with basotelencephalic, diencephalic and brainstem structures in turtles: HRP study. Rusa Zh Evol Biokhi Fiziol.

Bennett, A. F., & Ruben, J. A. (1979). Endothermy and activity in vertebrates. Science, 206(4419), 649–654. https://doi.org/10.1126/science.493968

Bennett, P. M., & Harvey, P. H. (2009). Brain size, development and metabolism in birds and mammals. Journal of Zoology, 207(4), 491–509. https://doi.org/10.1111/j.1469-7998.1985.tb04946.x

Benowitz, L. I., & Karten, H. J. (1976). Organization of the tectofugal visual pathway in the pigeon: A retrograde transport study. The Journal of Comparative Neurology, 167(4), 503–520. https://doi.org/10.1002/cne.901670407

Benson, R. B. J., Campione, N. E., Carrano, M. T., Mannion, P. D., & Sullivan, C. (2014). Rates of Dinosaur Body Mass Evolution Indicate 170 Million Years of Sustained Ecological Innovation on the Avian Stem Lineage. PLoS Biol, 12(5), 1001853. https://doi.org/10.1371/journal.pbio.1001853

Benton, M. J., & Donoghue, P. C. J. (2006). Paleontological Evidence to Date the Tree of Life. Molecular Biology and Evolution, 24(1), 26–53. https://doi.org/10.1093/molbev/msl150

Benton, M. J., Tverdokhlebov, V. P., & Surkov, M. V. (2004). Ecosystem remodelling among vertebrates at the Permian-Triassic boundary in Russia. Nature, 432(7013), 97–100. https://doi.org/10.1038/nature02950

Benton, M.J., & Clark, J. M. (1988). Archosaur phylogeny and the relationships of the Crocodylia. In M.J. Benton (Ed.), The phylogeny and classification of the tetrapods, Volume 1: Amphibians, Reptiles, Birds (Vol. 35, pp. 295–338). Oxford: Clarendon Press.

Benton, M.J., Donoghue, P. C. J., & Asher, R. J. (2009). Calibrating and constraining molecular clocks. In S. B. Hedges & S. Kumar (Eds.), The Timetree of Life, 35–86.

135

Benton, Michael J. (2012). No gap in the Middle Permian record of terrestrial vertebrates. Geology, 40(4), 339– 342. https://doi.org/10.1130/G32669.1

Benton, Michael J. (2020). The origin of endothermy in synapsids and archosaurs and arms races in the Triassic. Gondwana Research. https://doi.org/10.1016/j.gr.2020.08.003

Benton, Michael J., Dhouailly, D., Jiang, B., & McNamara, M. (2019). The Early Origin of Feathers. Trends in Ecology and Evolution, 34, 856–869. https://doi.org/10.1016/j.tree.2019.04.018

Benton, Michael J., Forth, J., & Langer, M. C. (2014, January 20). Models for the rise of the dinosaurs. Current Biology, 24, R87–R95. https://doi.org/10.1016/j.cub.2013.11.063

Berger, B., Gaspar, P., & Verney, C. (1991). Dopaminergic innervation of the cerebral cortex: unexpected differences between rodents and primates. Trends in Neurosciences, 14(1), 21–27. https://doi.org/10.1016/0166-2236(91)90179-X

Berti, V., Vanzi, E., Polito, C., & Pupi, A. (2013). Back to the future: The absolute quantification of cerebral metabolic rate of glucose. Clinical and Translational Imaging, 1(4), 289–296. https://doi.org/10.1007/s40336-013-0030-2

Bever, G. S., Lyson, T. R., Field, D. J., & Bhullar, B. A. S. (2015). Evolutionary origin of the turtle skull. Nature, 525(7568), 239–242. https://doi.org/10.1038/nature14900

Beyrand, V., Voeten, D. F. A. E., Bureš, S., Fernandez, V., Janáček, J., Jirák, D., … Tafforeau, P. (2019). Multiphase progenetic development shaped the brain of flying archosaurs. Scientific Reports, 9(1), 1–15. https://doi.org/10.1038/s41598-019-46959-2

Billings, B. K., Behroozi, M., Helluy, X., Bhagwandin, A., Manger, P. R., Güntürkün, O., & Ströckens, F. (2020). A three-dimensional digital atlas of the Nile crocodile (Crocodylus niloticus) forebrain. Brain Structure and Function, 225(2), 683–703. https://doi.org/10.1007/s00429-020-02028-3

Billings, B. K., Bhagwandin, A., Patzke, N., Ngwenya, A., Rook, N., von Eugen, K., … Manger, P. R. (2020). Nuclear organization and morphology of catecholaminergic neurons and certain pallial terminal networks in the brain of the Nile crocodile, Crocodylus niloticus. Journal of Chemical Neuroanatomy, 109, 101851. https://doi.org/10.1016/j.jchemneu.2020.101851

Bininda-Emonds, O. R. P., Cardillo, M., Jones, K. E., MacPhee, R. D. E., Beck, R. M. D., Grenyer, R., … Purvis, A. (2007). The delayed rise of present-day mammals. Nature, 446(7135), 507–512. https://doi.org/10.1038/nature05634

Bird, L. R., Roberts, W. A., Abroms, B., Kit, K. A., & Crupi, C. (2003). Spatial Memory for Food Hidden by Rats (Rattus norvegicus) on the Radial Maze: Studies of Memory for Where, What, and When. Journal of Comparative Psychology, 117(2), 176–187. https://doi.org/10.1037/0735-7036.117.2.176 birdbrain. (n.d.). Retrieved September 25, 2020, from Cambridge English Dictionary website: https://dictionary.cambridge.org/dictionary/english/birdbrain

Bischof, H.-J., & Herrmann, K. (1988). Isolation-dependent enhancement of 2-[14C]deoxyglucose uptake in the forebrain of zebra finch males. Behavioral and Neural Biology, 49(3), 386–397. https://doi.org/10.1016/S0163-1047(88)90402-5

136

Bishop, C. M. (2005, May 1). Circulatory variables and the flight performance of birds. Journal of Experimental Biology, 208, 1695–1708. https://doi.org/10.1242/jeb.01576

Blair Hedges, S., & Poling, L. L. (1999). A molecular phylogeny of reptiles. Science, 283(5404), 998–1001. https://doi.org/10.1126/science.283.5404.998

Blake, R. (1993). Cats Perceive Biological Motion. Psychological Science, 4(1), 54–57. https://doi.org/10.1111/j.1467-9280.1993.tb00557.x

Blois, S. de, Novak, M., & Bond, M. (1998). Object permanence in orangutans (Pongo pygmaeus) and squirrel monkeys (Saimiri sciureus). Journal of Comparative Neurology, 112(2), 137-152 https://doi.org/10.1037/0735-7036.112.2.137.

Bock, J., Schnabel, R., & Braun, K. (1997). Role of the dorso-caudal neostriatum in filial imprinting of the domestic chick: a pharmacological and autoradiographical approach focused on the involvement of NMDA-receptors. The European Journal of Neuroscience, 9, 1262–1272. https://doi.org/10.1111/j.1460-9568.1997.tb01481.x

Boire, D., & Baron, G. (1994). Allometric comparison of brain and main brain subdivisions in birds. Journal of Brain Research, 1, 49–66.

Bond, A. B., Kamil, A. C., & Balda, R. P. (2007). Serial Reversal Learning and the Evolution of Behavioral Flexibility in Three Species of North American Corvids (Gymnorhinus cyanocephalus, Nucifraga columbiana, Aphelocoma californica). Journal of Comparative Psychology, 121(4), 372–379. https://doi.org/10.1037/0735-7036.121.4.372

Bonnet, X., Shine, R., & Lourdais, O. (2002, January 1). Taxonomic chauvinism. Trends in Ecology and Evolution, 17, 1–3. https://doi.org/10.1016/S0169-5347(01)02381-3

Borowsky, I. W., & Collins, R. C. (1989). Metabolic anatomy of brain: A comparison of regional capillary density, glucose metabolism, and enzyme activities. Journal of Comparative Neurology, 288(3), 401–413. https://doi.org/10.1002/cne.902880304

Bottjer, S. W. (1993). The distribution of tyrosine hydroxylase immunoreactivity in the brains of male and female zebra finches. Journal of Neurobiology, 24(1), 51–69. https://doi.org/10.1002/neu.480240105

Bottjer, S. W., Brady, J. D., & Cribbs, B. (2000). Connections of a motor cortical region in zebra finches: Relation to pathways for vocal learning. Journal of Comparative Neurology, 420(2), 244–260. https://doi.org/10.1002/(SICI)1096-9861(20000501)420:2<244::AID-CNE7>3.0.CO;2-M

Boysen, S. T., & Berntson, G. G. (1989). Numerical competence in a chimpanzee (Pan troglodytes). Journal of Comparative Psychology, 103(1), 23–31. https://doi.org/10.1037/0735-7036.103.1.23

Bozek, K., Wei, Y., Yan, Z., Liu, X., Xiong, J., Sugimoto, M., … Khaitovich, P. (2014). Exceptional Evolutionary Divergence of Human Muscle and Brain Metabolomes Parallels Human Cognitive and Physical Uniqueness. PLoS Biology, 12(5), e1001871. https://doi.org/10.1371/journal.pbio.1001871

Bradbury, J. W., & Balsby, T. J. S. (2016, March 1). The functions of vocal learning in parrots. Behavioral Ecology and Sociobiology, 70, 293–312. https://doi.org/10.1007/s00265-016-2068-4

Braitenberg, V., & Schüz, A. (1998). Cortex: Statistics and Geometry of Neuronal Connectivity. In Cortex: Statistics and Geometry of Neuronal Connectivity. https://doi.org/10.1007/978-3-662-03733-1

137

Brannon, E. M., & Terrace, H. S. (1998). Ordering of the numerosities 1 to 9 by monkeys. Science, 282(5389), 746–749. https://doi.org/10.1126/science.282.5389.746

Braun, E. J., & Sweazea, K. L. (2008). Glucose regulation in birds. Comparative Biochemistry and Physiology Part B: Biochemistry and Molecular Biology, 151(1), 1–9. https://doi.org/10.1016/j.cbpb.2008.05.007

Braun, K., Bock, J., Metzger, M., Jiang, S., & Schnabel, R. (1999). The dorsocaudal neostriatum of the domestic chick: A structure serving higher associative functions. Behavioural Brain Research, 98(2), 211–218. https://doi.org/10.1016/S0166-4328(98)00086-2

Breuer, J., Grazioso, R., Zhang, N., Schmand, M., & Wienhard, K. (2010). Evaluation of an MR-compatible blood sampler for PET. Physics in Medicine and Biology, 55(19), 5883–5893. https://doi.org/10.1088/0031- 9155/55/19/017

Briscoe, S. D., Albertin, C. B., Rowell, J. J., & Ragsdale, C. W. (2018). Neocortical Association Cell Types in the Forebrain of Birds and Alligators. Current Biology, 28(5), 686-696.e6. https://doi.org/10.1016/j.cub.2018.01.036

Briscoe, S. D., & Ragsdale, C. W. (2018a). Molecular anatomy of the alligator dorsal telencephalon. Journal of Comparative Neurology, 526(10), 1613–1646. https://doi.org/10.1002/cne.24427

Briscoe, S. D., & Ragsdale, C. W. (2018b). Homology, neocortex, and the evolution of developmental mechanisms. Science, 362, 190–193. https://doi.org/10.1126/science.aau3711

Brocklehurst, N., Kammerer, C. F., & Fröbisch, J. (2013). The early evolution of synapsids, and the influence of sampling on their fossil record. Paleobiology, 39(3), 470–490. https://doi.org/10.1666/12049

Brocklehurst, R. J., Schachner, E. R., Codd, J. R., & Sellers, W. I. (2020). Respiratory evolution in archosaurs. Philosophical Transactions of the Royal Society B: Biological Sciences, 375(1793), 20190140. https://doi.org/10.1098/rstb.2019.0140

Brodmann, K. (1909). Vergleichende Lokalisationslehre der Grosshirnrinde in ihren Prinzipien dargestellt auf Grund des Zellenbaues: Leipzig: Barth.

Brown, J., Kaplan, G., Rogers, L. J., & Vallortigara, G. (2010). Perception of biological motion in common marmosets (Callithrix jacchus): by females only. Animal Cognition, 13(3), 555–564. https://doi.org/10.1007/s10071-009-0306-0

Bruce, L. L., & Neary, T. J. (1995). The limbic system of tetrapods: a comparative analysis of cortical and amygdalar populations. Brain, Behavior and Evolution, 46, 224–234. https://doi.org/10.1159/000113276

Bruce, Laura L., & Butler, A. B. (1984). Telencephalic connections in lizards. II. Projections to anterior dorsal ventricular ridge. Journal of Comparative Neurology, 229(4), 602–615. https://doi.org/10.1002/cne.902290412

Brusatte, S. L., O’Connor, J. K., & Jarvis, E. D. (2015, October 5). The origin and diversification of birds. Current Biology, 25, R888–R898. https://doi.org/10.1016/j.cub.2015.08.003

Bugnyar, T., Reber, S. A., & Buckner, C. (2016). Ravens attribute visual access to unseen competitors. Nature Communications, 7, 1–6. https://doi.org/10.1038/ncomms10506

138

Burghardt, G. M. (2014). Play in fishes, frogs and reptiles. Current Biology, 25, R9–R10. https://doi.org/10.1111/eth.12312

Burgin, C. J., Colella, J. P., Kahn, P. L., & Upham, N. S. (2018). How many species of mammals are there? Journal of Mammalogy, 99(1), 1–14. https://doi.org/10.1093/jmammal/gyx147

Burn, C. C. (2008). What is it like to be a rat? Rat sensory perception and its implications for experimental design and rat welfare. Applied Animal Behaviour Science, 112, 1–32. https://doi.org/10.1016/j.applanim.2008.02.007

Buschman, T. J., Siegel, M., Roy, J. E., & Miller, E. K. (2011). Neural substrates of cognitive capacity limitations. Proceedings of the National Academy of Sciences of the United States of America, 108(27), 11252–11255. https://doi.org/10.1073/pnas.1104666108

Butler, P.J. (1991). Exercise in Birds. Journal of Experimental Biology, 160(1), 233-262.

Butler, Patrick J. (2016). The physiological basis of bird flight. Philosophical Transactions of the Royal Society B: Biological Sciences, 371(1704), 20150384. https://doi.org/10.1098/rstb.2015.0384

Butter, C. M. (1969). Perseveration in extinction and in discrimination reversal tasks following selective frontal ablations in Macaca mulatta. Physiology and Behavior, 4(2), 163–171. https://doi.org/10.1016/0031- 9384(69)90075-4

Caminiti, R., Ghaziri, H., Galuske, R., Hof, P. R., & Innocenti, G. M. (2009). Evolution amplified processing with temporally dispersed slow neuronal connectivity in primates. Proceedings of the National Academy of Sciences of the United States of America, 106(46), 19551–19556. https://doi.org/10.1073/pnas.0907655106

Carlén, M. (2017). What constitutes the prefrontal cortex? Science, 358(6362), 478–482. https://doi.org/10.1126/science.aan8868

Carroll, R. L. (1988). Vertebrate paleontology and evolution.

Carter, B. C., & Bean, B. P. (2009). Sodium Entry during Action Potentials of Mammalian Neurons: Incomplete Inactivation and Reduced Metabolic Efficiency in Fast-Spiking Neurons. Neuron, 64(6), 898–909. https://doi.org/10.1016/j.neuron.2009.12.011

Chakraborty, M., & Jarvis, E. D. (2015). Brain evolution by brain pathway duplication. Philosophical Transactions of the Royal Society B: Biological Sciences, 370(1684), 20150056. https://doi.org/10.1098/rstb.2015.0056

Cherniak, C., Mokhtarzada, Z., Rodriguez-Esteban, R., & Changizi, K. (2004). Global optimization of cerebral cortex layout. Proceedings of the National Academy of Sciences of the United States of America, 101(4), 1081–1086. https://doi.org/10.1073/pnas.0305212101

Chiari, Y., Cahais, V., Galtier, N., & Delsuc, F. (2012a). Phylogenomic analyses support the position of turtles as the sister group of birds and crocodiles (Archosauria). BMC Biology, 10(1), 1–15. https://doi.org/10.1186/1741-7007-10-65

Chiari, Y., Cahais, V., Galtier, N., & Delsuc, F. (2012b). Phylogenomic analyses support the position of turtles as the sister group of birds and crocodiles (Archosauria). BMC Biology, 10(1), 1–15. https://doi.org/10.1186/1741-7007-10-65

139

Clark, D. A., Mitra, P. P., & Wang, S. S. H. (2001). Scalable architecture in mammalian brains. Nature, 411(6834), 189–193. https://doi.org/10.1038/35075564

Clark, J. M., & Ulinski, P. S. (1984). A Golgi study of anterior dorsal ventricular ridge in the alligator,Alligator mississippiensis. Journal of Morphology, 179(2), 153–174. https://doi.org/10.1002/jmor.1051790204

Clark, L., Cools, R., & Robbins, T. W. (2004). The neuropsychology of ventral prefrontal cortex: Decision-making and reversal learning. Brain and Cognition, 55(1), 41–53. https://doi.org/10.1016/S0278-2626(03)00284-7

Clark, R. E., & Zola, S. (1998). Trace eyeblink classical conditioning in the monkey: A nonsurgical method and behavioral analysis. Behavioral Neuroscience, 112(5), 1062–1068. https://doi.org/10.1037/0735- 7044.112.5.1062

Clarke, A., & Pörtner, H.-O. (2010). Temperature, metabolic power and the evolution of endothermy. Biological Reviews, 85(4), no-no. https://doi.org/10.1111/j.1469-185X.2010.00122.x

Clarke, A., & Rothery, P. (2007). Scaling of body temperature in mammals and birds. Functional Ecology, 22(1), 58-67. https://doi.org/10.1111/j.1365-2435.2007.01341.x

Clayton, N., & Dickinson, A. (1998). Episodic-like memory during cache recovery by scrub jays. Nature, 395(6699), 272-274.

