Professional Documents
Culture Documents
Ebook Avian Cognition 1St Edition Carel Ten Cate Online PDF All Chapter
Ebook Avian Cognition 1St Edition Carel Ten Cate Online PDF All Chapter
Cate
Visit to download the full and correct content document:
https://ebookmeta.com/product/avian-cognition-1st-edition-carel-ten-cate/
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
https://ebookmeta.com/product/hell-bent-1st-edition-cate-corvin/
https://ebookmeta.com/product/hell-bent-1st-edition-cate-
corvin-2/
https://ebookmeta.com/product/avian-immunology-3rd-edition-bernd-
kaspers/
https://ebookmeta.com/product/hell-to-pay-1st-edition-cate-
corvin/
Hell Hath No Fury 1st Edition Cate Corvin
https://ebookmeta.com/product/hell-hath-no-fury-1st-edition-cate-
corvin/
https://ebookmeta.com/product/all-hell-breaks-loose-1st-edition-
cate-corvin/
https://ebookmeta.com/product/all-hell-breaks-loose-1st-edition-
cate-corvin-2/
https://ebookmeta.com/product/sturkies-avian-physiology-seventh-
edition-colin-g-scanes/
https://ebookmeta.com/product/heterogeneous-contributions-to-
numerical-cognition-learning-and-education-in-mathematical-
cognition-1st-edition/
Avian Cognition
The cognitive abilities of birds are remarkable: hummingbirds integrate spatial and tem-
poral information about food sources, day-old chicks have a sense of numbers, parrots
can make and use tools and ravens have sophisticated insights in social relationships.
This volume describes the full range of avian cognitive abilities, the mechanisms behind
such abilities and how they relate to the ecology of the species.
Synthesising the latest research in avian cognition, a range of experts in the field
provide first-hand insights into experimental procedures, outcomes and theoretical
advances, including a discussion of how the findings in birds relate to the cognitive
abilities of other species, including humans. The authors cover a range of topics such
as spatial cognition, social learning, tool use, perceptual categorization and concept
learning, providing the broader context for students and researchers interested in the
current state of avian cognition research, its key questions and appropriate experimental
approaches.
Carel ten Cate is Professor of Animal Behaviour at Universiteit Leiden, the Nether-
lands. His research focuses on behavioural development and communication by vocal
and visual signals in species ranging from birds and fish to humans. He also uses birds
for comparative studies on the cognitive mechanisms relevant for speech and language
processing in humans.
Susan D. Healy is a Reader in Zoology at the University of St Andrews, Scotland and
Executive Editor of Animal Behaviour. Her work integrates theoretical and empirical
data from the fields of biology and psychology to investigate the behavioural ecology
and neurobiology of animal cognition, with a focus on birds.
Avian Cognition
Edited by
CAREL TEN CATE
Leiden University
SUSAN D. HEALY
University of St Andrews
University Printing House, Cambridge CB2 8BS, United Kingdom
One Liberty Plaza, 20th Floor, New York, NY 10006, USA
477 Williamstown Road, Port Melbourne, VIC 3207, Australia
4843/24, 2nd Floor, Ansari Road, Daryaganj, Delhi - 110002, India
79 Anson Road, #06-04/06, Singapore 079906
www.cambridge.org
Information on this title: www.cambridge.org/9781107092389
DOI: 10.1017/9781316135976
C Cambridge University Press 2017
Index 337
Contributors
Lucy M. Aplin
Edward Grey Institute, Department of Zoology, University of Oxford, Oxford, UK
Alice Auersperg
Messerli Research Institute, University of Veterinary Medicine, Medical University of
Vienna, University of Vienna, Vienna, Austria
Ulrike Aust
Messerli Research Institute, University of Veterinary Medicine, Medical University of
Vienna, University of Vienna, Vienna, Austria
Marc T. Avey
Ottawa Hospital Research Institute, Ottawa Hospital, Ottawa, Canada
Laurie L. Bloomfield
Department of Psychology, Algoma University, Sault Ste. Marie, Canada
Neeltje J. Boogert
Edward Grey Institute, Department of Zoology, University of Oxford, Oxford, UK
Thomas Bugnyar
Department of Cognitive Biology, University of Vienna, Vienna, Austria
Leyre Castro
Department of Psychology, University of Iowa, Iowa City, Iowa, USA
Cinzia Chiandetti
Department of Life Sciences, University of Trieste, Trieste, Italy
Andrea S. Griffin
School of Psychology, University of Newcastle, Callaghan, Australia
David Guez
School of Psychology, University of Newcastle, Callaghan, Australia
viii List of Contributors
Lauren M. Guillette
School of Biology, University of St Andrews, St Andrews, UK
Allison Hahn
Department of Psychology, University of Alberta, Edmonton, Canada
Christina G. Halpin
Centre for Behaviour and Evolution, Institute of Neuroscience, Newcastle University,
Newcastle, UK
Susan D. Healy
School of Biology, University of St Andrews, St Andrews, UK
Marisa Hoeschele
Department of Cognitive Biology, University of Vienna, Vienna, Austria
Ludwig Huber
Messerli Research Institute, University of Veterinary Medicine, Medical University of
Vienna, University of Vienna, Vienna, Austria
T. Andrew Hurly
Department of Biological Sciences, University of Lethbridge, Lethbridge, Canada
Debbie M. Kelly
University of Manitoba, Winnipeg, Canada
Louis Lefebvre
Department of Biology, McGill University, Montreal, Canada
Neil McMillan
Department of Psychology, University of Alberta, Edmonton, Canada
Irene M. Pepperberg
Department of Psychology, Harvard University, Cambridge, Massachusetts, USA
James F. Reichert
University of Manitoba, Winnipeg, Canada
Candy Rowe
Centre for Behaviour and Evolution, Institute of Neuroscience, Newcastle University,
Newcastle, UK
List of Contributors ix
Sebastian Schwarz
University of Manitoba, Winnipeg, Canada
David F. Sherry
Department of Psychology, Western University, London, Ontario, Canada
John Skelhorn
Centre for Behaviour and Evolution, Institute of Neuroscience, Newcastle University,
Newcastle, UK
Christopher B. Sturdy
Department of Psychology, Neuroscience and Mental Health Institute, University of
Alberta, Edmonton, Canada
Sabine Tebbich
Department of Cognitive Biology, University of Vienna, Vienna, Austria
Irmgard Teschke
Department of Cognitive Biology, University of Vienna, Vienna, Austria
Giorgio Vallortigara
Centre for Mind-Brain Sciences, University of Trento, Rovereto, Italy
Edward E. Wasserman
Department of Psychology, University of Iowa, Iowa City, Iowa, USA
Preface
In the 1980s the emergence of behavioural ecology significantly changed the study of
animal behaviour. It shifted the focus of many researchers from studying causal mech-
anisms to addressing the adaptive significance and costs and benefits of the way in
which animals behaved. We are currently witnessing a similar shift in focus, one in
which researchers are asking and addressing questions about the cognitive processes
that underlie animal behaviour. This shift is visible in the rapidly expanding field of
animal cognition where the focus is on analysing the nature and development of the
‘knowledge’ that enables animals to respond to the challenges they face in their daily
life: going to the right place at the right moment, getting the food they need, dealing
with conspecifics, and so on. These are intriguing questions and their answers not only
tell us something about how animals ‘think’, but may also provide a window on the
origin of our own thinking. Although primates, apes in particular, have long been the
main group to which these questions have been addressed, more recently birds have
emerged as the focus of attention. Birds show a diverse range of mental abilities, which
are proving accessible to systematic research. Against this background, we organised an
‘Avian Cognition’ symposium at the International Ethological Congress in Newcastle,
UK in 2013. It attracted a lot of interest and when we organised another one at the next
meeting in Cairns, Australia in 2015, it was even more popular.