Cloninger, C. R., & Gilligan, S. B. (1987). Neurogenetic mechanisms of learning: A phylogenetic perspective. Journal of Psychiatric Research, 21(4), 457–472. https://doi.org/10.1016/0022-3956(87)90094-X

Cnotka, J., Güntürkün, O., Rehkämper, G., Gray, R. D., & Hunt, G. R. (2008). Extraordinary large brains in tool- using New Caledonian crows (Corvus moneduloides). Neuroscience Letters, 433(3), 241–245. https://doi.org/10.1016/J.NEULET.2008.01.026

Cosans, C. E., & Ulinski, P. S. (1990). Spatial organization of axons in turtle visual cortex: Intralamellar and interlamellar projections. The Journal of Comparative Neurology, 296(4), 548–558. https://doi.org/10.1002/cne.902960404

Crawford, N. G., Faircloth, B. C., McCormack, J. E., Brumfield, R. T., Winker, K., & Glenn, T. C. (2012). More than 1000 ultraconserved elements provide evidence that turtles are the sister group of archosaurs. Biology Letters, 8(5), 783–786. https://doi.org/10.1098/rsbl.2012.0331

Crawford, N. G., Parham, J. F., Sellas, A. B., Faircloth, B. C., Glenn, T. C., Papenfuss, T. J., … Simison, W. B. (2015). A phylogenomic analysis of turtles. Molecular Phylogenetics and Evolution, 83, 250–257. https://doi.org/10.1016/j.ympev.2014.10.021

Cubo, J., Le Roy, N., Martinez-Maza, C., & Montes, L. (2012). Paleohistological estimation of bone growth rate in extinct archosaurs. Paleobiology, 38(2), 335–349. https://doi.org/10.1666/08093.1

Cunningham, V. J., & Cremer, J. E. (1981). A method for the simultaneous estimation of regional rates of glucose influx and phosphorylation in rat brain using radiolabeled 2-deoxyglucose. Brain Research, 221(2), 319– 330. https://doi.org/10.1016/0006-8993(81)90781-2

Daum, I., Schugens, M. M., Channon, S., Polkey, C. E., & Gray, J. A. (1991). T-Maze Discrimination and Reversal Learning After Unilateral Temporal or Frontal Lobe Lesions in Man. Cortex, 27(4), 613–622. https://doi.org/10.1016/S0010-9452(13)80010-X

140

Davies, D. C., Csillag, A., Sz�kely, A. D., & Kabai, P. (1997). Efferent connections of the domestic chick archistriatum: A phaseolus lectin anterograde tracing study. The Journal of Comparative Neurology, 389(4), 679–693. https://doi.org/10.1002/(SICI)1096-9861(19971229)389:4<679::AID-CNE10>3.0.CO;2-7

Davis, K. M., & Burghardt, G. M. (2007). Training and long-term memory of a novel food acquisition task in a turtle (Pseudemys nelsoni). Behavioural Processes, 75(2), 225–230. https://doi.org/10.1016/j.beproc.2007.02.021

Dawson, R. R., Field, D. J., Hull, P. M., Zelenitsky, D. K., Therrien, F., & Affek, H. P. (2020). Eggshell geochemistry reveals ancestral metabolic thermoregulation in Dinosauria. Science Advances, 6(7), eaax9361. https://doi.org/10.1126/sciadv.aax9361

Day, L. B., Crews, D., & Wilczynski, W. (1999). Spatial and reversal learning in congeneric lizards with different foraging strategies. Animal Behaviour, 57(2), 393–407. https://doi.org/10.1006/anbe.1998.1007

Day, L. B., Ismail, N., & Wilczynski, W. (2003). Use of Position and Feature Cues in Discrimination Learning by the Whiptail Lizard (Cnemidophorus inornatus). Journal of Comparative Psychology, 117(4), 440–448. https://doi.org/10.1037/0735-7036.117.4.440

De Groof, G., Jonckers, E., Güntürkün, O., Denolf, P., Van Auderkerke, J., & Van der Linden, A. (2013). Functional MRI and functional connectivity of the visual system of awake pigeons. Behavioural Brain Research, 239(1), 43–50. https://doi.org/10.1016/j.bbr.2012.10.044

DeCasien, A. R., & Higham, J. P. (2019). Primate mosaic brain evolution reflects selection on sensory and cognitive specialization. Nature Ecology and Evolution, 3(10), 1483–1493. https://doi.org/10.1038/s41559- 019-0969-0

Desfilis, E., Abellán, A., Sentandreu, V., & Medina, L. (2018). Expression of regulatory genes in the embryonic brain of a lizard and implications for understanding pallial organization and evolution. Journal of Comparative Neurology, 526(1), 166–202. https://doi.org/10.1002/cne.24329

Dial, K. P. (2003). Evolution of Avian Locomotion: Correlates of Flight Style, Locomotor Modules, Nesting Biology, Body Size, Development, and The Origin of Flapping Flight. The Auk, 120(4), 941–952. https://doi.org/10.1093/auk/120.4.941

Diamond, A. (2013). Executive Functions. Annual Review of Psychology, 64(1), 135–168. https://doi.org/10.1146/annurev-psych-113011-143750

Dias, R., Robbins, T. W., & Roberts, A. C. (1996). Dissociation in prefrontal cortex of affective and attentional shifts. Nature, 380(6569), 69–72. https://doi.org/10.1038/380069a0

Dicke, U., & Roth, G. (2016). Neuronal factors determining high intelligence. Philosophical Transactions of the Royal Society B: Biological Sciences, 371(1685), 20150180. https://doi.org/10.1098/rstb.2015.0180

Dickinson, A. (2012). Associative learning and animal cognition. Philosophical Transactions of the Royal Society B: Biological Sciences, 367(1603), 2733–2742. https://doi.org/10.1098/rstb.2012.0220

Diekamp, B, Kalt, T., & Güntürkün, O. (2002). Working memory neurons in pigeons. The Journal of Neuroscience, 22(4), RC210.

Diekamp, Bettina, Gagliardo, A., & Güntürkün, O. (2002). Nonspatial and subdivision-specific working memory deficits after selective lesions of the avian prefrontal cortex. The Journal of Neuroscience, 22(21), 9573– 9580. https://doi.org/22/21/9573

141

Dienel, G. A. (2019). Brain glucose metabolism: Integration of energetics with function. Physiological Reviews, 99(1), 949–1045. https://doi.org/10.1152/physrev.00062.2017

Divac, I. (1973). Delayed response in cats after frontal lesions extending beyond the gyrus proreus. Physiology & Behavior, 10(4), 717–720. https://doi.org/10.1016/0031-9384(73)90151-0

Divac, I., Mogensen, J., & Björklund, A. (1985). The prefrontal “cortex” in the pigeon. Biochemical evidence. Brain Research, 332(2), 365–368. https://doi.org/10.1016/0006-8993(85)90606-7

Donoghue, P. C. J., & Benton, M. J. (2007). Rocks and clocks: calibrating the Tree of Life using fossils and molecules. Trends in Ecology and Evolution, 22, 424–431. https://doi.org/10.1016/j.tree.2007.05.005 dos Reis, M., Inoue, J., Hasegawa, M., Asher, R. J., Donoghue, P. C. J., & Yang, Z. (2012). Phylogenomic datasets provide both precision and accuracy in estimating the timescale of placental mammal phylogeny. Proceedings of the Royal Society B: Biological Sciences, 279(1742), 3491–3500. https://doi.org/10.1098/rspb.2012.0683

Downs, C. T., Greaver, C., & Taylor, R. (2008). Body temperature and basking behaviour of Nile crocodiles (Crocodylus niloticus) during winter. Journal of Thermal Biology, 33(3), 185–192. https://doi.org/10.1016/j.jtherbio.2008.02.001

Ducatez, S., Sol, D., Sayol, F., & Lefebvre, L. (2020). Behavioural plasticity is associated with reduced extinction risk in birds. Nature Ecology and Evolution, 4(6), 788–793. https://doi.org/10.1038/s41559-020-1168-8

Dufour, V., Wascher, C. A. F., Braun, A., Miller, R., & Bugnyar, T. (2012). Corvids can decide if a future exchange is worth waiting for. Biology Letters, 8(2), 201–204. https://doi.org/10.1098/rsbl.2011.0726

Dugas-Ford, J., Rowell, J. J., & Ragsdale, C. W. (2012a). Cell-type homologies and the origins of the neocortex. Proceedings of the National Academy of Sciences of the United States of America, 109(42), 16974–16979. https://doi.org/10.1073/pnas.1204773109

Dugas-Ford, J., Rowell, J. J., & Ragsdale, C. W. (2012b). Cell-type homologies and the origins of the neocortex. Proceedings of the National Academy of Sciences of the United States of America, 109(42), 16974–16979. https://doi.org/10.1073/pnas.1204773109

Durstewitz, D., Kr??ner, S., Hemmings, H. C., & G??nt??rk??n, O. (1998). The dopaminergic innervation of the pigeon telencephalon: Distribution of DARPP-32 and co-occurrence with glutamate decarboxylase and tyrosine hydroxylase. Neuroscience, 83(3), 763–779. https://doi.org/10.1016/S0306-4522(97)00450-8

Durstewitz, D., Kröner, S., & Güntürkün, O. (1999). The dopaminergic innervation of the avian telencephalon. Progress in Neurobiology, 59(2), 161–195. https://doi.org/10.1016/S0301-0082(98)00100-2

Dykes, M., Klarer, A., Porter, B., Rose, J., & Colombo, M. (2018). Neurons in the pigeon nidopallium caudolaterale display value-related activity. Scientific Reports, 8(1). https://doi.org/10.1038/s41598-018- 23694-8

Emery, N. J., & Clayton, N. S. (2004). The mentality of crows: convergent evolution of intelligence in corvids and apes. Science, 306(5703), 1903–1907. https://doi.org/10.1126/science.1098410

Erwin, D. H. (1994). The Permo-Triassic extinction. Nature, 367, 231–236. https://doi.org/10.1038/367231a0

142

Evans, S. E. (2008). Evolution and Phylogeny of Amniotes. In Encyclopedia of Neuroscience (1192–1194). https://doi.org/10.1007/978-3-540-29678-2_3113

Faisal, A. A., White, J. A., & Laughlin, S. B. (2005). Ion-channel noise places limits on the miniaturization of the brain’s wiring. Current Biology, 15(12), 1143–1149. https://doi.org/10.1016/j.cub.2005.05.056

Fallon, J. H., & Moore, R. Y. (1978). Catecholamine innervation of the basal forebrain IV. Topography of the dopamine projection to the basal forebrain and neostriatum. The Journal of Comparative Neurology, 180(3), 545–579. https://doi.org/10.1002/cne.901800310

Faraci, F. M. (1991). Adaptations to hypoxia in birds: how to fly. In Annual Review of Physiology, 53(1), 59-70.

Farmer, C. G., & Sanders, K. (2010). Unidirectional airflow in the lungs of alligators. Science, 327(5963), 338– 340. https://doi.org/10.1126/science.1180219

Feenders, G., Liedvogel, M., Rivas, M., Zapka, M., Horita, H., Hara, E., … Jarvis, E. D. (2008). Molecular Mapping of Movement-Associated Areas in the Avian Brain: A Motor Theory for Vocal Learning Origin. PLoS ONE, 3(3), e1768. https://doi.org/10.1371/journal.pone.0001768

Felten, D. L., & Sladek, J. R. (1983). Monoamine distribution in primate brain V. Monoaminergic nuclei: Anatomy, pathways and local organization. Brain Research Bulletin, 10(2), 171–284. https://doi.org/10.1016/0361- 9230(83)90045-X

Field, D. J., Gauthier, J. A., King, B. L., Pisani, D., Lyson, T. R., & Peterson, K. J. (2014). Toward consilience in reptile phylogeny: miRNAs support an archosaur, not lepidosaur, affinity for turtles. Evolution & Development, 16(4), 189–196. https://doi.org/10.1111/ede.12081

Font, E., García-Roa, R., Pincheira-Donoso, D., & Carazo, P. (2019). Rethinking the Effects of Body Size on the Study of Brain Size Evolution. Brain, Behavior and Evolution, 93(4), 182–195. https://doi.org/10.1159/000501161

Ford, D. P., & Benson, R. B. J. (2020). The phylogeny of early amniotes and the affinities of Parareptilia and Varanopidae. Nature Ecology and Evolution, 4(1), 57–65. https://doi.org/10.1038/s41559-019-1047-3

Frey, E., Tarsitano, S. F., Oelofsen, B., & Riess, J. (2001). The origin of temporal fenestrae. South African Journal of Science, 97(7–8), 334–336.

Fristoe, T. S., Iwaniuk, A. N., & Botero, C. A. (2017). Big brains stabilize populations and facilitate colonization of variable habitats in birds. Nature Ecology and Evolution, 1(11), 1706–1715. https://doi.org/10.1038/s41559- 017-0316-2

Fu, X., Giavalisco, P., Liu, X., Catchpole, G., Fu, N., Ning, Z. Bin, … Khaitovich, P. (2011). Rapid metabolic evolution in human prefrontal cortex. Proceedings of the National Academy of Sciences of the United States of America, 108(15), 6181–6186. https://doi.org/10.1073/pnas.1019164108

Fueger, B. J., Czernin, J., Hildebrandt, I., Tran, C., Halpern, B. S., Stout, D., … Weber, W. A. (2006). Impact of animal handling on the results of 18F-FDG PET studies in mice. Journal of Nuclear Medicine, 47(6), 999– 1006.

Funke, K. (1989). Somatosensory areas in the telencephalon of the pigeon. II. Spinal pathways and afferent connections. Experimental Brain Research, 76(3), 620–638. https://doi.org/10.1007/bf00248918

143

Fuster, J M. (1973). Unit activity in prefrontal cortex during delayed-response performance: neuronal correlates of transient memory. Journal of Neurophysiology, 36(1), 61–78. https://doi.org/10.1152/jn.1973.36.1.61

Fuster, Joaquin M. (2015). The prefrontal cortex. Academic Press.

Fuster, Joaquin M., & Alexander, G. E. (1971). Excitation and inhibition of neuronal firing in visual cortex by reticular stimulation. Science, 133(3469), 2011–2012. https://doi.org/10.1126/science.133.3469.2011

Gagliardo, A., Bonadonna, F., & Divac, I. (1996). Behavioural effects of ablations of the presumed ‘prefrontal cortex’ or the corticoid in pigeons. Behavioural Brain Research, 78(2), 155–162. https://doi.org/10.1016/0166-4328(95)00243-X

Gaines Das, R., & North, D. (2007). Implications of experimental technique for analysis and interpretation of data from animal experiments: outliers and increased variability resulting from failure of intraperitoneal injection procedures. Laboratory Animals, 41(3), 312–320. https://doi.org/10.1258/002367707781282802

Gallyas, F. (1971). Silver staining of Alzheimer’s neurofibrillary changes by means of physical development. Acta Morphologica Academiae Scientiarum Hungaricae, 19(1), 1–8.

Garamszegi, L. Z., Eens, M., Erritzøe, J., & Møller, A. P. (2005). Sexually size dimorphic brains and song complexity in passerine birds. Behavioral Ecology, 16(2), 335–345. https://doi.org/10.1093/beheco/arh167

Gaspar, P., Stepniewska, I., & Kaas, J. H. (1992). Topography and collateralization of the dopaminergic projections to motor and lateral prefrontal cortex in owl monkeys. The Journal of Comparative Neurology, 325(1), 1–21. https://doi.org/10.1002/cne.903250102

Gatesy, S. M., & Dial, K. P. (1996). Locomotor modules and the evolution of avian flight. Evolution, 50(1), 331– 340. https://doi.org/10.1111/j.1558-5646.1996.tb04496.x

Genç, E., Fraenz, C., Schlüter, C., Friedrich, P., Hossiep, R., Voelkle, M. C., … Jung, R. E. (2018). Diffusion markers of dendritic density and arborization in gray matter predict differences in intelligence. Nature Communications, 9(1), 1–11. https://doi.org/10.1038/s41467-018-04268-8

Gill, R. E., Tibbitts, T. L., Douglas, D. C., Handel, C. M., Mulcahy, D. M., Gottschalck, J. C., … Piersma, T. (2009). Extreme endurance flights by landbirds crossing the Pacific Ocean: ecological corridor rather than barrier? Proceedings of the Royal Society B: Biological Sciences, 276(1656), 447–457. https://doi.org/10.1098/rspb.2008.1142

Gjedde, A., & Diemer, N. H. (1983). Autoradiographic determination of regional brain glucose content. Journal of Cerebral Blood Flow and Metabolism, 3(3), 303–310. https://doi.org/10.1038/jcbfm.1983.45

Glenn Northcutt, R., & Kaas, J. H. (1995). The emergence and evolution of mammalian neocortex. Trends in Neurosciences, 18(9), 373–379. https://doi.org/10.1016/0166-2236(95)93932-N

Godefroit, P., Sinitsa, S. M., Dhouailly, D., Bolotsky, Y. L., Sizov, A. V., McNamara, M. E., … Spagna, P. (2014). A Jurassic ornithischian dinosaur from Siberia with both feathers and scales. Science, 345(6195), 451–455. https://doi.org/10.1126/science.1253351

Goffinet, A. M. (2017). The evolution of cortical development: The synapsid-diapsid divergence. Development (Cambridge), 144(22), 4061–4077. https://doi.org/10.1242/dev.153908

144

Goldman-Rakic, P. (1987). Circuit basis of a cognitive function in non-human primates. In S. Stahl, S. Iversen, & E. Goodman (Eds.), Cognitive neurochemistry (pp. 90–110). Oxford: Oxford University Press.

Gormezano, I., Prokasy, W. F., & Thompson, R. F. (1987). Classical conditioning. Experimental methods and instrumentation in psychology, 1, 385-420.

Gossette, R. L., & Hombach, A. (1969). Successive discrimination reversal (SDR) performances of American alligators and American crocodiles on a spatial task. Perceptual and Motor Skills, 28(1), 63–67. https://doi.org/10.2466/pms.1969.28.1.63

Green, L., Fisher, E., Perlow, S., & Sherman, L. (1981). Preference reversal and self control: Choice as a function of reward amount and delay. - PsycNET. Behavioural Analysis Letters, 1, 43–51.

Green, R. E., Braun, E. L., Armstrong, J., Earl, D., Nguyen, N., Hickey, G., … Ray, D. A. (2014). Three crocodilian genomes reveal ancestral patterns of evolution among archosaurs. Science, 346(6215). https://doi.org/10.1126/science.1254449

Greggor, A. L., McIvor, G. E., Clayton, N. S., & Thornton, A. (2018). Wild jackdaws are wary of objects that violate expectations of animacy. Royal Society Open Science, 5(10), 181070. https://doi.org/10.1098/rsos.181070

Grillner, S., Robertson, B., & Stephenson-Jones, M. (2013, November). The evolutionary origin of the vertebrate basal ganglia and its role in action selection. Journal of Physiology, 591, 5425–5431. https://doi.org/10.1113/jphysiol.2012.246660

Grubb, B. R. (1983). Allometric relations of cardiovascular function in birds. American Journal of Physiology - Heart and Circulatory Physiology, 14(4). https://doi.org/10.1152/ajpheart.1983.245.4.h567

Güntürkün, O. (2000). Sensory Physiology: Vision. In Sturkie’s Avian Physiology (pp. 1–19). https://doi.org/10.1016/b978-012747605-6/50002-x

Güntürkün, O. (1997). Cognitive impairments after lesions of the neostriatum caudolaterale and its thalamic afferent in pigeons: functional similarities to the mammalian prefrontal system? Journal Fur Hirnforschung, 38(1), 133–143.

Güntürkün, O, Stacho, M., & Ströckens, F. (2017). The brains of reptiles and birds. In J. Kaas (Ed.), Evolution of Nervous Systems (2nd ed., pp. 171–221).

Güntürkün, O. (2005a). Avian and mammalian “prefrontal cortices”: Limited degrees of freedom in the evolution of the neural mechanisms of goal-state maintenance. Brain Research Bulletin, 66(4–6), 311–316. https://doi.org/10.1016/j.brainresbull.2005.02.004

Güntürkün, O. (2005b). The avian “prefrontal cortex” and cognition. Current Opinion in Neurobiology, 15(6), 686– 693. https://doi.org/10.1016/j.conb.2005.10.003

Güntürkün, O. (2012). The convergent evolution of neural substrates for cognition. Psychological Research, 76(2), 212–219. https://doi.org/10.1007/s00426-011-0377-9

Güntürkün, O, & Bugnyar, T. (2016). Cognition without Cortex. Trends in Cognitive Sciences, 20(4), 291–303. https://doi.org/10.1016/j.tics.2016.02.001

145

Güntürkün, O, Koenen, C., Iovine, F., Garland, A., & Pusch, R. (2018). The neuroscience of perceptual categorization in pigeons: A mechanistic hypothesis. Learning and Behavior, 46(3), 229–241. https://doi.org/10.3758/s13420-018-0321-6

Güntürkün, O., Ströckens, F., Scarf, D., & Colombo, M. (2017). Apes, feathered apes, and pigeons: differences and similarities. Current Opinion in Behavioral Sciences, 16, 35–40. https://doi.org/10.1016/j.cobeha.2017.03.003

Hahn, L. A., & Rose, J. (2020). Working Memory as an Indicator for Comparative Cognition – Detecting Qualitative and Quantitative Differences. Frontiers in Psychology, 11, 1–9. https://doi.org/10.3389/fpsyg.2020.01954

Hammond, K. A., & Diamond, J. (1997, April 3). Maximal sustained energy budgets in humans and animals. Nature, 386, 457–462. https://doi.org/10.1038/386457a0

Hankin, B. L., & Stein, W. D. (1972). On the temperature dependence of initial velocities of glucose transport in the human red blood cell. BBA - Biomembranes, 288(1), 127–136. https://doi.org/10.1016/0005- 2736(72)90230-1

Hartmann, B., & Güntürkün, O. (1998). Selective deficits in reversal learning after neostriatum caudolaterale lesions in pigeons: Possible behavioral equivalencies to the mammalian prefrontal system. Behavioural Brain Research, 96(1–2), 125–133. https://doi.org/10.1016/S0166-4328(98)00006-0

Hasenstaub, A., Otte, S., Callaway, E., & Sejnowski, T. J. (2010). Metabolic cost as a unifying principle governing neuronal biophysics. Proceedings of the National Academy of Sciences of the United States of America, 107(27), 12329–12334. https://doi.org/10.1073/pnas.0914886107

Hasselbalch, S. G., Knudsen, G. M., Holm, S., Hageman, L. P., Capaldo, B., & Paulson, O. B. (1996). Transport of D-glucose and 2-fluorodeoxyglucose across the blood-brain barrier in humans. Journal of Cerebral Blood Flow and Metabolism, 16(4), 659–666. https://doi.org/10.1097/00004647-199607000-00017

Hasselbalch, S. G., Madsen, P. L., Knudsen, G. M., Holm, S., & Paulson, O. B. (1998). Calculation of the FDG lumped constant by simultaneous measurements of global glucose and FDG metabolism in humans. Journal of Cerebral Blood Flow and Metabolism, 18(2), 154–160. https://doi.org/10.1097/00004647- 199802000-00005

Herculano-Houzel, S. (2009). The human brain in numbers: a linearly scaled-up primate brain. Frontiers in Human Neuroscience, 3, 31. https://doi.org/10.3389/neuro.09.031.2009

Herculano-Houzel, S. (2011). Scaling of brain metabolism with a fixed energy budget per neuron: Implications for neuronal activity, plasticity and evolution. PLoS ONE, 6(3), e17514. https://doi.org/10.1371/journal.pone.0017514

Herculano-Houzel, S., Catania, K., Manger, P. R., & Kaas, J. H. (2015). Mammalian Brains Are Made of These: A Dataset of the Numbers and Densities of Neuronal and Nonneuronal Cells in the Brain of Glires, Primates, Scandentia, Eulipotyphlans, Afrotherians and Artiodactyls, and Their Relationship with Body Mass. Brain, Behavior and Evolution, 86(3–4), 145–163. https://doi.org/10.1159/000437413

Herculano-Houzel, S., Collins, C. E., Wong, P., & Kaas, J. H. (2007). Cellular scaling rules for primate brains. Proceedings of the National Academy of Sciences of the United States of America, 104(9), 3562–3567. https://doi.org/10.1073/pnas.0611396104

146

Herculano-Houzel, S., Manger, P. R., & Kaas, J. H. (2014). Brain scaling in mammalian evolution as a consequence of concerted and mosaic changes in numbers of neurons and average neuronal cell size. Frontiers in Neuroanatomy, 8, 1–28. https://doi.org/10.3389/fnana.2014.00077

Herculano-Houzel, S., Mota, B., & Lent, R. (2006). Cellular scaling rules for rodent brains. Proceedings of the National Academy of Sciences of the United States of America, 103(32), 12138–12143. https://doi.org/10.1073/pnas.0604911103

Herold, C., Palomero-Gallagher, N., Hellmann, B., Kröner, S., Theiss, C., Güntürkün, O., … Zilles, K. (2011). The receptor architecture of the pigeons’ nidopallium caudolaterale: An avian analogue to the mammalian prefrontal cortex. Brain Structure and Function, 216(3), 239–254. https://doi.org/10.1007/s00429-011-0301- 5

Herold, C., Paulitschek, C., Palomero-Gallagher, N., Güntürkün, O., & Zilles, K. (2018). Transmitter receptors reveal segregation of the arcopallium/amygdala complex in pigeons (Columba livia). Journal of Comparative Neurology, 526(3), 439–466. https://doi.org/10.1002/cne.24344

Heyers, D., Kovjanic, D., & Redies, C. (2003). Cadherin expression coincides with birth dating patterns in patchy compartments of the developing chicken telencephalon. The Journal of Comparative Neurology, 460(2), 155–166. https://doi.org/10.1002/cne.10631

Hodgkin, A. L., & Huxley, A. F. (1952). The components of membrane conductance in the giant axon of Loligo. The Journal of Physiology, 116(4), 473–496. https://doi.org/10.1113/jphysiol.1952.sp004718

Hodos, W., Karten, H. J., & Bonbright, J. C. (1973). Visual intensity and pattern discrimination after lesions of the thalamofugal visual pathway in pigeons. The Journal of Comparative Neurology, 148(4), 447–467. https://doi.org/10.1002/cne.901480404

Holmes, P. A., & Bitterman, M. E. (1966). Spatial and visual habit reversal in the turtle. Journal of Comparative and Physiological Psychology, 62(2), 328–331. https://doi.org/10.1037/h0023675

Howarth, C., Gleeson, P., & Attwell, D. (2012). Updated energy budgets for neural computation in the neocortex and cerebellum. Journal of Cerebral Blood Flow and Metabolism, 32, 1222–1232. https://doi.org/10.1038/jcbfm.2012.35

Howat, W. J., & Wilson, B. A. (2014). Tissue fixation and the effect of molecular fixatives on downstream staining procedures. Methods, 70(1), 12–19. https://doi.org/10.1016/j.ymeth.2014.01.022

Huang, C. C., Wu, C. H., Huang, Y. Y., Tzen, K. Y., Chen, S. F., Tsai, M. L., & Wu, H. M. (2017). Performing repeated quantitative small-animal PET with an arterial input function is routinely feasible in rats. Journal of Nuclear Medicine, 58(4), 611–616. https://doi.org/10.2967/jnumed.116.182402

Hubrecht, R., & Kirkwood, J. (2010). The UFAW Handbook on the Care and Management of Laboratory and Other Research Animals. In R. Hubrecht & J. Kirkwood (Eds.), The UFAW Handbook on the Care and Management of Laboratory and Other Research Animals: Eighth Edition: Eighth Edition. https://doi.org/10.1002/9781444318777

Hudgins, M. N., Uhen, M. D., & Hinnov, L. A. (2020). The evolution of respiratory systems in Theropoda and Paracrocodylomorpha, the end-Triassic extinction, and the role of Late Triassic atmospheric O2 and CO2. Palaeogeography, Palaeoclimatology, Palaeoecology, 545, 109638. https://doi.org/10.1016/j.palaeo.2020.109638

147

Hull, J. H. (1977). Instrumental response topographies of rats. Animal Learning & Behavior, 5(2), 207–212. https://doi.org/10.3758/BF03214080

Hunt, G. R., & Gray, R. D. (2004). The crafting of hook tools by wild New Caledonian crows. Proceedings of the Royal Society of London. Series B: Biological Sciences, 271. https://doi.org/10.1098/rsbl.2003.0085

Hyder, F., Rothman, D. L., & Bennett, M. R. (2013, February 26). Cortical energy demands of signaling and nonsignaling components in brain are conserved across mammalian species and activity levels. Proceedings of the National Academy of Sciences of the United States of America, 110, 3549–3554. https://doi.org/10.1073/pnas.1214912110

Innocenti, G. M., Vercelli, A., & Caminiti, R. (2014). The diameter of cortical axons depends both on the area of origin and target. Cerebral Cortex, 24(8), 2178–2188. https://doi.org/10.1093/cercor/bht070

Isler, K., & van Schaik, C. (2006a). Costs of encephalization: the energy trade-off hypothesis tested on birds. Journal of Human Evolution, 51(3), 228–243. https://doi.org/10.1016/j.jhevol.2006.03.006

Isler, K., & van Schaik, C. P. (2006b). Metabolic costs of brain size evolution. Biology Letters, 2(4), 557–560. https://doi.org/10.1098/rsbl.2006.0538

Isler, K., & van Schaik, C. P. (2009). The Expensive Brain: A framework for explaining evolutionary changes in brain size. Journal of Human Evolution, 57(4), 392–400. https://doi.org/10.1016/j.jhevol.2009.04.009

Isler, K., & Van Schaik, C. P. (2009). Why are there so few smart mammals (but so many smart birds)? Biology Letters, 5(1), 125–129. https://doi.org/10.1098/rsbl.2008.0469

Iwaniuk, A. N. (2017). The Evolution of Cognitive Brains in Non-mammals. In Evolution of the Brain, Cognition, and Emotion in Vertebrates (pp. 101–124). Tokyo: Springer. https://doi.org/10.1007/978-4-431-56559-8_5

Iwaniuk, A. N., Dean, K. M., & Nelson, J. E. (2004). A mosaic pattern characterizes the evolution of the avian brain. Proceedings of the Royal Society B: Biological Sciences, 271, 148–151. https://doi.org/10.1098/rsbl.2003.0127

Iwaniuk, A. N., & Hurd, P. L. (2005). The evolution of cerebrotypes in birds. Brain, Behavior and Evolution, 65(4), 215–230. https://doi.org/10.1159/000084313

Izawa, E.-I., & Watanabe, S. (2007). A stereotaxic atlas of the brain of the jungle crow (Corvus macrorhynchos). Integration of Comparative Neuroanatomy and Cognition, 215–273.

Izquierdo, A., Brigman, J. L., Radke, A. K., Rudebeck, P. H., & Holmes, A. (2017, March 14). The neural basis of reversal learning: An updated perspective. Neuroscience, 345, 12–26. https://doi.org/10.1016/j.neuroscience.2016.03.021

Jarvis, E. D. (2009). Evolution of the pallium in birds and reptiles. In M. D. Binder, N. Hirokawa, U. Windhorst, & A. Butler (Eds.), New Encyclopedia of Neuroscience (pp. 1385–1400).

Jerison, H. (1973). Evolution of the brain and intelligence. Elsevier.