This volume arose out of the interest shown for these symposia. It became clear that
many people would like to know more about the questions and progress of the field. It
inspired us to invite colleagues working at the forefront of avian cognition to contribute
to a book on the topic. We got enthusiastic responses, as well as stimulating support
for the idea from Cambridge University Press, resulting in what has become this book.
Right from the start we wanted it to be a book that would be of interest to those working
in the field, reviewing the current state of knowledge, while at the same time being
accessible for a broader audience consisting of interested researchers from other areas
as well as suitable for (graduate) student seminars. We challenged our contributors to
review relevant work beyond their own studies, to draw comparisons both among bird
species as well as between birds and other species, including humans, and to keep their
writing understandable. We thank all our authors for their great contributions (and for
their patience with dealing with our comments). It has been a very rewarding exercise
to collect all the wonderful work presented in this book. We hope it will be similarly
rewarding to those reading it and that it will contribute to advancing the topic!
We, humans, are amazing animals. We devise and use tools to do things we would not
be able to do otherwise; we can remember how we solved a problem in the past and
use this knowledge to solve a current problem; we can plan ahead; we can distinguish
and categorise objects on abstract as well as on functional properties; we can commu-
nicate about events in the past, present or the future; we form complex social networks,
and so on. Not only do we take these abilities for granted, many people assume they
demonstrate our cognitive superiority over other animal species. However, many non-
human animals also do amazing things: New Caledonian crows can manufacture tools
to extract food from wood logs, chimpanzees tell each other what kind of predator is
in the vicinity, honey bees tell their sisters where to find food, racing pigeons return to
their home loft from hundreds of miles away and bottlenose dolphins coordinate their
hunting. These are just a very few of the examples that give rise to questions about
how these animals are able to do what they do: Do they have an ‘understanding’ of
the situation? What characterises this ‘understanding’? And how do the mechanisms
involved relate to those of humans? Such questions are core to the study of animal
cognition.
Animal Cognition
Analogous to the study of human cognition, the study of animal cognition examines
how animals perceive, process, learn, store and use information (Shettleworth, 2010).
It asks for the kind of knowledge and the mechanisms that enable animals to behave as
described above, how the relevant knowledge has been acquired and how that knowl-
edge is used to produce the behaviour observed. The focus in this field is on observing
behaviour followed by experimental manipulations to test how the behaviour is brought
about. Animals may or may not have private experiences like consciousness or feel-
ings related to their behaviour, but because these experiences are private it is difficult
to access them. Animals may behave as if they have ‘thought’ about how to solve a
problem, and their behaviour may suggest forward planning or causal reasoning, but we
can measure only their behavioural solutions to the problem and how and what kind
of previous experiences and contextual information affect that solution. Fortunately, as
demonstrated by the chapters in this book, this approach to the study of animal cognition
is very successful.
2 Carel ten Cate and Susan D. Healy
Research Questions
The wonder about the remarkable abilities that animals display in their daily lives is
what drives many researchers to study animal cognition. These animal abilities give rise
to different types of questions similar to those formulated by Tinbergen (1963) for the
study of animal behaviour: what are the underlying mechanisms (how is it done?), how
does the ability develop (what experiences affect its emergence?), how did it evolve
(what was its origin, what are the selection pressures?) and is it adaptive (i.e. do the
animals that have the cognitive ability leave more offspring?). While many researchers
find their inspiration for research questions in the natural abilities of animals, others
find their inspiration in the cognitive abilities of humans and wonder where they come
from and how they relate to those of non-human animal species: to what extent do
other species and humans solve the problem at hand by the same mechanisms? Are
there general principles that apply to humans and non-human animals alike? In the past,
the distinction between these two types of researchers was related to differences in the
research traditions that gave rise to them: behavioural ecology and the study of animal
behaviour versus comparative psychology. Both approaches have a lot to offer and it is
therefore encouraging to see that there has been a merger of these approaches over the
years. This is also reflected in the contents of this book, with contributions by people
from a wide range of backgrounds.
Why Birds?
If examples of remarkable cognitive abilities can be found among many species, ranging
from insects to mammals, why then single out birds? One trivial reason is that the field
of animal cognition is flourishing and broad. Rather than attempting to capture every-
thing, a focus on a single clade enables a more comprehensive and coherent treatment
of the main topics addressed for that clade. Birds form a very interesting group for such
a focus. With over 10,000 species, birds are a species-rich and without doubt the most
conspicuous, vertebrate clade. They are present on every continent and in all environ-
ments, whether it is the center of a busy city, a remote oceanic island or a pristine rain
forest, and their presence is usually well visible and audible. It is thus no wonder that
they have always drawn the interest of researchers of animal behaviour. Both Niko Tin-
bergen and Konrad Lorenz, two of the three men awarded the Nobel Prize for founding
the study of animal behaviour, derived many of their ideas from studying birds. Later
researchers have followed in their footsteps with the result being an extensive knowl-
edge about bird behaviour and a wealth of examples of apparently sophisticated abilities
like tool use, spatial orientation, concept formation, episodic-like memory and others.
Across the chapters in this book, we aim to present an overview of what is known about
the cognitive processes underlying such abilities.
We focus on birds also because studies on their cognitive abilities are increasingly
demonstrating that many birds appear able to match some or all of the abilities of the
primates, often considered the most ‘intelligent’ of animals (e.g. Emery & Clayton,
2004). These cognitive similarities are intriguing as there is an extensive phylogenetic
1 Introduction: Avian Cognition – Why and What? 3
gap that separates birds from mammals. The current-day insights are that the ancestor
of modern birds arose from theropod dinosaurs during the Cretaceous period, around
100 million years ago, and lived alongside the earliest mammals. Dinosaurs and mam-
mals, in turn, are assumed to have evolved from a common reptile-like stem amniote
ancestor that lived over 310 million years ago. Although it is possible that cognitive
abilities that are shared between some, but not all, mammals and birds might origi-
nate from this common ancestor, it is more likely that such similar abilities arose by
independent, but convergent evolution. This will certainly hold for similarities in more
specialised abilities. Comparing birds with other taxa thus provides the opportunity to
examine which selection pressures might have been at work to shape particular cognitive
abilities, to determine whether functionally similar behaviours in different taxa result
from similar cognitive mechanisms, and to compare their neural instantiation. Take, for
example, vocal learning. Humans acquire speech and language by being exposed to spo-
ken language. Although such vocal learning occurs in a few other mammal groups, like
cetaceans and bats, humans are the only primate species that learn their vocalisations.