Jetz, W., Thomas, G. H., Joy, J. B., Hartmann, K., & Mooers, A. O. (2012). The global diversity of birds in space and time. Nature, 491(7424), 444–448. https://doi.org/10.1038/nature11631

148

Jiao, Y., , Z., Lee, T., Fusco, F. R., Kimble, T. D., Meade, C. A., … Reiner, A. (1999). A simple and sensitive antigen retrieval method for free-floating and slide-mounted tissue sections. Journal of Neuroscience Methods, 93(2), 149–162. https://doi.org/10.1016/S0165-0270(99)00142-9

Johnston, M., Anderson, C., & Colombo, M. (2017). Neural correlates of sample-coding and reward-coding in the delay activity of neurons in the entopallium and nidopallium caudolaterale of pigeons (Columba livia). Behavioural Brain Research, 317, 382–392. https://doi.org/10.1016/j.bbr.2016.10.003

Jønsson, K. A., Fabre, P. H., & Irestedt, M. (2012). Brains, tools, innovation and biogeography in crows and ravens. BMC Evolutionary Biology, 12(1). https://doi.org/10.1186/1471-2148-12-72

Kaas, J.H. (2020). Evolution of Visual Cortex in Primates. In Evolutionary Neuroscience (pp. 547–564). https://doi.org/10.1016/b978-0-12-820584-6.00023-4

Kaas, Jon H. (2011). Neocortex in early mammals and its subsequent variations. Annals of the New York Academy of Sciences, 1225(1), 28–36. https://doi.org/10.1111/j.1749-6632.2011.05981.x

Kaas, Jon H. (2012, June 26). Evolution of columns, modules, and domains in the neocortex of primates. Proceedings of the National Academy of Sciences of the United States of America, 109, 10655–10660. https://doi.org/10.1073/pnas.1201892109

Kaas, Jon H. (2013). The evolution of brains from early mammals to humans. Wiley Interdisciplinary Reviews: Cognitive Science, 4(1), 33–45. https://doi.org/10.1002/wcs.1206

Kaas, Jon H. (2019). The origin and evolution of neocortex: From early mammals to modern humans. In Progress in brain research (pp. 61–81). https://doi.org/10.1016/bs.pbr.2019.03.017

Kabadayi, C., & Osvath, M. (2017). Ravens parallel great apes in flexible planning for tool-use and bartering. Science, 357(6347), 202–204. https://doi.org/10.1126/science.aam8138

Kalenscher, T., Diekamp, B., & Gunturkun, O. (2003). Neural architecture of choice behaviour in a concurrent interval schedule. European Journal of Neuroscience, 18(9), 2627–2637. https://doi.org/10.1046/j.1460- 9568.2003.03006.x

Kalenscher, T., Güntürkün, O., Calabrese, P., Gehlen, W., Kalt, T., & Diekamp, B. (2005).Neural correlates of a default response in a delayed go/no-go task. Journal of the Experimental Analysis of Behavior, 84(3), 521– 535. https://doi.org/10.1901/jeab.2005.86-04

Kalenscher, T., Windmann, S., Diekamp, B., Rose, J., Güntürkün, O., Colombo, M., … Colombo, M. (2005). Single units in the pigeon brain integrate reward amount and time-to-reward in an impulsive choice task. Current Biology, 15(7), 594–602. https://doi.org/10.1016/j.cub.2005.02.052

Karakuyu, D., Diekamp, B., & Güntürkün, O. (2003). Comparative neurochemistry of the avian forebrain and striatum: a microdialysis study. Neuroscience Research Communications, 33(2), 139–146. https://doi.org/10.1002/nrc.10090

Karten, H. J. (1969). The organization of the avian telencephalon and some speculations on the phylogeny of the amniote telencephalon. Annals of the New York Academy of Sciences, 167(1), 164–179. https://doi.org/10.1111/j.1749-6632.1969.tb20442.x

Karten, H. J. (2013). Neocortical Evolution: Neuronal Circuits Arise Independently of Lamination. Current Biology, 23(1), R12–R15. https://doi.org/10.1016/j.cub.2012.11.013

149

Karten, H. J., Karten, H. J., & Hodos, W. (1967). Stereotaxic atlas of the brain of the pigeon (Columba livia).

Kasthuri, N., Hayworth, K. J., Berger, D. R., Schalek, R. L., Conchello, J. A., Knowles-Barley, S., … Lichtman, J. W. (2015). Saturated Reconstruction of a Volume of Neocortex. Cell, 162(3), 648–661. https://doi.org/10.1016/j.cell.2015.06.054

Kemp, T. S. (2006). The origin and early radiation of the therapsid mammal-like reptiles: a palaeobiological hypothesis. Journal of Evolutionary Biology, 19(4), 1231–1247. https://doi.org/10.1111/j.1420- 9101.2005.01076.x

Kerly, M., & Leaback, D. H. (1957). The hexose specificity of hexokinase in brain from various species. The Biochemical Journal, 67(2), 250–252. https://doi.org/10.1042/bj0670250

Kilgore, D. L., Bernstein, M. H., & Hudson, D. M. (1976). Brain temperatures in birds. Journal of Comparative Physiology ■ B, 110(2), 209–215. https://doi.org/10.1007/BF00689309

Kitt, C. A., & Brauth, S. E. (1986). Telencephalic projections from midbrain and isthmal cell groups in the pigeon. II. The nigral complex. The Journal of Comparative Neurology, 247(1), 92–110. https://doi.org/10.1002/cne.902470106

Klyachko, V. A., & Stevens, C. F. (2003). Connectivity optimization and the positioning of cortical areas. Proceedings of the National Academy of Sciences of the United States of America, 100(13), 7937–7941. https://doi.org/10.1073/pnas.0932745100

Koenen, C., Millar, J., & Colombo, M. (2013). How bad do you want it? Reward modulation in the avian nidopallium caudolaterale. Behavioral Neuroscience, 127(4), 544–554. https://doi.org/10.1037/a0033551

Kono, T., Nishida, M., Nishiki, Y., Seki, Y., Sato, K., & Akiba, Y. (2005). Characterisation of glucose transporter (GLUT) gene expression in broiler chickens. British Poultry Science, 46(4), 510–515. https://doi.org/10.1080/00071660500181289

Koteja, P. (2004). The evolution of concepts on the evolution of endothermy in birds and mammals. Physiological and Biochemical Zoology, 77(6), 1043–1050. https://doi.org/10.1086/423741

Krohn, K. A., Muzi, M., & Spence, A. M. (2007). What is in a number? The FDG lumped constant in the rat brain. Journal of Nuclear Medicine : Official Publication, Society of Nuclear Medicine, 48(1), 5–7. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/17204692

Kröner, S., & Güntürkün, O. (1999). Afferent and efferent connections of the caudolateral neostriatum in the pigeon (Columba uvia): A retro- and anterograde pathway tracing study. Journal of Comparative Neurology, 407(2), 228–260. https://doi.org/10.1002/(SICI)1096-9861(19990503)407:2<228::AID-CNE6>3.0.CO;2-2

Krupenye, C., Kano, F., Hirata, S., Call, J., & Tomasello, M. (2016). Great apes anticipate that other individuals will act according to false beliefs. Science , 354(6308), 110–114. https://doi.org/10.1126/science.aaf8110

Ksepka, D. T., Balanoff, A. M., Smith, N. A., Bever, G. S., Bhullar, B. A. S., Bourdon, E., … Smaers, J. B. (2020). Tempo and Pattern of Avian Brain Size Evolution. Current Biology, 30(11), 2026-2036.e3. https://doi.org/10.1016/j.cub.2020.03.060

Kuenzel, W. J., Medina, L., Csillag, A., Perkel, D. J., & Reiner, A. (2011, November 18). The avian subpallium: New insights into structural and functional subdivisions occupying the lateral subpallial wall and their embryological origins. Brain Research, 1424, 67–101. https://doi.org/10.1016/j.brainres.2011.09.037

150

Kuntner, C. (2014). Kinetic modeling in pre-clinical positron emission tomography. Zeitschrift Fur Medizinische Physik, 24(4), 274–285. https://doi.org/10.1016/j.zemedi.2014.02.003

Kuwabara, H., Evans, A. C., & Gjedde, A. (1990). Michaelis-Menten constraints improved cerebral glucose metabolism and regional lumped constant measurements with [18F]fluorodeoxyglucose. Journal of Cerebral Blood Flow and Metabolism, 10(2), 180–189. https://doi.org/10.1038/jcbfm.1990.33

Lanuza, E. (1998). Identification of the reptilian basolateral amygdala: An anatomical investigation of the afferents to the posterior dorsal ventricular ridge of the lizard Podarcis hispanica. European Journal of Neuroscience, 10(11), 3517–3534. https://doi.org/10.1046/j.1460-9568.1998.00363.x

Lasiewski, R. C., & Calder, W. A. (1971). A preliminary allometric analysis of respiratory variables in resting birds. Respiration Physiology, 11(2), 152–166. https://doi.org/10.1016/0034-5687(71)90020-X

Laude, J. R., Stagner, J. P., Rayburn-Reeves, R., & Zentall, T. R. (2014). Midsession reversals with pigeons: Visual versus spatial discriminations and the intertrial interval. Learning and Behavior, 42(1), 40–46. https://doi.org/10.3758/s13420-013-0122-x

Laughlin, S. B. (2001). Energy as a constraint on the coding and processing of sensory information. Current Opinion in Neurobiology, 11, 475–480. https://doi.org/10.1016/S0959-4388(00)00237-3

Laughlin, S. B., & Sejnowski, T. J. (2003). Communication in neuronal networks. Science, 301, 1870–1874. https://doi.org/10.1126/science.1089662

LeBlanc, M. M., Goode, C. T., MacDougall-Shackleton, E. A., & Maney, D. L. (2007). Estradiol modulates brainstem catecholaminergic cell groups and projections to the auditory forebrain in a female songbird. Brain Research, 1171, 93–103. https://doi.org/10.1016/J.BRAINRES.2007.06.086

Lee, M. S. Y., Cau, A., Naish, D., & Dyke, G. J. (2014). Sustained miniaturization and anatomical innovation in the dinosaurian ancestors of birds. Science, 345(6196), 562–566. https://doi.org/10.1126/science.1252243

Lefebvre, L., Reader, S. M., & Sol, D. (2004). Brains, innovations and evolution in birds and primates. Brain, Behavior and Evolution, 63(4), 233–246. https://doi.org/10.1159/000076784

Lefèvre, C. M., Sharp, J. A., & Nicholas, K. R. (2010). Evolution of Lactation: Ancient Origin and Extreme Adaptations of the Lactation System. Annual Review of Genomics and Human Genetics, 11(1), 219–238. https://doi.org/10.1146/annurev-genom-082509-141806

Legendre, L. J., Gucrossed D Signnard, G., Botha-Brink, J., & Cubo, J. (2016). Palaeohistological Evidence for Ancestral High Metabolic Rate in Archosaurs. Systematic Biology, 65(6), 989–996. https://doi.org/10.1093/sysbio/syw033

Leising, K. J., Elmore, L. C., Rivera, J. J., Magnotti, J. F., Katz, J. S., & Wright, A. A. (2013). Testing visual short- term memory of pigeons (Columba livia) and a rhesus monkey (Macaca mulatta) with a location change detection task. Animal Cognition, 16(5), 839–844. https://doi.org/10.1007/s10071-013-0644-9

Lengersdorf, D., Marks, D., Uengoer, M., Stüttgen, M. C., & Güntürkün, O. (2015). Blocking NMDA-receptors in the pigeon’s “prefrontal” caudal nidopallium impairs appetitive extinction learning in a sign-tracking paradigm. Frontiers in Behavioral Neuroscience, 9, 85. https://doi.org/10.3389/fnbeh.2015.00085

151

Lengersdorf, D., Pusch, R., Güntürkün, O., & Stüttgen, M. C. (2014). Neurons in the pigeon nidopallium caudolaterale signal the selection and execution of perceptual decisions. European Journal of Neuroscience, 40(9), 3316–3327. https://doi.org/10.1111/ejn.12698

Lengersdorf, D., Stüttgen, M. C., Uengoer, M., & Güntürkün, O. (2014). Transient inactivation of the pigeon hippocampus or the nidopallium caudolaterale during extinction learning impairs extinction retrieval in an appetitive conditioning paradigm. Behavioural Brain Research, 265, 93–100. https://doi.org/10.1016/J.BBR.2014.02.025

Lennie, P. (2003). The cost of cortical computation. Current Biology, 13(6), 493–497. https://doi.org/10.1016/S0960-9822(03)00135-0

Letzner, S., Simon, A., & Güntürkün, O. (2016a). Connectivity and neurochemistry of the commissura anterior of the pigeon ( Columba livia ). Journal of Comparative Neurology, 524(2), 343–361. https://doi.org/10.1002/cne.23858

Leutgeb, S., Husband, S., Riters, L. V., Shimizu, T., & Bingman, V. P. (1996). Telencephalic afferents to the caudolateral neostriatum of the pigeon. Brain Research, 730(1–2), 173–181. https://doi.org/10.1016/S0006- 8993(96)00444-1

Levenson, R. M., Krupinski, E. A., Navarro, V. M., & Wasserman, E. A. (2015). Pigeons (Columba livia) as Trainable Observers of Pathology and Radiology Breast Cancer Images. PLOS ONE, 10(11), e0141357. https://doi.org/10.1371/journal.pone.0141357

Levy, W. B., & Baxter, R. A. (1996). Energy Efficient Neural Codes. Neural Computation, 8(3), 531–543. https://doi.org/10.1162/neco.1996.8.3.531

Lind, J., Enquist, M., & Ghirlanda, S. (2015). Animal memory: A review of delayed matching-to-sample data. Behavioural Processes, 117, 52–58. https://doi.org/10.1016/j.beproc.2014.11.019

Lindvall, O., Björklund, A., & Divac, I. (1978). Organization of catecholamine neurons projecting to the frontal cortex in the rat. Brain Research, 142(1), 1–24. https://doi.org/10.1016/0006-8993(78)90173-7

Lissek, S, Diekamp, B., & Güntürkün, O. (2002). Impaired learning of a color reversal task after NMDA receptor blockade in the pigeon (Columba livia) associative forebrain (neostriatum caudolaterale). Behavioral Neuroscience, 116(4), 523–529. https://doi.org/10.1037//0735-7044.116.4.523

Lissek, Silke, & Güntürkün, O. (2003). Dissociation of extinction and behavioral disinhibition: the role of NMDA receptors in the pigeon associative forebrain during extinction. The Journal of Neuroscience : The Official Journal of the Society for Neuroscience, 23(22), 8119–8124. https://doi.org/10.1523/JNEUROSCI.23-22- 08119.2003

Lissek, Silke, & Güntürkün, O. (2004). Maintenance in working memory or response selection?: Functions of NMDA receptors in the pigeon “prefrontal cortex.” Behavioural Brain Research, 153(2), 497–506. https://doi.org/10.1016/J.BBR.2004.01.007

Lissek, Silke, & Güntürkün, O. (2005). Out of Context: NMDA Receptor Antagonism in the Avian “Prefrontal Cortex” Impairs Context Processing in a Conditional Discrimination Task. Behavioral Neuroscience, 119(3), 797–805. https://doi.org/10.1037/0735-7044.119.3.797

152

Lohman, A. H. M., & Smeets, W. J. A. J. (1991). The Dorsal Ventricular Ridge and Cortex of Reptiles in Historical and Phylogenetic Perspective. In The Neocortex (pp. 59–74). https://doi.org/10.1007/978-1-4899-0652-6_7

Lovegrove, B. G. (2017). A phenology of the evolution of endothermy in birds and mammals. Biological Reviews, 92(2), 1213–1240. https://doi.org/10.1111/brv.12280

Lu, X. C. M., Slotnick, B. M., & Silberberg, A. M. (1993). Odor matching and odor memory in the rat. Physiology and Behavior, 53(4), 795–804. https://doi.org/10.1016/0031-9384(93)90191-H

Lucas, S. G., & Tanner, L. H. (2018). The Missing Mass Extinction at the Triassic-Jurassic Boundary. In The Late Triassic World (pp. 721-785). https://doi.org/10.1007/978-3-319-68009-5_15

Luo, Z. X. (2007). Transformation and diversification in early mammal evolution. Nature, 450, 1011–1019. https://doi.org/10.1038/nature06277

Macphail, E. (1982). Brain and intelligence in vertebrates. Oxford: Clarendon Press.