Among non-mammalian vertebrates, vocal learning is known only from birds, where it
occurs in hummingbirds, parrots, songbirds and some suboscines. Investigation of vocal
learning by songbirds has revealed striking similarities with vocal learning in humans:
the learning proceeds without instruction, there is a sensitive phase for learning early
in life, the learning process is canalised with respect to the sounds that are most readily
learned, social interactions affect the model chosen for copying, perceptual learning of
the relevant sounds precedes the production, and the development is characterised by a
babbling phase in which the output is gradually shaped into the adult form by auditory
feedback (e.g. Doupe & Kuhl, 1999). Comparing such commonalities between birds and
distant taxa can provide insights into the essential or inevitable components of cognitive
abilities or on the selection pressures giving rise to them.
Cognitive similarities at the behavioural level also raise questions about the underlying
neural mechanisms. And here is another reason why birds are of interest. In their gross
anatomy, the bird and mammal brain share a general vertebrate brain structure, consist-
ing of a hindbrain, midbrain, cerebellum, thalamus and telencephalon. Of these regions,
the vertebrate telencephalon is the most variable. Mammals show a strong proliferation
of the outer areas of the telencephalon, which includes a layered neocortex. This neo-
cortex is involved in many cognitive processes and for a long time, the bird brain was
considered to be more ‘primitive’ with relatively large basal ganglia but no neocortex.
In 2004, however, based on detailed studies of nervous connectivity plus neuromolecu-
lar and developmental evidence, a large consortium of avian neuroscientists concluded
that a large part of the avian telencephalon should be considered similar in its neurobi-
ological characteristics as well as its functionality to the mammalian neocortex (Jarvis
et al., 2005). So, although superficially different, the brains of birds and mammals may
be homologous. This has been confirmed for vocal learning, as discussed above, for
which known functionally convergent neural circuits in songbirds and humans also show
4 Carel ten Cate and Susan D. Healy
convergent molecular changes and expression of multiple genes. These similarities are
both striking and intriguing as they suggest that brain circuits for complex traits may
be constrained in how they evolved from a common ancestor (Pfenning et al., 2014),
despite millions of years of evolutionary separation. It makes birds a group that is all
the more interesting for examining whether other behavioural similarities between birds
and mammals are also based on neuromolecular similarities or whether functionally
similar behaviour results from different underlying mechanisms.
This Book
The breadth of research in avian cognition is reflected in the contents of this book,
which aims to provide an overview of the current state of the field. Its emphasis is on
the behavioural rather than the neural analysis of cognitive processes.
The first chapters that follow this one concern spatial orientation and food storing. In
Chapter 2, Reichert et al. deal with the ways in which birds orient themselves in space:
what environmental features do they use and how are various types of information (the
use of landmarks and geometry) weighted against each other? While this chapter cov-
ers a lot of laboratory experiments on the issue, in Chapter 3, Healy and Hurly show
that spatial cognition can be studied in the field and address how spatial and tempo-
ral information about food sources is integrated. In Chapter 4, Sherry describes what
food-storing birds remember about caches and discusses the neural basis and processes
underlying the spatial learning and memory consolidation involved in food storing. In
Chapter 5, Rowe et al. discuss how aversive experience with insect prey, combined with
specific perceptual and learning biases of their avian predators, may affect the evolution
of warning colours in those prey, thereby demonstrating how cognitive processes may
shape evolutionary outcomes. In the two chapters that follow, the authors deal with the
relationship between innovation in the context of acquiring novel feeding behaviours
and the cognitive processes involved: in Chapter 6, Lefebvre and Aplin focus in par-
ticular on how innovations may spread within populations via social learning, while in
Chapter 7, Griffin and Guez discuss the relationship between innovation and cognition.
Vallortigara and Chiandetti, in Chapter 8, examine the extent to which several basic cog-
nitive abilities that relate to physical cognition, space and numbers are already present
in day-old chicks and whether this indicates their universal nature. Physical cognition
is also the topic of Chapter 9, in which Auersperg et al. discuss the processes underly-
ing tool use by birds. Pepperberg reviews numerical cognition in birds in Chapter 10.
Numerical cognition also deals with the question as to whether birds can form abstract
number concepts. Abstract concepts are also needed to deal with objects and their rela-
tionships, such as being the same or different. The three chapters that follow cover
the cognitive processes of concept formation and categorisation: in Chapter 11, Huber
and Aust describe whether and how birds can form perceptual categories at different
levels of abstraction and in Chapter 12, Castro and Wasserman cover relational con-
cept learning. While these chapters concentrate on the processing of visual input, in
Chapter 13 ten Cate concentrates on the processing of auditory input by birds and the
1 Introduction: Avian Cognition – Why and What? 5
extent to which birds can detect abstract regularities in auditory input. In Chapter 14,
Avey et al. also deal with auditory processing, but with emphasis on the bioacoustics
and perceptual mechanisms involved in processing natural vocalisations. Boogert in
Chapter 15 examines the relationship between song and other cognitive abilities espe-
cially with regard to mate choice. Finally, in Chapter 16, Bugnyar and Massen review
what is known about the cognitive abilities that birds display in their social relationships.
What’s Next?
The contents of this book reflect the insights obtained in many domains of avian cog-
nition as well as the ways in which studies of avian cognition contribute to insights in
cognitive processes in general. The contents also reflect a number of questions that are
still unanswered and topics that have only just begun to be explored. They show that the
field is dynamic and also that views of different researchers are sometimes divergent.
There are thus ample topics for future research. Apart from specific abilities of (some)
birds there are also general questions waiting to be resolved, such as whether and how
various cognitive abilities are related to each other; what socioecological factors drive
the evolution of cognitive abilities; and what is the relationship between variation in
personalities and cognitive abilities. Of course, many more topics lend themselves for
further study and are worthwhile pursuing. The presence of many bird species that can
be studied both in the laboratory and in the field and under a wide range of conditions
make them very well suited to address such questions. Their study will enable compar-
ison with similar studies in other animal taxa, thereby shedding light on how universal
certain abilities are and how they might have evolved. They may also provide ideas on
the origin and evolution of human cognitive abilities. We hope this book will provide a
useful and inspiring basis for such studies.
References
Doupe, A. J. and Kuhl, P. K. (1999). Birdsong and human speech: Common themes and mecha-
nisms. Annual Review of Neuroscience, 22, 567–631. DOI:10.1146/annurev.neuro.22.1.567
Emery, N. J. and Clayton, N. S. (2004). The mentality of crows: Convergent evolution of intelli-
gence in corvids and apes. Science, 306, 1903–1907.
Jarvis, E. D., Güntürkün, O., Bruce, L., et al. (2005). Avian brains and a new understanding of
vertebrate brain evolution. Nature Reviews Neuroscience, 6, 151–159. DOI:10.1038/nrn1606
Pfenning, A. R., Hara, E., Whitney, O., et al. (2014). Convergent transcriptional spe-
cializations in the brains of humans and song-learning birds. Science, 346, 1256846.
DOI:10.1126/science.1256846
Shettleworth, S. J. (2010). Cognition, Evolution and Behavior, 2nd edn. Oxford: Oxford Univer-
sity Press.
Tinbergen, N. (1963). On aims and methods of Ethology. Zeitschrift für Tierpsychologie, 20,
410–433.
2 Spatial Cognition in Birds
James F. Reichert, Sebastian Schwarz and Debbie M. Kelly
Feats of avian spatial cognition rank as some of the most impressive in the animal
world. Homing pigeons (Columbia livia) are able to travel hundreds of kilometers from
distant locations, over varying types of landscape and weather conditions, eventually
arriving at their home loft. Migratory birds are able to cover substantial distances and
travel seasonally between breeding and wintering sites. These journeys require the abil-
ity to combine spatial information from earth-based cues (olfactory cues: Ioalè et al.,
1990; Gagliardo et al., 2011; magnetic cues: Wiltschko & Wiltschko, 1978, 1996),
visual landmarks (Biro et al., 2004; Lipp et al., 2004) or sky-based cues (position
of the sun: Schmidt-Koenig, 1958; skylight polarization patterns: Kreithen & Keeton,
1974; Able, 1982). Sensitivity to such a range of cues provides birds with an impres-
sive array of navigational tools to maintain their bearings and reach their destination.
Indeed, as one of the several avian long-distance travellers, pigeons are capable of true
navigation – as witnessed by their ability to find their way home even when displaced
far off a known route (Bingman & Cheng, 2005). Food-storing birds face a similarly
important challenge of spatial cognition: these birds create food stores during times of
resource abundance for later retrieval during times of scarcity. The birds must encode
the location of these food stores in a manner that will allow them to recover the caches
within a landscape that undergoes seasonal changes between autumn and winter. Yet
despite this seemingly insurmountable demand on their spatial memory skills, long-
term food-storing birds routinely locate thousands of previously hidden food caches
with a high degree of precision, and do so year after year (Tomback, 1980; Van-
der Wall, 1982). How different avian species are able to accomplish these kinds of
tasks hinges on their ability to efficiently process the spatial relationships within their
environment.
In this chapter we examine the different cues birds use to encode spatial information
as well as the factors that influence this process. Firstly, we describe how birds use fea-
tural and geometric information for orientation, specifically how near (proximal) and
far (distal) landmarks are relied upon. Secondly, we examine how spatial information is
extracted from landmark arrays and continuous surfaces. Thirdly, we discuss the extent
to which birds use geometric and featural cues, and how those cues interact with each
other. Finally, we discuss the use of panoramic views and view-matching as an addi-
tional strategy for understanding the spatial abilities of birds.
2 Spatial Cognition in Birds 7
Before a bird can navigate it must first be able to orient itself to its surroundings, which
is the initial stage of any type of navigational endeavor. It does so by remembering
specific landmarks such as trees, rocks or flat surfaces such as those formed by moun-
tain sides. The distinctive qualities of these individual landmarks, such as color, pattern
and texture, are referred to as featural cues. In addition to featural cues, a bird may
also use geometric cues, which comprise the geometric relations between landmarks
and surfaces such as distance and directional information. By successfully encoding the
identity of individual landmarks (features) within its environment, as well as the spa-
tial location of those landmarks relative to other landmarks and surfaces (geometry),
birds have many potential sources of information at their disposal when trying to stay
oriented.
Cheng (1988, 1989) showed that pigeons could rely on the positions of both nearby
and distant landmarks to pinpoint the location of a hidden goal. Positional estimates
include both a distance component and a direction component, which together form a
vector. The Vector Sum Model (Cheng, 1989, 1994) proposes that birds can code dis-
tance and direction coordinates independently from individual landmarks to a specific
location. According to the model, the more landmarks that a bird has available, the more
accurate its estimation is likely to be. For this reason, the encoding of the locations of
multiple landmarks allows a bird more flexibility when it is trying to pinpoint a precise
location such as a food source (Kamil & Cheng, 2001).
Although birds can make use of multiple landmarks, both near and far from a goal
location, it appears that landmarks closer to a goal carry the greatest weight. Cheng
(1989) demonstrated this point by training pigeons to find food hidden between two
equal-sized landmarks, with one landmark located west and closer (10 cm) to the hid-
den food and the second landmark located east and farther away (40 cm). During test
trials each landmark was shifted from its original position an equal distance away from
the hidden goal (i.e., the closer landmark was shifted farther to the west and the more
distant landmark shifted farther to the east). The result of this landmark shift was that
the birds’ search location shifted as a consequence, with a bias toward the landmark
that had been closer to the hidden goal during training. Furthermore, Gould-Beierle
and Kamil (1999) showed that search accuracy by food-storing Clark’s nutcrackers
(Nucifraga columbiana) was much better when the birds could use a nearby landmark
as a reference as opposed to one that was farther away. Cheng (1992) provided a psy-
chophysical explanation for such findings by showing that the amount of error involved
when an animal estimates a vector from a landmark to a goal location increases propor-
tionally as the distance between the two locations increases.
Analogous to these open-field type tasks during which active locomotion is possible,
Spetch (1995) used a two-dimensional computer-based touch-screen task to show that
pigeons also relied on visual landmarks when searching for a hidden goal on the screen.
The pigeons were required to peck on the screen of a computer monitor at a particu-
lar point at a consistent vector from an array of landmarks, with the landmark array
8 James F. Reichert, Sebastian Schwarz and Debbie M. Kelly
Figure 2.1 Rendition of a training trial for a landmark array task presented to pigeons using a
computer monitor equipped with a touch-screen. Pecks directed at a precise location defined by
the landmark array results in food presented in a nearby grain hopper (modified after Spetch,
1995; photo credit: H. Hobson).
appearing at different screen locations across trials (see Figure 2.1 for a photographic
illustration). Therefore, although the absolute location of the goal relative to the screen
changed from trial to trial, the location relative to the landmark array remained con-
stant. Test trials consisted of select exposure to only certain landmarks within the array
to determine how much control individual landmarks had acquired over the pigeons’
search strategies. The pigeons not only relied more on near rather than farther land-
marks, but the learning of the nearer landmarks overshadowed learning of landmarks
that were more distant from the goal.
Pigeons are excellent navigators, able to travel hundreds of kilometers using a variety
of earth-based (i.e., magnetic and olfactory) and landmark cues, to arrive at a single,
consistent location, which is typically their home loft. For food-storing birds such as
the Clark’s nutcracker, the spatial challenge is quite different as they need to remember
and update a series of changing food locations. Pine seeds are the main food source for
Clark’s nutcrackers, which they store in individual caches during the fall when availabil-
ity is high and then retrieve during the winter when food sources are scarce (Tomback,
1978). This type of behavior requires that the birds form a memory based on a spatial
2 Spatial Cognition in Birds 9
When a prominent object, such as a tree or a rock, is positioned very close to a goal
location (e.g., a hidden food source), it can serve as a beacon for that location, in which
case a bird would simply need to fly directly to the landmark in order to reach the goal.
But when an object is far enough from a goal that a beaconing strategy is no longer
feasible, a bird must accurately estimate the metric coordinates from that object to the
goal location in order to make effective use of it as a landmark. These types of distance
and direction estimates can be accomplished by using either absolute or relative metrics.