Maina, J. N. (1984). Morphometrics of the avian lung. 3. The structural design of the passerine lung. Respiration Physiology, 55(3), 291–307. https://doi.org/10.1016/0034-5687(84)90052-5

Malatesta, M., Mori, G., Acquotti, D., Campanini, B., Peracchi, A., Antin, P. B., & Percudani, R. (2020). Birth of a pathway for sulfur metabolism in early amniote evolution. Nature Ecology and Evolution, 1–8. https://doi.org/10.1038/s41559-020-1232-4

Mandelblat-Cerf, Y., & Fee, M. S. (2014). An Automated Procedure for Evaluating Song Imitation. PLoS ONE, 9(5), e96484. https://doi.org/10.1371/journal.pone.0096484

Mandelblat-Cerf, Y., Las, L., Denisenko, N., & Fee, M. S. (2014). A role for descending auditory cortical projections in songbird vocal learning. ELife, 3. https://doi.org/10.7554/eLife.02152

Maney, D., & Pinaud, R. (2011). Estradiol-dependent modulation of auditory processing and selectivity in songbirds. Frontiers in Neuroendocrinology, 32(3), 287–302. https://doi.org/10.1016/j.yfrne.2010.12.002

Marino, L. (2017). Thinking chickens: a review of cognition, emotion, and behavior in the domestic chicken. Animal Cognition, 20(2), 127–147. https://doi.org/10.1007/s10071-016-1064-4

Martin-Ordas, G., Haun, D., Colmenares, F., & Call, J. (2010). Keeping track of time: evidence for episodic-like memory in great apes. Animal Cognition, 13(2), 331–340. https://doi.org/10.1007/s10071-009-0282-4

Martin Wild, J. (1987). The avian somatosensory system: connections of regions of body representation in the forebrain of the pigeon. Brain Research, 412(2), 205–223. https://doi.org/10.1016/0006-8993(87)91127-9

Martin Wild, J., Karten, H. J., & Frost, B. J. (1993). Connections of the auditory forebrain in the pigeon (columba livia). The Journal of Comparative Neurology, 337(1), 32–62. https://doi.org/10.1002/cne.903370103

Mascalzoni, E., Regolin, L., & Vallortigara, G. (2010). Innate sensitivity for self-propelled causal agency in newly hatched chicks. Proceedings of the National Academy of Sciences of the United States of America, 107(9), 4483–4485. https://doi.org/10.1073/pnas.0908792107

Matragrano, L. L., Sanford, S. E., Salvante, K. G., Sockman, K. W., & Maney, D. L. (2011). Estradiol-dependent catecholaminergic innervation of auditory areas in a seasonally breeding songbird. European Journal of Neuroscience, 34(3), 416–425. https://doi.org/10.1111/j.1460-9568.2011.07751.x

153

McDougall, S., Riad, W. V., Silva-Gotay, A., Tavares, E. R., Harpalani, D., Li, G. L., & Richardson, H. N. (2018). Myelination of axons corresponds with faster transmission speed in the prefrontal cortex of developing male rats. ENeuro, 5(4). https://doi.org/10.1523/ENEURO.0203-18.2018

McGrew, W. C. (2004). Primatology: Advanced ape technology. Current Biology, 14(24), 1046–1047. https://doi.org/10.1016/j.cub.2004.11.042

McNab, B.K. (1994). Energy Conservation and the Evolution of Flightlessness in Birds. The American Naturalist, 144(4), 628–642.

McNab, Brian K. (1966). An Analysis of the Body Temperatures of Birds. The Condor, 68(1), 47–55. https://doi.org/10.2307/1365174

McNab, Brian K., & Salisbury, C. A. (1995). Energetics of new zealand’s temperate parrots. New Zealand Journal of Zoology, 22(3), 339–349. https://doi.org/10.1080/03014223.1995.9518050

McNab, Brian Keith. (2009). Ecological factors affect the level and scaling of avian BMR. Comparative Biochemistry and Physiology - A Molecular and Integrative Physiology, 152(1), 22–45. https://doi.org/10.1016/j.cbpa.2008.08.021

Medina, L., Abellán, A., Vicario, A., Castro-Robles, B., & Desfilis, E. (2016). The Amygdala. In Evolution of Nervous Systems: Second Edition (pp. 427–478). https://doi.org/10.1016/B978-0-12-804042-3.00019-1

Medina, Loreta, & Abellán, A. (2009). Development and evolution of the pallium. Seminars in Cell and Developmental Biology, 20, 698–711. https://doi.org/10.1016/j.semcdb.2009.04.008

Medina, Loreta, & Reiner, A. (2000). Do birds possess homologues of mammalian primary visual, somatosensory and motor cortices? Trends in Neurosciences, 23, 1–12. https://doi.org/10.1016/S0166-2236(99)01486-1

Mehlhorn, J., Hunt, G. R., Gray, R. D., Rehkämper, G., & Güntürkün, O. (2010). Tool-Making New Caledonian Crows Have Large Associative Brain Areas. Brain, Behavior and Evolution, 75(1), 63–70. https://doi.org/10.1159/000295151

Meléndez-Morales, D., de Paz-Lugo, P., & Meléndez-Hevia, E. (2009). Glycolysis activity in flight muscles of birds according to their physiological function. An experimental model in vitro to study aerobic and anaerobic glycolysis activity separately. Molecular and Cellular Biochemistry, 328(1–2), 127–135. https://doi.org/10.1007/s11010-009-0082-9

Mello, C. V., Kaser, T., Buckner, A. A., Wirthlin, M., & Lovell, P. V. (2019). Molecular architecture of the zebra finch arcopallium. Journal of Comparative Neurology, 527(15), 2512–2556. https://doi.org/10.1002/cne.24688

Mello, C. V., Vates, E., Okuhata, S., & Nottebohm, F. (1998). Descending auditory pathways in the adult male zebra finch (Taeniopygia Guttata). The Journal of Comparative Neurology, 395(2), 137–160. https://doi.org/10.1002/(SICI)1096-9861(19980601)395:2<137::AID-CNE1>3.0.CO;2-3

Merker, B. (1983). Silver staining of cell bodies by means of physical development. Journal of Neuroscience Methods, 9(3), 235–241. https://doi.org/10.1016/0165-0270(83)90086-9

Mess, A., Blackburn, D. G., & Zeller, U. (2003). Evolutionary transformations of fetal membranes and reproductive strategies. Journal of Experimental Zoology, 299A(1), 3–12. https://doi.org/10.1002/jez.a.10287

154

Metzger, M., Jiang, S., & Braun, K. (2002). A quantitative immuno-electron microscopic study of dopamine terminals in forebrain regions of the domestic chick involved in filial imprinting. Neuroscience, 111(3), 611– 623. https://doi.org/10.1016/S0306-4522(01)00611-X

Metzger, M., Jiang, S., & Braun, K. (1998). Organization of the dorsocaudal neostriatal complex: A retrograde and anterograde tracing study in the domestic chick with special emphasis on pathways relevant to imprinting. Journal of Comparative Neurology, 395(3), 380–404. https://doi.org/10.1002/(SICI)1096- 9861(19980808)395:3<380::AID-CNE8>3.0.CO;2-Z

Metzger, M., Jiang, S., Wang, J., & Braun, K. (1996). Organization of the dopaminergic innervation of forebrain areas relevant to learning: A combined immunohistochemical/retrograde tracing study in the domestic chick. The Journal of Comparative Neurology, 376(1), 1–27. https://doi.org/10.1002/(SICI)1096- 9861(19961202)376:1<1::AID-CNE1>3.0.CO;2-7

Micheva, K. D., Wolman, D., Mensh, B. D., Pax, E., Buchanan, J., Smith, S. J., & Bock, D. D. (2016). A large fraction of neocortical myelin ensheathes axons of local inhibitory neurons. ELife, 5. https://doi.org/10.7554/eLife.15784

Miller, E. K., & Cohen, J. D. (2001). An Integrative Theory of Prefrontal Cortex Function. Annual Review of Neuroscience, 24, 167–170.

Miller, E. K., Erickson, C. A., & Desimone, R. (1996). Neural mechanisms of visual working memory in prefrontal cortex of the macaque. The Journal of Neuroscience, 16(16), 5154–5167. https://doi.org/10.1523/JNEUROSCI.16-16-05154.1996

Mink, J. W., Blumenschine, R. J., & Adams, D. B. (1981a). Ratio of central nervous system to body metabolism in vertebrates: its constancy and functional basis. The American Journal of Physiology, 241(3), R203-12. https://doi.org/10.1152/ajpregu.1981.241.3.R203

Mogensen, J., & Divac, I. (1982). The Prefrontal “Cortex” in the Pigeon. Brain, Behavior and Evolution, 21(3), 60– 66. https://doi.org/10.1159/000121617

Moll, F. W., & Nieder, A. (2015). Cross-modal associative mnemonic signals in crow endbrain neurons. Current Biology, 25(16), 2196–2201. https://doi.org/10.1016/j.cub.2015.07.013

Moll, F. W., & Nieder, A. (2017). Modality-invariant audio-visual association coding in crow endbrain neurons. Neurobiology of Learning and Memory, 137, 65–76. https://doi.org/10.1016/j.nlm.2016.11.011

Montiel, J. F., & Aboitiz, F. (2018). Homology in Amniote Brain Evolution: The Rise of Molecular Evidence. Brain, Behavior and Evolution, 91(2), 59–64. https://doi.org/10.1159/000489116

Moons, L., van Gils, J., Ghijsels, E., & Vandesande, F. (1994). Immunocytochemical localization of L-DOPA and dopamine in the brain of the chicken (Gallus domesticus). The Journal of Comparative Neurology, 346(1), 97–118. https://doi.org/10.1002/cne.903460107

Moore, J. W., Joyner, R. W., Brill, M. H., Waxman, S. D., & Najar-Joa, M. (1978). Simulations of conduction in uniform myelinated fibers. Relative sensitivity to changes in nodal and internodal parameters. Biophysical Journal, 21(2), 147–160. https://doi.org/10.1016/S0006-3495(78)85515-5

155

Morand-Ferron, J., Cole, E. F., & Quinn, J. L. (2016). Studying the evolutionary ecology of cognition in the wild: A review of practical and conceptual challenges. Biological Reviews, 91(2), 367–389. https://doi.org/10.1111/brv.12174

Mota, B., & Herculano-Houzel, S. (2014). All brains are made of this: a fundamental building block of brain matter with matching neuronal and glial . Frontiers in Neuroanatomy, 8, 127. https://doi.org/10.3389/fnana.2014.00127

Mrozek, S., Vardon, F., & Geeraerts, T. (2012). Brain temperature: Physiology and pathophysiology after brain injury. Anesthesiology Research and Practice, 2012. https://doi.org/10.1155/2012/989487

Mulcahy, D. G., Noonan, B. P., Moss, T., Townsend, T. M., Reeder, T. W., Sites, J. W., & Wiens, J. J. (2012). Estimating divergence dates and evaluating dating methods using phylogenomic and mitochondrial data in squamate reptiles. Molecular Phylogenetics and Evolution, 65(3), 974–991. https://doi.org/10.1016/j.ympev.2012.08.018

Mulcahy, N. J., & Call, J. (2006). Apes save tools for future use. Science, 312(5776), 1038–1040. https://doi.org/10.1126/science.1125456

Müller, N. G., & Knight, R. T. (2006). The functional neuroanatomy of working memory: Contributions of human brain lesion studies. Neuroscience, 139(1), 51–58. https://doi.org/10.1016/J.NEUROSCIENCE.2005.09.018

Nagy, K. A., Girard, I. A., & Brown, T. K. (1999). Energetics of free-ranging mammals, reptiles, and birds. Annual Review of Nutrition, 19(1), 247-277.

Naumann, R. K., Ondracek, J. M., Reiter, S., Shein-Idelson, M., Tosches, M. A., Yamawaki, T. M., & Laurent, G. (2015). The reptilian brain. Current Biology, 25, R317–R321. https://doi.org/10.1016/j.cub.2015.02.049

Navarrete, A., Van Schaik, C. P., & Isler, K. (2011). Energetics and the evolution of human brain size. Nature, 480(7375), 91–93. https://doi.org/10.1038/nature10629

Nealen, P. M., & Ricklefs, R. E. (2001). Early diversification of the avian brain:body relationship. Journal of Zoology, 253(3), S095283690100036X. https://doi.org/10.1017/S095283690100036X

Nespolo, R. F., Bacigalupe, L. D., Figueroa, C. C., Koteja, P., & Opazo, J. C. (2011). Using new tools to solve an old problem: The evolution of endothermy in vertebrates. Trends in Ecology and Evolution, 26, 414–423. https://doi.org/10.1016/j.tree.2011.04.004

Nespolo, R. F., Gonzaìlez-Lagos, C., Solano-Iguaran, J. J., Elfwing, M., Garitano-Zavala, A., Mañosa, S., … Altimiras, J. (2018). Aerobic power and flight capacity in birds: A phylogenetic test of the heart-size hypothesis. Journal of Experimental Biology, 221(1). https://doi.org/10.1242/jeb.162693

Ngwenya, A., Patzke, N., Manger, P. R., & Herculano-Houzel, S. (2016). Continued Growth of the Central Nervous System without Mandatory Addition of Neurons in the Nile Crocodile (Crocodylus niloticus). Brain, Behavior and Evolution, 87(1), 19–38. https://doi.org/10.1159/000443201

Nieder, A. (2017). Inside the corvid brain—probing the physiology of cognition in crows. Current Opinion in Behavioral Sciences, 16, 8–14. https://doi.org/10.1016/j.cobeha.2017.02.005

Nieder, A., Wagener, L., & Rinnert, P. (2020). A neural correlate of sensory consciousness in a corvid bird. Science, 369(6511), 1626–1629. https://doi.org/10.1126/science.abb1447

156

Niven, J. E., Anderson, J. C., & Laughlin, S. B. (2007). Fly Photoreceptors Demonstrate Energy-Information Trade-Offs in Neural Coding. PLoS Biology, 5(4), e116. https://doi.org/10.1371/journal.pbio.0050116

Niven, J. E., & Laughlin, S. B. (2008). Energy limitation as a selective pressure on the evolution of sensory systems. The Journal of Experimental Biology, 211(11), 1792–1804. https://doi.org/10.1242/jeb.017574

Nixdorf-Bergweiler, B. E., & Bischof, H.-J. (2007). A stereotaxic atlas of the brain of the zebra finch, Taeniopygia guttata, with special emphasis on telencephalic visual and song system nuclei in transverse and sagittal sections. Bethesda (MD).

Noda, A., Takamatsu, H., Minoshima, S., Tsukada, H., & Nishimura, S. (2003). Determination of Kinetic Rate Constants for 2-[ 18F]fluoro-2-deoxy-D-glucose and Partition Coefficient of Water in Conscious Macaques and Alterations in Aging or Anesthesia Examined on Parametric Images with an Anatomic Standardization Technique. Journal of Cerebral Blood Flow and Metabolism, 23(12), 1441–1447. https://doi.org/10.1097/01.WCB.0000090623.86921.47

Nomura, T., Murakami, Y., Gotoh, H., & Ono, K. (2014). Reconstruction of ancestral brains: Exploring the evolutionary process of encephalization in amniotes. Neuroscience Research, 86, 25–36. https://doi.org/10.1016/j.neures.2014.03.004

Northcutt, R. G. (2013). Variation in Reptilian Brains and Cognition. Brain, Behavior and Evolution, 82(1), 45–54. https://doi.org/10.1159/000351996

Nottebohm, F. (1972). The Origins of Vocal Learning. The American Naturalist, 106(947), 116–140. https://doi.org/10.1086/282756

O’Connor, J., Chiappe, L. M., & Bell, A. (2011). Pre-modern Birds: Avian Divergences in the Mesozoic. In Living Dinosaurs (pp. 39–114). https://doi.org/10.1002/9781119990475.ch3

Olkowicz, S., Kocourek, M., Luèan, R. K., Porteš, M., Fitch, W. T., Herculano-Houzel, S., … Němec, P. (2016). Birds have primate-like numbers of neurons in the forebrain. Proceedings of the National Academy of Sciences of the United States of America, 113(26), 7255–7260. https://doi.org/10.1073/pnas.1517131113

Osvath, M., & Osvath, H. (2008). Chimpanzee (Pan troglodytes) and orangutan (Pongo abelii) forethought: Self- control and pre-experience in the face of future tool use. Animal Cognition, 11(4), 661–674. https://doi.org/10.1007/s10071-008-0157-0

Ott, T., & Nieder, A. (2019). Dopamine and Cognitive Control in Prefrontal Cortex. Trends in Cognitive Sciences, 23(3), 213–234. https://doi.org/10.1016/J.TICS.2018.12.006

Passingham, R., & Wise, S. (2012). The neurobiology of the prefrontal cortex: anatomy, evolution, and the origin of insight. Oxford: Oxford University Press.