For example, assume that a bird hides a food cache midway between two trees. If it has
encoded the food location using an absolute metric, it will attempt to remember the
location of the food as being an exact distance from either of the trees. However, if the
bird has encoded the food location using a relative metric, it will attempt to remember
the food location as being approximately midway between the two trees. At first glance
this second, relational strategy may seem the simpler of the two strategies, but it is
actually considered to be a more sophisticated and flexible form of learning (Kelly and
Spetch, 2001), specifically because it represents the formation of an abstract rule (i.e.,
the “middle rule”) that can be applied across similar situations. It should be noted that
a hallmark of human cognition is our ability to quickly adopt relational rules, and adult
humans have been shown to preferentially default to a relational strategy during spatial
tasks (e.g., Spetch et al., 1997; also see Gouteux et al., 2001 for a study with rhesus
monkeys [Macaca mulatta]).
Kamil and Jones (1997) were the first to show that birds could use a relational rule to
solve a spatial learning problem. Clark’s nutcrackers were trained to search for food hid-
den at the midway point between two colored PVC pipes, which served as landmarks.
During training, the inter-landmark distance was randomly varied in 20-cm increments
from a minimum distance of 20 cm to a maximum distance of 120 cm. During test
10 James F. Reichert, Sebastian Schwarz and Debbie M. Kelly
training testing
Figure 2.2 Schematic illustration of training and testing arrays in expansion tests. During training,
the food location (black star) was centered in the landmark array. In expansion tests, birds could
either use the “middle rule” and search for the now absent food in the relative center of the
expanded array or they could search at the absolute location indicated by one single landmark.
Grey squares represent hypothetical search peaks of tested birds (modified after Spetch et al.,
1997).
trials when the birds were presented with novel inter-landmark distances (both shorter
and longer compared to training), they continued to search at the midpoint between the
two landmarks, thus establishing that the birds had applied a relational rule to remem-
ber the location of the food during training. Jones et al. (2002) went on to show that
pigeons could similarly apply a middle rule when searching for food hidden between two
landmarks, albeit not as accurately as nutcrackers. These studies confirmed that birds
could learn the distance between two landmarks when directional information remained
unchanged (i.e., the goal was always situated between the two landmarks). Researchers
have also examined the use of directional cues by birds, as shown when Clark’s nutcrack-
ers and pigeons learned to search for food located at either a constant bearing or a
constant distance relative to two landmarks such that the relationship between the three
points formed a triangle (Kamil & Jones, 2000; Spetch et al., 2003). These studies found
that both pigeons and nutcrackers could solve these tasks using either directional or dis-
tance cues, although again, nutcrackers searched more accurately than did the pigeons.
Although birds are capable of applying relational rules when encoding landmark
arrays, laboratory experiments have established that a relational strategy is usually not
a default preference and that some species rely more on absolute vectors. For exam-
ple, pigeons were trained to search for food hidden in the center of a four-landmark
array in the overall shape of a square (Spetch et al., 1997). During expansion tests
the distance between the landmarks was doubled, resulting in an array twice the size
as compared to training. If the pigeons used a relational rule (i.e., the “middle rule”)
to encode the landmark array, then they would be expected to search in the center
of the expanded array, just as they had during training since the relative center of a
square does not change as a function of size. But instead of searching in the center, the
pigeons directed their searches to a location that maintained an absolute vector from a
single landmark that was consistent with a landmark-to-center vector experienced dur-
ing training (see Figure 2.2). A similar use of absolute geometric properties was shown
2 Spatial Cognition in Birds 11
a b
training testing
Figure 2.3 Schematic illustrations of (a) a rectangular training enclosure and (b) L-shaped testing
enclosure for studies examining the use of local geometry, medial axes and principal axes.
Dashed lines represent principal axes and solid line medial axis. Black circles in (a) can be
defined by local geometry, medial or principal axes. Black circles in (b) define corners defined
by local geometry or medial axis but not by principal axis. (modified after Kelly et al., 2011).
bisected horizontally by the minor principal axis at the center point of the space. Theo-
retically, any space – regardless of its size or shape – can be defined by its principal axes,
with perhaps the simplest examples being either a rectangular or square-shaped environ-
ment (see dashed lines in Figure 2.3). Contrary to the principal axes, the medial axes
are a more localized shape parameter that resembles a hypothetical trunk and branch
structure, with a central trunk positioned down the length of a space with branches radi-
ating to corners (solid lines in Figure 2.3). But what informational value would global
(principal axis) or local (medial axis) shape encoding serve to an animal when trying to
stay oriented?
(1) local geometry – recognition that the rewarded corner was defined by distinctive
wall length and sense properties (e.g., a long wall either to the left or right of a short
wall). Note this strategy would yield two “correct” corners, although only one was ever
baited during training (see black circles in Figure 2.3a); (2) medial axes – recognition of
the geometric shape of the rewarded corner as defined by local axes of symmetry (see
solid line and black circles in Figure 2.3a and b); and (3) principal axes – recognition
of the overall shape of the space as defined globally by the axis of symmetry (see dashed
line in Figure 2.3a and b). Of these three possibilities, chicks showed a primary reliance
on the local geometry (i.e., differential wall lengths and left-right sense), and a sec-
ondary reliance on medial axes to remember their rewarded corner, whereas pigeons
relied on medial axes, which suggests differences in how species use geometric cues to
orient. Importantly, Kelly et al. (2011) showed that neither the chicks nor pigeons used
the global shape of the rectangular training arena to locate the correct corner, but instead
they relied on more local geometric information to guide orientation.
cues may naturally hold more relevance for them over the short term compared to
species that store food over longer time periods during which local features may become
unreliable. Brodbeck (1994; Experiment 4) used a similar paradigm to investigate cue
reliance of black-capped chickadees (Poecile atricapillus), a food-storing bird. Similar
to Spetch and Edwards (1986), a single wooden feeder was baited and included within
an array of other feeders (a total of four) and both local and global geometric cues
were available to the birds. But unlike the previous study there were three types of cues
that could be used to distinguish the correct feeder: (1) its spatial position within the
four-feeder array (local geometry), (2) its spatial position within the larger experimen-
tal room (global geometry), and (3) its distinct color pattern relative to the other feeders
(featural). During testing, the different cues were placed in conflict and the birds were
able to select the cue on which they relied most. For black-capped chickadees it was the
global spatial position within the experimental room, rather than local spatial position
within the array or distinct features, which was the cue they preferred. In a follow-
up experiment (Brodbeck, 1994; Experiment 5), a non-storing bird species (dark-eyed
juncos [Junco hyemalis]) did not exhibit this same preference for global geometry, but
instead weighed each type of information equally.
Hummingbirds, as foragers, also need to make use of the local and global properties
of an environment in order to remember the location of nectar-producing flowers. To be
efficient, hummingbirds must refrain from visiting those flowers that are either empty
of nectar or close to being empty, and instead visit only those flowers that are more
likely to contain nectar, a spatial memory requirement not unlike that experienced by
food-storing birds. Hurly and Healy (1996, see also Chapter 3, Healy and Hurly) tested
whether hummingbirds (Selasphorus rufus) rely more on local featural cues associated
with flowers, such as color or color pattern, or whether they rely more on the global
spatial position of the flower relative to other naturally positioned landmarks within the
environment. The experiment was conducted in a field using free-living hummingbirds
as subjects. Four featurally distinct flowers were positioned in the shape of a square
and only one of the flowers was baited with a sucrose solution. Hummingbirds were
freely able to approach the array and discover the baited flower which contained enough
sucrose that the hummingbirds would require more than a single visit to consume it.