Paterson, A. K., & Bottjer, S. W. (2017). Cortical inter-hemispheric circuits for multimodal vocal learning in songbirds. Journal of Comparative Neurology, 525(15), 3312–3340. https://doi.org/10.1002/cne.24280

Patlak, C. S., & Blasberg, R. G. (1985). Graphical evaluation of blood-to-brain transfer constants from multiple- time uptake data. Generalizations. Journal of Cerebral Blood Flow and Metabolism, 5(4), 584–590. https://doi.org/10.1038/jcbfm.1985.87

157

Pepperberg, I. M., & Funk, M. S. (1990). Object permanence in four species of psittacine birds: An African Grey parrot (Psittacus erithacus), an Illiger mini macaw (Ara maracana), a parakeet (Melopsittacus undulatus), and a cockatiel (Nymphicus hollandicus). Animal Learning & Behavior, 18(1), 97–108. https://doi.org/10.3758/BF03205244

Perge, J. A., Niven, J. E., Mugnaini, E., Balasubramanian, V., & Sterling, P. (2012). Why do axons differ in caliber? Journal of Neuroscience, 32(2), 626–638. https://doi.org/10.1523/JNEUROSCI.4254-11.2012

Petkov, C. I., & Jarvis, E. D. (2012). Birds, primates, and spoken language origins: Behavioral phenotypes and neurobiological substrates. Frontiers in Evolutionary Neuroscience, 4, 12. https://doi.org/10.3389/fnevo.2012.00012

Pfenning, A. R., Hara, E., Whitney, O., Rivas, M. V., Wang, R., Roulhac, P. L., … Jarvis, E. D. (2014). Convergent transcriptional specializations in the brains of humans and song-learning birds. Science, 346(6215). https://doi.org/10.1126/science.1256846

Phelps, M. E., Huang, S. C., Hoffman, E. J., Selin, C., Sokoloff, L., & Kuhl, D. E. (1979). Tomographic measurement of local cerebral glucose metabolic rate in humans with (F‐18)2‐fluoro‐2‐deoxy‐D‐glucose: Validation of method. Annals of Neurology, 6(5), 371–388. https://doi.org/10.1002/ana.410060502

Pinaud, R., & Terleph, T. A. (2008). A songbird forebrain area potentially involved in auditory discrimination and memory formation. Journal of Biosciences, 33(1), 145–155. https://doi.org/10.1007/s12038-008-0030-y

Pinshow, B., Bernstein, M. H., Lopez, G. E., & Kleinhaus, S. (1982). Regulation of brain temperature in pigeons: Effects of corneal convection. American Journal of Physiology - Regulatory Integrative and Comparative Physiology, 11(3). https://doi.org/10.1152/ajpregu.1982.242.5.r577

Pollok, B., Prior, H., & Güntürkün, O. (2000). Development of object permanence in food-storing magpies (Pica pica). Journal of Comparative Psychology, 114(2), 148–157. https://doi.org/10.1037/0735-7036.114.2.148

Pontzer, H. (2017). Economy and Endurance in Human Evolution. Current Biology, 27, R613–R621. https://doi.org/10.1016/j.cub.2017.05.031

Pontzer, H., Brown, M. H., Raichlen, D. A., Dunsworth, H., Hare, B., Walker, K., … Ross, S. R. (2016). Metabolic acceleration and the evolution of human brain size and life history. Nature, 533(7603), 390–392. https://doi.org/10.1038/nature17654

Pozner, T., Vistoropsky, Y., Moaraf, S., Heiblum, R., & Barnea, A. (2018). Questioning Seasonality of Neuronal Plasticity in the Adult Avian Brain. Scientific Reports, 8(1), 11289. https://doi.org/10.1038/s41598-018- 29532-1

Preuss, T. M. (1995). Do Rats Have Prefrontal Cortex? The Rose-Woolsey-Akert Program Reconsidered. Journal of Cognitive Neuroscience, 7(1), 1–24. https://doi.org/10.1162/jocn.1995.7.1.1

Prinzinger, R., Preßmar, A., & Schleucher, E. (1991). Body temperature in birds. Comparative Biochemistry and Physiology -- Part A: Physiology, 99, 499–506. https://doi.org/10.1016/0300-9629(91)90122-S

Pritz, M. B. (1974). Ascending connections of a thalamic auditory area in a crocodile, Caiman crocodilus. Journal of Comparative Neurology, 153(2), 199–213. https://doi.org/10.1002/cne.901530205

158

Pritz, M. B. (1975). Anatomical identification of a telencephalic visual area in crocodiles: Ascending connections of nucleus rotundus in Caiman crocodilus. Journal of Comparative Neurology, 164(3), 323–338. https://doi.org/10.1002/cne.901640305

Pritz, M. B. (2016). Thalamic reticular nucleus in Caiman crocodilus: Forebrain connections. Neuroscience Letters, 627, 65–70. https://doi.org/10.1016/j.neulet.2016.05.046

Pritz, M. B., & Stritzel, M. E. (1992). A second auditory area in the non-cortical telencephalon of a reptile. Brain Research, 569(1), 146–151. https://doi.org/10.1016/0006-8993(92)90381-I

Pritz, M. B., & Stritzel, M. E. (1994). Anatomical identification of a telenceplialic somatosensory area in a reptile, caiman crocodilus. Brain, Behavior and Evolution, 43(2), 107–127. https://doi.org/10.1159/000113628

Procyk, E., & Goldman-Rakic, P. S. (2006). Modulation of dorsolateral prefrontal delay activity during self- organized behavior. The Journal of Neuroscience, 26(44), 11313–11323. https://doi.org/10.1523/JNEUROSCI.2157-06.2006

Prum, R. O., Berv, J. S., Dornburg, A., Field, D. J., Townsend, J. P., Lemmon, E. M., & Lemmon, A. R. (2015). A comprehensive phylogeny of birds (Aves) using targeted next-generation DNA sequencing. Nature, 526(7574), 569–573. https://doi.org/10.1038/nature15697

Puelles, L. (2007). The chick brain in stereotaxic coordinates : an atlas featuring neuromeric subdivisions and mammalian homologies. Academic Press.

Puelles, L., Kuwana, E., Puelles, E., & Rubenstein, J. L. R. (1999). Comparison of the Mammalian and Avian Telencephalon from the Perspective of Gene Expression Data. European Journal of Morphology, 37(2–3), 139–150. https://doi.org/10.1076/ejom.37.2.139.4756

Puelles, L., Sandoval, J. E., Ayad, A., del Corral, R., Alonso, A., Ferran, J. L., & Martínez-de-la-Torre, M. (2017). The Pallium in Reptiles and Birds in the Light of the Updated Tetrapartite Pallium Model. In Evolution of Nervous Systems: Second Edition (pp. 519–555). https://doi.org/10.1016/B978-0-12-804042-3.00014-2

Puelles, L. (2014). Development and Evolution of the Claustrum. In The Claustrum: Structural, Functional, and Clinical Neuroscience (pp. 119–176). https://doi.org/10.1016/B978-0-12-404566-8.00004-0

Puelles, L., Kuwana, E., Puelles, E., Bulfone, A., Shimamura, K., Keleher, J., … Rubenstein, J. L. R. (2000). Pallial and subpallial derivatives in the embryonic chick and mouse telencephalon, traced by the expression of the genes Dlx-2, Emx-1, Nkx-2.1, Pax-6, and Tbr-1. Journal of Comparative Neurology, 424(3), 409–438. https://doi.org/10.1002/1096-9861(20000828)424:3<409::AID-CNE3>3.0.CO;2-7

Puig, M. V., Rose, J., Schmidt, R., & Freund, N. (2014). Dopamine modulation of learning and memory in the prefrontal cortex: Insights from studies in primates, rodents, and birds. Frontiers in Neural Circuits, 8, 93. https://doi.org/10.3389/fncir.2014.00093

Rattenborg, N. C., & Martinez-Gonzalez, D. (2011). A bird-brain view of episodic memory. Behavioural Brain Research, 222, 236–245. https://doi.org/10.1016/j.bbr.2011.03.030

Rayburn-Reeves, R. M., Stagner, J. P., Kirk, C. R., & Zentall, T. R. (2013). Reversal learning in rats (rattus norvegicus) and pigeons (columba livia): Qualitative differences in behavioral flexibility. Journal of Comparative Psychology, 127(2), 202–211. https://doi.org/10.1037/a0026311

159

Rayner, J. M. V. (1988). Form and Function in Avian Flight. In Current Ornithology (pp. 1–66). https://doi.org/10.1007/978-1-4615-6787-5_1

Redies, C., Kovjanic, D., Heyers, D., Medina, L., Hirano, S., Suzuki, S. T., & Puelles, L. (2002). Patch/matrix patterns of gray matter differentiation in the telencephalon of chicken and mouse. Brain Research Bulletin, 57(3–4), 489–493. https://doi.org/10.1016/S0361-9230(01)00680-3

Redies, C., & Puelles, L. (2001). Modularity in vertebrate brain development and evolution. BioEssays, 23(12), 1100–1111. https://doi.org/10.1002/bies.10014

Rehkämper, G., Frahm, H. D., & Zilles, K. (1991). Quantitative Development of Brain and Brain Structures in Birds (Galliformes and Passeriformes) Compared to that in Mammals (Insectivores and Primates) (Part 1 of 2). Brain, Behavior and Evolution, 37(3), 125–134. https://doi.org/10.1159/000114353

Rehkämper, G., & Zilles, K. (1991). Parallel evolution in mammalian and avian brains: comparative cytoarchitectonic and cytochemical analysis. Cell and Tissue Research, 263(1), 3–28. https://doi.org/10.1007/BF00318396

Reilly, S., & Elias, J. (1998). Locomotion in alligator mississippiensis: kinematic effects of speed and posture and their relevance to the sprawling-to-erect paradigm. Journal of Experimental Biology, 201(18).

Reiner, A. (2005). A new avian brain nomenclature: Why, how and what. Brain Research Bulletin, 66(4–6), 317– 331. https://doi.org/10.1016/j.brainresbull.2005.05.007

Reiner, A. (2016). The Conservative Evolution of the Vertebrate Basal Ganglia. In Handbook of Behavioral Neuroscience (Vol. 24, pp. 63–97). https://doi.org/10.1016/B978-0-12-802206-1.00004-0

Reiner, A., Perkel, D. J., Bruce, L. L., Butler, A. N. N. B., Kuenzel, W., Medina, L., … Durand, S. (2004). The Avian Brain Nomenclature Forum: Terminology for a New Century in Comparative Neuroanatomy. 1–8. https://doi.org/10.1002/cne.20119.The

Reiner, A., Perkel, D. J., Mello, C. V., & Jarvis, E. D. (2004). Songbirds and the revised avian brain nomenclature. Annals of the New York Academy of Sciences, 1016, 77–108. https://doi.org/10.1196/annals.1298.013

Reisz, R. R., & Modesto, S. P. (1996). Archerpeton anthracos from the Joggins Formation of Nova Scotia: a microsaur, not a reptile. Canadian Journal of Earth Sciences, 33(5), 703–709. https://doi.org/10.1139/e96- 053

Reisz, R. R. (1997). The origin and early evolutionary history of amniotes. Trends in Ecology and Evolution, 12, 218–222. https://doi.org/10.1016/S0169-5347(97)01060-4

Reisz, R. R. (2014). “Pelycosaur”-Grade Synapsids: Introduction. In Vertebrate Paleobiology and Paleoanthropology (pp. 3–5). https://doi.org/10.1007/978-94-007-6841-3_1

Reivich, M., Alavi, A., Wolf, A., Fowler, J., Russell, J., Arnett, C., … Anand, A. (1985). Glucose metabolic rate kinetic model parameter determination in humans: The lumped constants and rate constants for [18F]fluorodeoxyglucose and [11C]deoxyglucose. Journal of Cerebral Blood Flow and Metabolism, 5(2), 179–192. https://doi.org/10.1038/jcbfm.1985.24

Renne, P. R., Deino, A. L., Hilgen, F. J., Kuiper, K. F., Mark, D. F., Mitchell, W. S., … Smit, J. (2013). Time scales of critical events around the cretaceous-paleogene boundary. Science, 339(6120), 684–687. https://doi.org/10.1126/science.1230492

160

Rezende, E. L., Bacigalupe, L. D., Nespolo, R. F., & Bozinovic, F. (2020). Shrinking dinosaurs and the evolution of endothermy in birds. Science Advances, 6(1), eaaw4486. https://doi.org/10.1126/sciadv.aaw4486

Ricklefs, R. E. (2004). The cognitive face of avian life histories. The Wilson Bulletin, 116(2), 119–133. https://doi.org/10.1676/04-054

Rigo, M., Onoue, T., Tanner, L. H., Lucas, S. G., Godfrey, L., Katz, M. E., … Bertinelli, A. (2020). The Late Triassic Extinction at the Norian/Rhaetian boundary: Biotic evidence and geochemical signature. Earth- Science Reviews, 204, 103180. https://doi.org/10.1016/j.earscirev.2020.103180

Rinnert, P., Kirschhock, M. E., & Nieder, A. (2019). Neuronal Correlates of Spatial Working Memory in the Endbrain of Crows. Current Biology, 29(16), 2616-2624.e4. https://doi.org/10.1016/J.CUB.2019.06.060

Rose, J., & Colombo, M. (2005). Neural correlates of executive control in the avian brain. PLoS Biology, 3(6), 1139–1146. https://doi.org/10.1371/journal.pbio.0030190

Rosser, B. W. C., & George, J. C. (1986). The avian pectoralis: histochemical characterization and distribution of muscle fiber types. Canadian Journal of Zoology, 64(5), 1174–1185. https://doi.org/10.1139/z86-176

Rosvold, H., & Szwarcbart, M. (1964). Neural structures involved in delayed-response performance. The Frontal Granular Cortex and Behaviour, 1–15.

Rowe, T. B., Macrini, T. E., & Luo, Z. X. (2011). Fossil evidence on origin of the mammalian brain. Science, 332(6032), 955–957. https://doi.org/10.1126/science.1203117

Rugani, R., Vallortigara, G., Priftis, K., & Regolin, L. (2015). Number-space mapping in the newborn chick resembles humans’ mental number line. Science, 347(6221), 534–536. https://doi.org/10.1126/SCIENCE.AAA1379

Sadananda, M., Korte, S., & Bischof, H.-J. (2007). Afferentation of a caudal forebrain area activated during courtship behavior: A tracing study in the zebra finch (Taeniopygia guttata). Brain Research, 1184, 108– 120. https://doi.org/10.1016/J.BRAINRES.2007.09.040

Sakurai, Y., & Sugimoto, S. (1986). Multiple unit activity of prefrontal cortex and dorsomedial thalamus during delayed go/no-go alternation in the rat. Behavioural Brain Research, 20(3), 295–301. https://doi.org/10.1016/0166-4328(86)90229-9

Salerno, M., Ferrer, E., Wei, S., Li, X., Gao, W., Ouellette, D., … Vaska, P. (2019). Behavioral neuroimaging in birds using PET. Journal of Neuroscience Methods, 317(June 2018), 157–164. https://doi.org/10.1016/j.jneumeth.2019.01.017

Sathyanesan, A., Ogura, T., & Lin, W. (2012). Automated measurement of nerve fiber density using line intensity scan analysis. Journal of Neuroscience Methods, 206(2), 165–175. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/22613744

Sayol, F. (2016). Relative Brain Size and Its Relation with the Associative Pallium in Birds. Brain, behavior and evolution, 87(2), 69–77. https://doi.org/10.1159/000444670

Sayol, F., Downing, P. A., Iwaniuk, A. N., Maspons, J., & Sol, D. (2018). Predictable evolution towards larger brains in birds colonizing oceanic islands. Nature Communications, 9(1), 1–7. https://doi.org/10.1038/s41467-018-05280-8

161

Scanes, C. G. (2015). Carbohydrate Metabolism. In Sturkie’s Avian Physiology: Sixth Edition (pp. 421–441). https://doi.org/10.1016/B978-0-12-407160-5.00018-X

Scarf, D., Hayne, H., & Colombo, M. (2011). Pigeons on par with primates in numerical competence. Science, 334(6063), 1664. https://doi.org/10.1126/science.1213357

Scarf, D., Miles, K., Sloan, A., Goulter, N., Hegan, M., Seid-Fatemi, A., … Colombo, M. (2011). Brain cells in the avian “prefrontal cortex” code for features of slot-machine-like gambling. PLoS ONE, 6(1), e14589. https://doi.org/10.1371/journal.pone.0014589

Schall, U., Gunturkun, O., & Delius, J. D. (1986). Sensory projections to the nucleus basalis prosencephali of the pigeon. Cell and Tissue Research, 245(3), 539–546. https://doi.org/10.1007/BF00218555

Schiffer, W. K., Mirrione, M. M., & Dewey, S. L. (2007). Optimizing experimental protocols for quantitative behavioral imaging with18F-FDG in rodents. Journal of Nuclear Medicine, 48(2), 277–287.