Between visits the experimenters drained and washed the flower of sucrose and switched
its position with that of another flower within the array. Upon its return, the humming-
bird could choose the flower that was either featurally correct (same color or color
pattern) or spatially correct (position in the array relative to other natural landmarks);
results showed that it was the spatial position of the flower that the hummingbirds relied
upon most, a finding that has been replicated since by the same authors (e.g., Healy
& Hurly, 1998). Indeed, encoding spatial position is so vital to a hummingbird’s forag-
ing strategy that it will return to a rewarded flower’s previous location even when the
flower has been removed from the site, a behavior that can be prompted by just a single
previous experience of feeding from the rewarded flower (Flores-Abreu et al., 2012).
In summary, the degree to which birds rely on local and global cues depends in
part on their general feeding ecology. For food-storing birds, which have to remem-
ber multiple locations in which food has been hidden, global spatial properties appear
to be of prime importance. This is also true of hummingbirds which similarly need to
2 Spatial Cognition in Birds 15
remember multiple locations of the most promising nectar-producing flowers. For non-
storing birds, it is the more local cues such as distinct features that seem to carry the
greatest weight. And yet, feeding ecology alone cannot always determine how a bird will
rely on different environmental cues. For example, Hodgson and Healy (2005) exam-
ined great tits (Parus major), a non-storing bird species, on their cue preference during a
simple spatial task. A tray consisting of several evenly-spaced shallow wells was placed
inside each bird’s cage; one of the wells was baited with food and covered by a colored
cloth which the bird learned to remove during training to obtain the food. During test-
ing, the colored cloth associated with the rewarded well now covered a different well,
whereas a novel colored cloth now covered the rewarded well. Given a choice between
the correct spatial location and the correct feature, the birds preferred to search at the
correct spatial location. Similarly, food-storing birds have also been shown to prioritize
featural cues over geometric cues during a spatial search task (LaDage et al., 2009), as
have hummingbirds (Hurly et al., 2014). So, although the food-gathering requirements
of a bird clearly influence its spatial decision-making, other contextual factors also play
a role, as will be addressed further in this chapter.
To test concurrent feature and geometry use in a controlled laboratory setting, Cheng
(1986) devised the rectangular reorientation paradigm. Within this paradigm, the experi-
mental apparatus consists of a rectangular walled enclosure in which all cues external to
the enclosure, both visual and auditory, are masked. Inside the enclosure often a single
corner is baited with hidden food for which an animal is free to search. Prior to entering
the enclosure the animal is first passively disoriented by being slowly rotated inside a
darkened transport box, a procedure that is used to reduce the animal’s ability to use
inertial information. Once inside the enclosure the animal is expected to search as effi-
ciently as possible by locating the baited corner without visiting any of the non-baited
corners first.
To examine the use of visual features using the rectangular reorientation paradigm, a
featural cue is located at each corner, in the form of either a distinct wall or a distinct
object; if the animal has encoded the featural information it can then use this infor-
mation to search in the baited corner without having to visit any of the other corners.
However, the animal can also use the geometric properties of the enclosure to identify
the baited corner. In the case of a rectangular enclosure, the baited corner will be located
at the juncture of one longer wall, either to the left or right of a shorter wall, depending
on which side of the enclosure that the baited corner is located. Within the rectangular
reorientation paradigm, the geometric cues are not deemed to be as informative as the
featural cues since two corners share identical geometric properties (the baited corner
and the corner diagonally opposite; see Figure 2.4 for details and an example) whereas
in the traditional procedure, only one corner contains the distinct feature associated
with the baited corner. Therefore, the best that an animal can ever do if it relies strictly
on geometric cues is to direct its initial search to either the baited corner or its diag-
onal counterpart. Despite this apparent imbalance favoring featural cues, the majority
16 James F. Reichert, Sebastian Schwarz and Debbie M. Kelly
a b c
* * *
Figure 2.4 Schematic illustrations of rectangular enclosures used for the study of geometric and
featural cues. Open circles represent containers from which the bird learns to search for hidden
food; one corner is consistently baited with food during training (for illustrative purposes, the
corner marked with “” denotes the baited corner in this example). Individual corners can be
distinguished by either (a) distinct featural cues located at the corners or (b) identical featural
cues. Birds provided only with identical features must learn the location of the baited corner
based on the geometric properties of the arena along with left-right sense (e.g., corner with long
wall to the left and short wall to the right); since the corner diagonally opposite the baited corner
contains the same geometric properties, this leads the bird to commit “rotational errors” by
choosing the baited corner and its rotational equivalent at an equal rate. During cue conflict
testing, (c) distinct featural and geometric cues are pitted against each other by relocating each
feature to the next clockwise corner which is an incorrect geometric location relative to what the
bird learned during training.
(i) Factors that Affect Feature and Geometry Use: Arena Size
Studies with newborn chicks have shown that their encoding of featural and geomet-
ric cues is sensitive to the size of the environmental arena. For example, Vallortigara
et al. (2005) trained chicks to search for food at one corner of a rectangular enclo-
sure that contained a distinct featural cue in each corner. One group of chicks was
trained inside a small enclosure and a second group was trained inside a large enclosure.
The groups were subsequently tested inside an enclosure the same size as their training
enclosure, as well as being presented with a cue conflict situation whereby the features
were relocated one corner clockwise. This latter test was conducted to examine whether
the chicks relied more on the geometric or featural properties of the enclosure when
searching for the correct corner. Chicks that had been trained and tested in the large
enclosure followed the rewarded feature to its new (geometrically incorrect) location,
2 Spatial Cognition in Birds 17
thus showing a primary reliance on features; conversely, chicks that had been trained and
tested in a small enclosure divided their choices between the correct geometric corner
and its rotational equivalent corner (neither corner contained the correct feature), thus
showing primary reliance on geometry. Taken together, these results show that geom-
etry was more salient to the birds in the small enclosure, whereas features were more
salient to birds in the large enclosure. Similar cue conflict tests using Clark’s nutcrackers
(Lambinet et al., 2014) have shown a similar general tendency for geometry to be the
more dominant cue in smaller enclosures as opposed to larger ones. Such an effect of
salience on the use of geometric information in smaller spaces has also been demon-
strated in non-avian species (fish: Sovrano et al., 2005; human children and adults:
Learmonth et al., 2001, 2002; Ratliff & Newcombe, 2008).
(ii) Factors that Affect Feature and Geometry Use: Rearing Experience
If an animal is reared within an environment that lacks informative spatial geometry, is it
able to use geometric cues effectively to orient later in life? This question was addressed
by Chiandetti and Vallortigara (2010) using newborn chicks. They reared one group of
chicks within a geometrically impoverished environment (in a circular enclosure) and
another group of chicks within a “normal” environment that contained a variety of geo-
metric properties such as right corner angles and walls of different length. When later
tested for their use of geometric cues, the circular enclosure–reared chicks did not differ
in their cue use compared to those chicks reared in the geometrically informative envi-
ronment: in small-shaped environments both groups placed higher reliance on geome-
try and in larger-sized environments they relied more on features (see also Chapter 8,
Vallortigara and Chiandetti). Chicks are the only avian species to have been tested in
this manner to date and more information is needed from other species before firm con-
clusions can be made about the effect that rearing environment has on spatial cognition.
However, other non-avian species show that rearing environments can have a significant
impact on later cue reliance. For instance, fish that have been raised in a circular envi-
ronment can learn to use both geometry and features inside a rectangular-shaped tank,
but they rely less on geometric cues than fish raised in a rectangular (i.e., geometrically
informative) environment (Brown et al., 2007). More dramatically, mice raised in a cir-
cular environment fail to learn geometry at all (but do learn features) when trained with
both features and geometry in a rectangular-shaped enclosure (Twyman et al., 2013).
(iii) Factors that Affect Feature and Geometry Use: Training Experience
Whereas rearing paradigms are designed to prevent an animal from gaining experience
with specific cues prior to experimentation, another type of manipulation involves alter-
ing the type of training that an animal receives during the experiment itself. Recall that
the original rectangular reorientation paradigm (Cheng, 1986) was designed such that
an animal could learn to associate a specific rewarded corner based on either a dis-
tinctive featural cue located at the corner or the geometric properties that formed the
corner. Although all animals tested with this paradigm will encode the featural cue,
they also encode the geometric information despite it being incidental to the solution
18 James F. Reichert, Sebastian Schwarz and Debbie M. Kelly
of the task. But what if an animal received explicit geometry training whereby it was
rewarded solely based on its geometrically correct choices? This was the question posed
by Kelly et al. (1998) using pigeons as subjects. One group of pigeons (group Feature)
was trained with distinct features at the corners whereas a second group of pigeons was
trained with identical features at the corners and only later re-trained with distinct fea-
tures (group Geometry-Feature). For those birds trained with identical features, geome-
try learning was explicitly reinforced whereas birds that were trained with distinct fea-
tures only learned geometry incidentally. Following training, all birds experienced cue
conflict testing in which the correct feature was relocated to an incorrect geometric cor-
ner. The group Feature birds chose the corner that contained the correct feature instead
of the correct geometry. However, the choices of the group Geometry-Feature birds,
which had received explicit training with geometry, were equally feature-and-geometry
directed. When birds were explicitly trained to make geometry-based choices, therefore,
their reliance on geometric information increased relative to their use of featural cues.
But what about food-storing birds, which might naturally possess a heightened pre-
paredness to use environmental geometry given its relative stability compared to fea-
tures? Reichert and Kelly (2015) tested Clark’s nutcrackers in a version of the rectangu-
lar reorientation paradigm similar to that of Kelly et al. (1998). One group of nutcrack-
ers (group Distinct) was trained in an enclosure with distinct features in the corners and
rewarded for choosing the specific corner that contained the correct feature. Two other
groups of nutcrackers received differential training: birds in group Distinct-Identical
were trained with distinct features at the corners but, after having learned the task, were
retrained with an enclosure in which the features in each corner were identical. These
birds were rewarded for their geometric choices. The birds in group Identical-Distinct
were trained with the same procedure as for group Distinct-Identical, only in reverse
order. Just as with the pigeons in the previous experiment, the explicit training to use
geometry was sufficient to shift reliance toward the use of geometry and away from
features. For the birds in group Identical-Distinct this shift was even more dramatic as
they weighed geometry significantly higher than features. These results are important
because they highlight a distinction between the use of geometric information when it
is learned incidentally and when it is learned more purposively. Experiments that have
varied training experience have been critical for demonstrating that, although birds will
learn about the geometry of their environment, it is only when geometry-based choices
are made integral to solving a task that the weight of geometrical information can rival
or even surpass that of features.
Reorientation tasks that are used to investigate the spatial processing of featural and
geometric cues may seem quite abstract considering that birds live in habitats where
they rarely encounter rectangular experimental arenas. Hence it is worth discussing
the extent to which geometric and featural cues might be involved in ecologically-
relevant navigational tasks. Using an outdoor field study, Hurly et al. (2014) investigated
whether hummingbirds rely more on geometric or featural cues in their natural habitat,
2 Spatial Cognition in Birds 19
which includes a surrounding panorama that contains a wealth of nearby and distant
landmarks. The birds were trained to approach one of four identical flowers that were
positioned in the shape of a rectangle. The rewarded flower was baited with a sucrose
solution and identified by a landmark situated nearby (a red cube), and during training
the hummingbirds learned to approach only the rewarded flower. During tests in which
the landmark was removed the hummingbirds continued to choose the rewarded flower
at a high rate and showed no evidence for using the configural geometry of the array
as a guide (i.e., they did not commit rotational errors) but instead appeared to rely at
least partly on the distal panorama. These results not only demonstrate the value of
conducting experiments on birds in their natural habitat (to compare and contrast with
laboratory research), but also show how outdoor field studies can be used to examine
the extent to which birds rely on the larger panoramic views afforded by their natural
environment.
In order to understand how view-matching could be used by birds during small or
large-scale navigational tasks it is necessary to understand how view-matching has been
studied in the past. Thus far, the vast majority of research examining view-matching
ability in animals has been conducted on insects. Insects, and especially social insects
(ants and bees), are known to rely predominantly on vision for their foraging and hom-
ing routes (Collett et al., 2006; Zeil, 2012). It was once assumed that ants and bees
navigate by using nearby terrestrial objects as landmarks and rely on the more distal
panoramic view only as a contextual reference that initiates the recall of the appropriate
landmark memory when navigating through an environment (Collett & Collett, 2002;
Collett et al., 2006; Cheng, 2012a). Using modern advanced image analysing techniques
(Zeil, 2003) that are able to obtain a better representation of what the animal actually
sees during navigation (Zollikofer et al., 1995; Schwarz et al., 2011), it appears that
view-matching using the whole panorama is likely to be the main navigational strategy
in insects (Philippides et al., 2011; Wystrach et al., 2011a; Wystrach et al., 2013).
View-matching can facilitate a compass orientation process by which the current per-
ceived retinotopic view is compared to a memorized view of a familiar environment
to set a heading direction (Zeil, 2003). Consequently, heading directions usually cor-
respond to directions with the lowest mismatch between the perceived and memorized
views. The use of the visual-compass does not require a distinction between discrete
landmarks and the rest of the panoramic view; instead, landmarks and other terres-
trial objects are naturally embedded in the current and memorized views and together
serve as a reference for heading direction. Views also include the high-contrast skyline
between the terrestrial landscape and the celestial sky. For instance, the mere contour
and shape of the skyline provide enough information for ants to home without the infor-
mation of discrete landmarks or other terrestrial objects (Graham & Cheng, 2009; Reid
et al., 2011; Lent et al., 2013). Thus, according to this approach, concepts such as geom-
etry and features do not apply to view-matching as there is no segregation between them.