Schmidt-Hieber, C., & Bischofberger, J. (2010). Fast sodium channel gating supports localized and efficient axonal action potential initiation. Journal of Neuroscience, 30(30), 10233–10242. https://doi.org/10.1523/JNEUROSCI.6335-09.2010

Schmidt-Nielsen, K. (1972). Locomotion: Energy cost of swimming, flying, and running. Science, 177, 222–228. https://doi.org/10.1126/science.177.4045.222

Schnabel, R., Metzger, M., Jiang, S., Hemmings, H. C., Greengard, P., & Braun, K. (1997). Localization of dopamine D1 receptors and dopaminoceptive neurons in the chick forebrain. Journal of Comparative Neurology, 388(1), 146–168. https://doi.org/10.1002/(SICI)1096-9861(19971110)388:1<146::AID- CNE10>3.0.CO;2-T

Schoenbaum, G., Roesch, M. R., Stalnaker, T. A., & Takahashi, Y. K. (2009). A new perspective on the role of the orbitofrontal cortex in adaptive behaviour. Nature Reviews Neuroscience, 10, 885–892. https://doi.org/10.1038/nrn2753

Seebacher, F. (1999). Behavioural postures and the rate of body temperature change in wild freshwater crocodiles, Crocodylus johnstoni. Physiological and Biochemical Zoology, 72(1), 57–63. https://doi.org/10.1086/316638

Seed, A., Emery, N., & Clayton, N. (2009). Intelligence in Corvids and Apes: A Case of Convergent Evolution? Ethology, 115(5), 401–420. https://doi.org/10.1111/j.1439-0310.2009.01644.x

Sen, S., Parishar, P., Pundir, A. S., Reiner, A., & Iyengar, S. (2019). The expression of tyrosine hydroxylase and DARPP‐32 in the house crow ( Corvus splendens ) brain. Journal of Comparative Neurology, 527(11), 1801–1836. https://doi.org/10.1002/cne.24649

Sengupta, B., Faisal, A. A., Laughlin, S. B., & Niven, J. E. (2013). The effect of cell size and channel density on neuronal information encoding and energy efficiency. Journal of Cerebral Blood Flow and Metabolism, 33(9), 1465–1473. https://doi.org/10.1038/jcbfm.2013.103

Sengupta, B., Stemmler, M., Laughlin, S. B., & Niven, J. E. (2010). Action Potential Energy Efficiency Varies Among Neuron Types in Vertebrates and Invertebrates. PLoS Computational Biology, 6(7), e1000840. https://doi.org/10.1371/journal.pcbi.1000840

162

Senn, D. G., & Northcutt, R. G. (1973). The forebrain and midbrain of some squamates and their bearing on the origin of snakes. Journal of Morphology, 140(2), 135–151. https://doi.org/10.1002/jmor.1051400202

Sereno, P. C. (1999). The evolution of dinosaurs. Science, 284, 2137–2147. https://doi.org/10.1126/science.284.5423.2137

Seymour, R. S. (2016). Cardiovascular Physiology of Dinosaurs. Physiology, 31(6), 430–441. https://doi.org/10.1152/physiol.00016.2016

Seymour, R. S., Bennett-Stamper, C. L., Johnston, S. D., Carrier, D. R., & Grigg, G. C. (2004). Evidence for endothermic ancestors of crocodiles at the stem of archosaur evolution. Physiological and Biochemical Zoology, 77(6), 1051–1067. https://doi.org/10.1086/422766

Shanahan, M., Bingman, V. P., Shimizu, T., Wild, M., & Güntürkün, O. (2013). Large-scale network organisation in the avian forebrain: A connectivity matrix and theoretical analysis. Frontiers in Computational Neuroscience, 7(JUN), 1–17. https://doi.org/10.3389/fncom.2013.00089

Shen, S. Z., Crowley, J. L., Wang, Y., Bowring, S. A., Erwin, D. H., Sadler, P. M., … Jin, Y. G. (2011). Calibrating the end-Permian mass extinction. Science, 334(6061), 1367–1372. https://doi.org/10.1126/science.1213454

Shettleworth, S J. (2010). Clever animals and killjoy explanations in comparative psychology. Trends in Cognitive Sciences, 14(11).

Shettleworth, Sara J. (2001). Animal cognition and animal behaviour. Animal Behaviour, 61, 277–286. https://doi.org/10.1006/anbe.2000.1606

Shimizu, T., Cox, K., & Karten, H. J. (1995). Intratelencephalic projections of the visual wulst in pigeons (Columba livia). The Journal of Comparative Neurology, 359(4), 551–572. https://doi.org/10.1002/cne.903590404

Shu, Y., Yu, G., Yang, J., & McCormick, D. A. (2007). Selective control of cortical axonal spikes by a slowly inactivating K + current. Proceedings of the National Academy of Sciences of the United States of America, 104(27), 11453–11458. https://doi.org/10.1073/pnas.0702041104

Skov-Rackette, S. I., Miller, N. Y., & Shettleworth, S. J. (2006). What-where-when memory in pigeons. Journal of Experimental Psychology: Animal Behavior Processes, 32(4), 345–358. https://doi.org/10.1037/0097- 7403.32.4.345

Smaers, J. B., Dechmann, D. K. N., Goswami, A., Soligo, C., & Safi, K. (2012). Comparative analyses of evolutionary rates reveal different pathways to encephalization in bats, carnivorans, and primates. Proceedings of the National Academy of Sciences of the United States of America, 109(44), 18006–18011. https://doi.org/10.1073/pnas.1212181109

Smaers, J. B., Gómez-Robles, A., Parks, A. N., & Sherwood, C. C. (2017). Exceptional Evolutionary Expansion of Prefrontal Cortex in Great Apes and Humans. Current Biology, 27(5), 714–720. https://doi.org/10.1016/j.cub.2017.01.020

Smeets, W. J. A. J. A. J., & González, A. (2000). Catecholamine systems in the brain of vertebrates: New perspectives through a comparative approach. Brain Research Reviews, 33(2–3), 308–379. https://doi.org/10.1016/S0165-0173(00)00034-5

Smith, G. T. (1996). Seasonal plasticity in the song nuclei of wild rufous-sided towhees. Brain Research, 734(1– 2), 79–85. https://doi.org/10.1016/0006-8993(96)00613-0

163

Smulders, T. V., Sasson, A. D., & DeVoogd, T. J. (1995). Seasonal variation in hippocampal volume in a food‐ storing bird, the black‐capped chickadee. Journal of Neurobiology, 27(1), 15–25. https://doi.org/10.1002/neu.480270103

Smulders, T. V., Shiflett, M. W., Sperling, A. J., & Devoogd, T. J. (2000). Seasonal changes in neuron numbers in the hippocampal formation of a food-hoarding bird: The black-capped chickadee. Journal of Neurobiology, 44(4), 414–422. https://doi.org/10.1002/1097-4695(20000915)44:4<414::AID-NEU4>3.0.CO;2-I

Sockol, M. D., Raichlen, D. A., & Pontzer, H. (2007). Chimpanzee locomotor energetics and the origin of human bipedalism. Proceedings of the National Academy of Sciences of the United States of America, 104(30), 12265–12269. https://doi.org/10.1073/pnas.0703267104

Sokoloff, L., Reivich, M., Kennedy, C., Rosiers, M. H. D., Patlak, C. S., Pettigrew, K. D., … Shinohara, M. (1977). THe [14C]deoxyglucose method for the measurement of local cerebral glucose utilization: theroy, procedure, and normal values in the conscious and anesthetized albino rat. Journal of Neurochemistry, 28(5), 897–916. https://doi.org/10.1111/j.1471-4159.1977.tb10649.x

Sol, D. (2009). The cognitive-buffer hypothesis for the evolution of large brains. In R. Dukas & J. M. Ratcliff (Eds.), Cognitive Ecology 2 (pp. 111–136).

Sol, D., Timmermans, S., & Lefebvre, L. (2002). Behavioural flexibility and invasion success in birds. Animal Behaviour, 63(3), 495–502. https://doi.org/10.1006/anbe.2001.1953

Sporns, O., & Zwi, J. D. (2004). The small world of the cerebral cortex. Neuroinformatics, 2, 145–162. https://doi.org/10.1385/NI:2:2:145

Springer, M. S., Emerling, C. A., Meredith, R. W., Janečka, J. E., Eizirik, E., & Murphy, W. J. (2017). Waking the undead: Implications of a soft explosive model for the timing of placental mammal diversification. Molecular Phylogenetics and Evolution, 106, 86–102. https://doi.org/10.1016/j.ympev.2016.09.017

Stacho, M., Herold, C., Rook, N., Wagner, H., Axer, M., Amunts, K., & Güntürkün, O. (2020). A cortex-like canonical circuit in the avian forebrain. Science, 369(6511). https://doi.org/10.1126/science.abc5534

Starosta, S., Bartetzko, I., Stüttgen, M. C., & Güntürkün, O. (2017). Integration of contextual cues into memory depends on “prefrontal” N-methyl-D-aspartate receptors. Neurobiology of Learning and Memory, 144, 19– 26. https://doi.org/10.1016/j.nlm.2017.05.012

Starosta, S., Güntürkün, O., & Stüttgen, M. C. (2013). Stimulus-Response-Outcome Coding in the Pigeon Nidopallium Caudolaterale. PLoS ONE, 8(2), e57407. https://doi.org/10.1371/journal.pone.0057407

Starosta, S., Stüttgen, M. C., & Güntürkün, O. (2014). Recording single neurons’ action potentials from freely moving pigeons across three stages of learning. Journal of Visualized Experiments, (88), e51283. https://doi.org/10.3791/51283

Stetner, M., & Fee, M. (2017). Quantification of single unit activity related to head and body movements in the intermediate arcopallium of the zebra finch. Society for Neuroscience.

Strausfeld, N. J., & Hirth, F. (2013). Deep homology of arthropod central complex and vertebrate basal ganglia. Science, 340, 157–161. https://doi.org/10.1126/science.1231828

Striedter, G.F. (2005). Principles of brain evolution. Sinauer associates.

164

Striedter, G F, & Keefer, B. P. (2000). Cell migration and aggregation in the developing telencephalon: pulse- labeling chick embryos with bromodeoxyuridine. The Journal of Neuroscience, 20(21), 8021–8030. https://doi.org/10.1523/JNEUROSCI.20-21-08021.2000

Striedter, G. F. (2016). Evolution of the hippocampus in reptiles and birds. Journal of Comparative Neurology, 524, 496–517. https://doi.org/10.1002/cne.23803

Sweazea, K. L., & Braun, E. J. (2006). Glucose transporter expression in English sparrows (Passer domesticus). Comparative Biochemistry and Physiology - B Biochemistry and Molecular Biology, 144(3), 263–270. https://doi.org/10.1016/j.cbpb.2005.12.027

Szabo, B., & Whiting, M. J. (2020). Do lizards have enhanced inhibition? A test in two species differing in ecology and sociobiology. Behavioural Processes, 172. https://doi.org/10.1016/j.beproc.2020.104043

Tomasi, S., Caminiti, R., & Innocenti, G. M. (2012). Areal differences in diameter and length of corticofugal projections. Cerebral Cortex, 22(6), 1463–1472. https://doi.org/10.1093/cercor/bhs011

Tosches, M. A., & Laurent, G. (2019). Evolution of neuronal identity in the cerebral cortex. Current Opinion in Neurobiology, 56, 199–208. https://doi.org/10.1016/j.conb.2019.04.009

Toyama, H., Ichise, M., Liow, J.-S., Modell, K. J., Vines, D. C., Esaki, T., … Innis, R. B. (2004). Absolute quantification of regional cerebral glucose utilization in mice by 18F-FDG small animal PET scanning and 2- 14C-DG autoradiography. Journal of Nuclear Medicine, 45(8), 1398–1405.

Tramontin, A. D., & Brenowitz, E. A. (2000). Seasonal plasticity in the adult brain. Trends in Neurosciences, 23, 251–258. https://doi.org/10.1016/S0166-2236(00)01558-7

Triana, E., & Pasnak, R. (1981). Object permanence in cats and dogs. Animal Learning & Behavior, 9(1), 135– 139. https://doi.org/10.3758/BF03212035

Uddin, M., Wildman, D. E., Liu, G., Xu, W., Johnson, R. M., Hof, P. R., … Goodman, M. (2004). Sister grouping of chimpanzees and humans as revealed by genome-wide phylogenetic analysis of brain gene expression profiles. Proceedings of the National Academy of Sciences of the United States of America, 101(9), 2957– 2962. https://doi.org/10.1073/pnas.0308725100

Ulinski, P. (1983). Dorsal ventricular ridge: a treatise on forebrain organization in reptiles and birds. New York: J. Wiley.

Ulinski, P. S. (1978). Organization of anterior dorsal ventricular ridge in snakes. The Journal of Comparative Neurology, 178(3), 411–449. https://doi.org/10.1002/cne.901780303

Uylings, H. B. M., Groenewegen, H. J., & Kolb, B. (2003). Do rats have a prefrontal cortex? Behavioural Brain Research, 146(1–2), 3–17. https://doi.org/10.1016/J.BBR.2003.09.028

Vates, G. E., Broome, B. M., Mello, C. V., & Nottebohm, F. (1996). Auditory pathways of caudal telencephalon and their relation to the song system of adult male zebra finches (Taenopygia guttata). The Journal of Comparative Neurology, 366(4), 613–642. https://doi.org/10.1002/(SICI)1096- 9861(19960318)366:4<613::AID-CNE5>3.0.CO;2-7

Veenman, C. L., Wild, J. M., & Reiner, A. (1995). Organization of the avian corticostriatal projection system: A retrograde and anterograde pathway tracing study in pigeons. The Journal of Comparative Neurology, 354(1), 87–126. https://doi.org/10.1002/cne.903540108

165

Veit, L., Hartmann, K., & Nieder, A. (2014). Neuronal correlates of visual working memory in the corvid endbrain. Journal of Neuroscience, 34(23), 7778–7786. https://doi.org/10.1523/JNEUROSCI.0612-14.2014

Veit, L., & Nieder, A. (2013). Abstract rule neurons in the endbrain support intelligent behaviour in corvid songbirds. Nature Communications, 4(1), 2878. https://doi.org/10.1038/ncomms3878

Veit, L., Pidpruzhnykova, G., & Nieder, A. (2015). Associative learning rapidly establishes neuronal representations of upcoming behavioral choices in crows. Proceedings of the National Academy of Sciences of the United States of America, 112(49), 15208–15213. https://doi.org/10.1073/pnas.1509760112

Verkhratsky, A., & Kettenmann, H. (1996). Calcium signalling in glial cells. Trends in Neurosciences, 19, 346– 352. https://doi.org/10.1016/0166-2236(96)10048-5

Vetter, P., Roth, A., & Häusser, M. (2001). Propagation of action potentials in dendrites depends on dendritic morphology. Journal of Neurophysiology, 85(2), 926–937. https://doi.org/10.1152/jn.2001.85.2.926

Vincze, O. (2016). Light enough to travel or wise enough to stay? Brain size evolution and migratory behavior in birds. Evolution, 70(9), 2123–2133. https://doi.org/10.1111/evo.13012 von Eugen, K., Tabrik, S., Güntürkün, O., & Ströckens, F. (2020). A comparative analysis of the dopaminergic innervation of the executive caudal nidopallium in pigeon, chicken, zebra finch, and carrion crow. Journal of Comparative Neurology, cne.24878. https://doi.org/10.1002/cne.24878

Wagener, L., Loconsole, M., Ditz, H. M., & Nieder, A. (2018). Neurons in the Endbrain of Numerically Naive Crows Spontaneously Encode Visual Numerosity. Current Biology, 28(7), 1090-1094.e4. https://doi.org/10.1016/j.cub.2018.02.023

Wagnon, G. S., & Brown, C. R. (2020). Smaller brained cliff swallows are more likely to die during harsh weather. Biology Letters, 16(7), 20200264. https://doi.org/10.1098/rsbl.2020.0264

Waldmann, C., & Güntürkün, O. (1993). The dopaminergic innervation of the pigeon caudolateral forebrain: immunocytochemical evidence for a “prefrontal cortex” in birds? Brain Research, 600(2), 225–234. https://doi.org/10.1016/0006-8993(93)91377-5

Wallis, J. D., Anderson, K. C., & Miller, E. K. (2001). Single neurons in prefrontal cortex encode abstract rules. Nature, 411(6840), 953–956. https://doi.org/10.1038/35082081

Walsh, S. A., & Knoll, F. (2017). The Evolution of Avian Intelligence and Sensory Capabilities: The Fossil Evidence. In Digital Endocasts (pp. 59–69). https://doi.org/10.1007/978-4-431-56582-6_5

Wang, R., Chen, C.-C., Hara, E., Rivas, M. V., Roulhac, P. L., Howard, J. T., … Jarvis, E. D. (2015). Convergent differential regulation of SLIT-ROBO axon guidance genes in the brains of vocal learners. Journal of Comparative Neurology, 523(6), 892–906. https://doi.org/10.1002/cne.23719

Wang, Y., Zorio, D. A. R., & Karten, H. J. (2017). Heterogeneous organization and connectivity of the chicken auditory thalamus ( Gallus gallus ). Journal of Comparative Neurology, 525(14), 3044–3071. https://doi.org/10.1002/cne.24262

Wang, Z., Pascual-Anaya, J., Zadissa, A., Li, W., Niimura, Y., Huang, Z., … Irie, N. (2013). The draft genomes of soft-shell turtle and green sea turtle yield insights into the development and evolution of the turtle-specific body plan. Nature Genetics, 45(6), 701–706. https://doi.org/10.1038/ng.2615

166

Watanabe, A., Gignac, P. M., Balanoff, A. M., Green, T. L., Kley, N. J., & Norell, M. A. (2018). Are endocasts good proxies for brain size and shape in archosaurs throughout ontogeny? Journal of Anatomy, 234(3), 291–305. https://doi.org/10.1111/joa.12918

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. The Journal of Comparative Neurology, 236(2), 241–257. https://doi.org/10.1002/cne.902360208

Watanabe, S., Sakamoto, J., & Wakita, M. (1995). Pigeon’s discrimination of paintings by Monet and Picasso. Journal of the Experimental Analysis of Behavior, 63(2), 165–174. https://doi.org/10.1901/jeab.1995.63-165

Waxman, S. G., & Bennett, M. V. I. (1972). Relative conduction velocities of small myelinated and non-myelinated fibres in the central nervous system. Nature New Biology, 238(85), 217–219. https://doi.org/10.1038/newbio238217a0

Wedel, M. J. (2005). Postcranial skeletal pneumaticity in sauropods and its implication for mass estimates. In K. Curry Rogers & J. Wilson (Eds.), The Sauropods: Evolution and Paleobiology (pp. 201–228).

Welch, K. C., Allalou, A., Sehgal, P., Cheng, J., & Ashok, A. (2013). Glucose Transporter Expression in an Avian Nectarivore: The Ruby-Throated Hummingbird (Archilochus colubris). PLoS ONE, 8(10), e77003. https://doi.org/10.1371/journal.pone.0077003

Welch, K. C., & Altshuler, D. L. (2009). Fiber type homogeneity of the flight musculature in small birds. Comparative Biochemistry and Physiology - B Biochemistry and Molecular Biology, 152(4), 324–331. https://doi.org/10.1016/j.cbpb.2008.12.013

White, C. R., Phillips, N. F., & Seymour, R. S. (2006). The scaling and temperature dependence of vertebrate metabolism. Biology Letters, 2(1), 125–127. https://doi.org/10.1098/rsbl.2005.0378

Wikmark, R. G. E., Divac, I., & Weiss, R. (1973). Retention of Spatial Delayed Alternation in Rats with Lesions in the Frontal Lobes. Brain, Behavior and Evolution, 8(5), 329–339. https://doi.org/10.1159/000124362

Wild, J. M., Arends, J. J. A., & Zeigler, H. P. (1985). Telencephalic connections of the trigeminal system in the pigeon (Columba livia): A trigeminal sensorimotor circuit. The Journal of Comparative Neurology, 234(4), 441–464. https://doi.org/10.1002/cne.902340404

Wild, J. Martin, & Farabaugh, S. M. (1996). Organization of afferent and efferent projections of the nucleus basalis prosencephali in a passerine,Taeniopygia guttata. The Journal of Comparative Neurology, 365(2), 306–328. https://doi.org/10.1002/(SICI)1096-9861(19960205)365:2<306::AID-CNE8>3.0.CO;2-9

Wild, J.M., & Williams, M. N. (1999). Rostral wulst of passerine birds: II. Intratelencephalic projections to nuclei associated with the auditory and song systems. The Journal of Comparative Neurology, 413(4), 520–534. https://doi.org/10.1002/(SICI)1096-9861(19991101)413:4<520::AID-CNE3>3.0.CO;2-B

Willemet, R. (2013). Reconsidering the evolution of brain, cognition, and behavior in birds and mammals. Frontiers in Psychology, 4, 1–26. https://doi.org/10.3389/fpsyg.2013.00396

Wilson, B., Mackintosh, N. J., & Boakes, R. A. (1985). Transfer of relational rules in matching and oddity learning by pigeons and corvids. The Quarterly Journal of Experimental Psychology Section B, 37(4), 313–332. https://doi.org/10.1080/14640748508401173

167

Wilson, J. E. (2003). Isozymes of mammalian hexokinase: Structure, subcellular localization and metabolic function. Journal of Experimental Biology, 206, 2049–2057. https://doi.org/10.1242/jeb.00241

Wise, S. P. (2008). Forward frontal fields: phylogeny and fundamental function. Trends in Neurosciences, 31(12), 599–608. https://doi.org/10.1016/j.tins.2008.08.008

Wong, K. P., Sha, W., Zhang, X., & Huang, S. C. (2011). Effects of administration route, dietary condition, and blood glucose level on kinetics and uptake of 18F-FDG in mice. Journal of Nuclear Medicine, 52(5), 800– 807. https://doi.org/10.2967/jnumed.110.085092

Wright, A. A., & Elmore, L. C. (2016). Pigeon visual short-term memory directly compared to primates. Behavioural Processes, 123, 84–89. https://doi.org/10.1016/j.beproc.2015.09.002

Wright, A. A., Magnotti, J. F., Katz, J. S., Leonard, K., Vernouillet, A., & Kelly, D. M. (2017). Corvids Outperform Pigeons and Primates in Learning a Basic Concept. Psychological Science, 28(4), 437–444. https://doi.org/10.1177/0956797616685871

Wylie, D. R., Gutierrez-Ibanez, C., & Iwaniuk, A. N. (2015). Integrating brain, behaviour and phylogeny to understand the evolution of sensory systems in birds. Frontiers in Neuroscience, 9, 281. https://doi.org/10.3389/fnins.2015.00281

Wynne, B., & Güntürkün, O. (1995). Dopaminergic Innervation of the Telencephalon of the Pigeon (Columba- Livia) - A Study with Antibodies Against Tyrosine-Hydroxylase and Dopamine. Journal of Comparative Neurology, 357(3), 446–464. Retrieved from isi:A1995RF04000008

Yamamoto, K., & Bloch, S. (2017). Overview of Brain Evolution: Lobe-Finned Fish vs. Ray-Finned Fish. In S. Watanabe, M. A. Hofman, & T. Shimizu (Eds.), Evolution of the Brain, Cognition, and Emotion in Vertebrates (pp. 3–33). https://doi.org/10.1007/978-4-431-56559-8_1

Yamashita, S. (2007). Heat-induced antigen retrieval: Mechanisms and application to histochemistry. Progress in Histochemistry and Cytochemistry, 41(3), 141–200. https://doi.org/10.1016/j.proghi.2006.09.001

Yang, Z., Jiang, B., McNamara, M. E., Kearns, S. L., Pittman, M., Kaye, T. G., … Benton, M. J. (2019). Pterosaur integumentary structures with complex feather-like branching. Nature Ecology and Evolution, 3(1), 24–30. https://doi.org/10.1038/s41559-018-0728-7

Yonezawa, T., Segawa, T., Mori, H., Campos, P. F., Hongoh, Y., Endo, H., … Hasegawa, M. (2017). Phylogenomics and Morphology of Extinct Paleognaths Reveal the Origin and Evolution of the Ratites. Current Biology, 27(1), 68–77. https://doi.org/10.1016/j.cub.2016.10.029

Yu, Y., Herman, P., Rothman, D. L., Agarwal, D., & Hyder, F. (2017). Evaluating the gray and white matter energy budgets of human brain function. Journal of Cerebral Blood Flow & Metabolism, 38(8), 1339–1353. https://doi.org/10.1177/0271678X17708691

Yu, Y., Hill, A. P., & McCormick, D. A. (2012). Warm body temperature facilitates energy efficient cortical action potentials. PLoS Computational Biology, 8(4), 1002456. https://doi.org/10.1371/journal.pcbi.1002456

Yu, Y., Karbowski, J., Sachdev, R. N. S., & Feng, J. (2014). Effect of temperature and glia in brain size enlargement and origin of allometric body-brain size scaling in vertebrates. BMC Evolutionary Biology, 14(1), 1–14. https://doi.org/10.1186/s12862-014-0178-z

168

Zeier, H., & Karten, H. J. (1971). The archistriatum of the pigeon: Organization of afferent and efferent connections. Brain Research, 31(2), 313–326. https://doi.org/10.1016/0006-8993(71)90185-5

Zentall, T. R., Clement, T. S., Bhatt, R. S., & Allen, J. (2001). Episodic-like memory in pigeons. Psychonomic Bulletin & Review, 8(4), 685–690. https://doi.org/10.3758/BF03196204

Zentall, T. R., & Raley, O. L. (2019). Object permanence in the pigeon (Columba livia): Insertion of a delay prior to choice facilitates visible- and invisible-displacement accuracy. Journal of Comparative Psychology, 133(1), 132–139. https://doi.org/10.1037/com0000134

Zoli, M., Torri, C., Ferrari, R., Jansson, A., Zini, I., Fuxe, K., & Agnati, L. F. (1998). The emergence of the volume transmission concept. Brain Research Reviews, 26(2–3), 136–147. https://doi.org/10.1016/S0165- 0173(97)00048-9

169

170

Appendix

171

172

A Curriculum Vitae

ACADEMIC University of Washington, Seattle, USA – College of the APPOINTMENTS Environment Visiting Scholar (November 2017)

Utrecht University, the Netherlands – Department of Farm Animal Health Graduate research assistant (2014)

EDUCATION Ruhr-University Bochum & International Graduate School of Neuroscience, Germany, Ph. D (2016 – now) Dissertation: Understanding the building blocks of avian complex cognition: the executive caudal nidopallium and the neuronal energy budget

Utrecht University, the Netherlands, Environmental Biology – Behavioural Ecology, MSc. (2013 – 2016), summa cum laude

University College Utrecht, the Netherlands, Bachelor of Science, BSc. (2010 – 2013), magna cum laude, honours college

University of California Berkeley, USA (2012) Semester exchange

TEACHING Ruhr-University Bochum, Psychology / Cognitive Science 1. Comparative Cognition (BSc. Spring 2020) 2. Can pigeons learn rules? (BSc. Spring 2017) 3. The evolution of higher cognitive functions in vertebrates (MSc. Autumn 2016, Autumn 2017, Autumn 2018, Autumn 2019, Autumn 2020)

Ruhr-University Bochum, Psychology / Cognitive Science Supervisor Master Thesis (2020) Supervisor Bachelor Thesis (2019)

Utrecht University, Department of veterinary science Supervision Bachelor Thesis (2014)

FELLOWSHIPS Member U/Select Honours Programme Life Sciences, Utrecht University (2014)

RESEARCH PLUS funding, Travel & Conference Grant, €29.517,- (RUB Research FUNDING School, 2018) PLUS funding, Travel & Lab visit Grant, €3.304,- (RUB Research School, 2017)

173

INVITED ‘Glucose metabolism of the avian brain: an 18F-FDG-PET study in PRESENTATIONS pigeons’, Comparative Imaging, Ruhr-University Bochum, Germany (May 2019)

‘The executive nidopallium in the carrion crow: the NCL and beyond’, Avian Cognitive Neuroscience, Ruhr-University Bochum, Germany (April 2018)

‘Differences in location, trajectory and size of the NCL in a variety of bird species’, University of Cambridge, UK (November 2016)

CONFERENCE ‘Glucose metabolism of the avian brain: an 18F-FDG-PET study’, PRESENTATIONS Society for Neuroscience, San Diego, USA (November 2018) (SELECTION) ‘The avian prefrontal cortex revisited’, Society for Neuroscience, Washington D.C., USA (November 2017)

‘A qualitative study into differences in NCL size between different bird species’, Rational animals, Ruhr-University Bochum, Germany (September 2016)

CONFERENCE International conference Comparative Imaging, Ruhr-University ORGANIZED Bochum, Germany (May 2019)

Bi-annual international conference Avian Cognitive Neuroscience, Ruhr-University Bochum, Germany (April 2018, April 2020)

SERVICE TO THE Referee for Animals PROFESSION Referee for Agriculture Referee for the RUB research school

LANGUAGES Dutch (native), English (fluent), German (proficient), French (intermediate), Russian (basic)

174

B List of publications

Billings, B.K., Bhagwandin, A., Patzke, N., Ngwena, A., Rook, N., von Eugen, K., Tabrik, S., Güntürkün, O., & Manger, P. Nuclear organization and morphology of catecholaminergic neurons and certain pallial terminal networks in the brain of the Nile crocodile, (Crocodylus niloticus). J. Chem. Neur., 109, 101851. von Eugen, K., Tabrik, S., Güntürkün, & Ströckens, F. A comparative analysis of the dopaminergic innervation of the executive caudal nidopallium in pigeon, chicken, zebra finch, and carrion crow (2020). J. Comp. Neur. von Eugen, K., Nordquist, R.E., Zeinstra, E., & van der Staay, F.J. Stocking density affects stress and anxious behaviour in the laying hen chick during rearing (2019). Animals, 9, 53. van Ruijssevelt, L., Chen, Y., von Eugen, K., XXX Güntürkün, O., Woolley, S.C., van der Linden, A. fMRI reveals a novel region for evaluating acoustic information for mate choice in a female songbird (2018). Curr. Biol., 28, 711-721.

175

176

C Acknowledgements

Writing this part of the thesis implies that I am really close to the finish line, which currently feels a bit surreal. I now also wonder if I should have written this in the beginning, since now my brain is complete mush and I am afraid I will not find the right words or that I would even forget people. Okay, well, here goes nothing.

First, I would like to thank Onur Güntürkün for giving me the opportunity to work in the Biopsychology lab, and to get immersed and involved in the field of avian neuroscience. My enthusiasm for the field was already sparked during my master thesis, but the past four years have strengthened it to an extend that I think I will never be able to get rid of it. Also, I would like to thank Carsten Theiꞵ and the other members of my committee for taking the time to read and evaluate my dissertation. Next, a deepest and sincerest thank you is expressed to Felix Ströckens, who has been my on-the-ground supervisor during the past four years, and already during my master thesis. I think it is fair to say that without you I would not be where I am right now. Thank you for having come up with these super exciting projects that I was able to work at, for getting the money to fund it, and for sharing your ideas and knowledge. I am also very grateful for your shared enthusiasm for the things we discovered and the fruitful discussion we had around them. I hope this will no be the last project we collaborate on, and that we continue to explore how the avian brain is able to facilitate complex cognition (I am counting on you to not get snatched by the human brain people in Dusseldorf). I am also very grateful to Elisa Wiebeck and Kim Walusiacki for their assistance in the wet lab, and Sabine Kesch for taking such good care of the animals, and providing assistance with housing the animals in Cologne. I must also thank Gabi Salzmann, Juliane Brenscheidt, Aiko Gastberg and Sandra Linn for keeping the lab running and helping me navigate the maze that is German bureaucracy.

Besides my own lab, there were other research groups I would like to thank that have welcomed me and assisted my during my PhD time. First, I am very thankful for the group at the University of Cologne, and Heike Endepols and Heiko Backes in particular. Heike has been a crucial part in the PET project and taught me everything around PET scanning, and always with a infectious smile and cheerful stories. Heiko, also known as the math wizard, has been extremely patient with the joint task of figuring out the data and coming up with different approaches to model pigeon cerebral metabolism. I also really appreciated that he always remained kind and optimistic, even though it was sometimes just extremely frustrating. A big thanks goes out to all the other people that assisted the project in Cologne. I would also like to thank Jonas Rose for getting me involved in two projects that I have been (and still am) very excited about: the crow atlas and the Avian Cognitive Neuroscience conference. Being able to

177

organize and be involved in the organization of ACN gave me the opportunity to plan a conference that I would be the most excited about, and that was exactly what ACN 2018 was. I truly hope we can pick it up again, and that it will be the start of a whole community. I also have to thank Jonas for bringing Lukas & Aylin to Bochum, who became my close friends, and fellow down-toe-hookers and rosé-lovers. A special thanks goes to Lukas who brightened up many gloomy days with sunny outside lunches, really really bad jokes and fierce discussions.

This brings me back to the Biopsychology lab. I have been extremely lucky with my office- mates, who provided me with a mix of warmth, fun, science, and support. This brings me to Sara Letzner, who has such a warm heart and is at the same time a total badass in science. And Meng Gao, who always surprised me with her strong moral code and unusual humour. Last but not least I want to express my appreciation for Juan, who has sometimes driven me mad, but he has also been disarmingly honest and always up for a chat about literally any topic imaginable. I am also very thankful for the rest of the Biopsychology group, for the Christmas dinners, movie night, and retreats. It is truly a remarkable group and I am grateful to having been a part of it.

Lastly, I really want to thank my friends and family for their love, support and company, and putting up with me rambling about birds and dinosaurs. First, thank you family: Lucie for helping my out with the entire lay-out and pulling me through the last final days when Word decided to mess everything up, and thanks to Luigi for giving the final styling tips and making amazing spinach pancakes. Thank you Caspar for listening and making jokes when it was badly needed, and mam and pap for being proud of me no matter what I do, and always making me feel loved and supported. I am also extremely grateful for Feli, for too many things to list here, but in this context especially for reminding me that fear is a bad counsellor. Then, thank you Thijs for proof-reading, your feedback and your enthusiasm. And thank you Mardy and Claire for remote digital love and energizing words.

Finally, thank you birds!

178