View-matching has been proposed as an alternative explanation for the rotational
errors committed by birds and other animals during rectangular reorientation experi-
ments. As mentioned previously, birds that learn the geometric shape of a rectangular
arena fail to distinguish between the two geometric identical corners as they provide
20 James F. Reichert, Sebastian Schwarz and Debbie M. Kelly
the same geometric information. When ants and bees were similarly tested in rectan-
gular arenas they also displayed rotational errors as reported for vertebrates (Wystrach
& Beugnon, 2009; Wystrach et al., 2011c; Sovrano et al., 2012; Sovrano et al., 2013;
Dittmar et al., 2014). However, when the researchers examined the results from these
insect studies using photographic image analyses, they concluded that a view-matching
interpretation may also account for the data (Cheung et al., 2008; Stürzl et al., 2008).
In its simplest form the photographic image analysis is based on a pixel-by-pixel com-
parison between an image of the goal location and the current image of the location of
the animal. As the animal approaches the goal location, the two images will become
more similar, and finally result in a match when the animal is at the goal position. Thus,
this approach uses images as a tool to represent an animal’s viewpoint as it navigates
(Wystrach & Graham, 2012a). Importantly, these images are constructed to take into
consideration the animal’s visual capacities as well as to include visual cues such as
contours and edges of the arena walls. An animal that uses view-matching will move
toward the goal location (in the case of a rectangular arena this would be one of the
corners) by heading to the direction with the lowest mismatch between the current and
remembered view. Note that this is only an alternative explanation as to why animals
might make rotational errors and does not necessarily exclude the encoding of geomet-
ric properties such as the length of the walls or the surface layout of the arena.
Although it is clear that view-matching is not the omnipotent answer for all naviga-
tional abilities in birds given that birds can detect a variety of sensory cues (Cheng et al.,
2013), view-matching does represent a simple and parsimonious strategy for successful
navigation through familiar terrain (Cheng, 2012b; Wystrach & Graham, 2012b). Since
view-matching combines the whole panoramic view for orientation it is less prone to
errors caused by changes or removal of individual landmarks. Furthermore, relying on
a panoramic view instead of individual landmarks would be beneficial for food-storing
birds that need to recover food after a significant period of time has elapsed between
storage and retrieval, a period in which the local properties of landmarks might change
considerably. Indeed, research with ants has shown that their ability to accurately navi-
gate is robust even under conditions of extreme alterations to the panoramic view. Thus,
panoramic views alone may provide a high degree of spatial information for successful
homing (e.g., Wystrach et al., 2011b, Schwarz et al., 2014).
Previous results mentioned earlier in this chapter involving geometric encoding of
landmark arrays and walled arenas have also been studied using a view-matching
approach. During an elegant series of studies conducted by Pecchia and Vallortigara
(2010, 2011, 2012) the use of a view-matching strategy was specifically tested using
both pigeons and chicks. Each bird was individually trained to locate food hidden inside
one of four identical cylinders that together formed a rectangular-shaped array. Each
cylinder contained four openings, with only one opening granting access to the food
reward during training. To examine the use of view-matching during training, birds
were trained with either a variable or a stable viewpoint when accessing the food; dur-
ing variable training, birds accessed the food from variable viewpoints whereas for birds
trained with a stable viewpoint, the viewpoint at which they accessed the food remained
consistent (see Figure 2.5). If birds relied on an allocentric (i.e., object-to-object)
strategy to encode the geometry of the array they should commit rotational errors
training
a stable viewpoint b variable viewpoint
trial 1
trial 2
trial 3
trial 4
testing
equally distributed
rotational error
search
Figure 2.5 Schematic illustration of rectangular array of cylinders used for the importance of
stable viewpoints during spatial cue encoding. Black stars represent hidden food locations and
correct geometric corners. Small arrows indicate accessible food openings; the other three
openings were blocked by a transparent film. In (a) the access to food at the correct cylinder was
stable across trials (i.e., positions of the small arrows). In (b) access to food at the correct
cylinder was variable across trials (i.e., positions of the small arrows). In tests when all openings
were accessible, birds chose the correct and its geometrically identical corner more often than
the remaining two corners (dashed arrows) after stable training (a). Birds chose equally between
all four corners (dashed arrows) when tested after variable training (b) indicating that a stable
viewpoint is important for spatial encoding of an environment (modified from Pecchia &
Vallortigara, 2011).
22 James F. Reichert, Sebastian Schwarz and Debbie M. Kelly
(indicative of geometric encoding) regardless of the type of training they had received.
However, if the birds relied on a viewpoint strategy to encode the geometry of the array
they could only do so when provided with a stable viewpoint during training. Indeed,
only those birds that were trained with a stable viewpoint had learned the geometry
of the array as evidenced by their pattern of rotational errors. Therefore, reorientation
within a rectangular arena may be accomplished by using a view-matching strategy. Yet
it should be noted that further experiments are required to establish view-matching as
mechanism for reorientation in birds. Future studies need to investigate the interaction
of geometric cues, features and landmarks, as well as panoramic views by setting these
sources of information into conflict – preferably across different behavioral contexts
(e.g., caching, route following, etc.).
With technological advances in GPS tracking, small GPS loggers can document the
flight paths of homing pigeons allowing for the investigation into the types of naviga-
tional cues the birds are using when homing. By doing so researchers have demonstrated
that pigeons use visual landmarks and large visual properties of their environment to
navigate (Biro et al., 2004; Lipp et al., 2004). Such landmarks can include manmade
constructions such as highways and buildings. Theories about how the birds use the
visual information span from “mosaic like map” or “map-and-compass” strategies to
piloting. The “map-and-compass” model consists of certain waypoints along a familiar
route that provide a bird with associated and memorized compass directions toward the
home loft. These compass directions are based on all visible and non-visible characteris-
tics of the waypoint and result in certain vector bearings. Piloting, however, can be inde-
pendent of compass information as it involves only the use of visual landmarks along a
known route. Alternatively, pigeons could use a view-matching strategy by comparing
their current view with a remembered view and head in the direction of least mismatch.
It should be noted that route fidelity is also common in insects (Collett, 1996; Kohler &
Wehner, 2005; Wystrach et al., 2011b; Mangan & Webb, 2012) and several studies with
ants demonstrate that a simple visual compass without the computation of site-specific
vectors at certain waypoints along the route can account for most of the observed hom-
ing behaviors (Zeil, 2003; Wystrach et al., 2012; Wystrach et al., 2013). Piloting is
based on one landmark that functions as a visual attractor whereas in view-matching
the whole panoramic view is used to obtain the direction with the least mismatch. How-
ever, currently it is quite challenging to use the photographic image analysis technique
to investigate view-matching in flying birds as thus far it has been specifically tailored
for the study of insects.
Spatial cognition in birds is influenced by a variety of factors that range from the avail-
ability and reliability of environmental cues, to ecological and evolutionary pressures,
2 Spatial Cognition in Birds 23
and finally to an individual bird’s previous experience. Since different bird species are
confronted with a variety of spatially-based survival demands, the type of cues that
they rely on most are likely to be reflective of those demands. For example, the spa-
tial challenges that homing pigeons and migratory birds are faced with arise from the
need to navigate under changing weather patterns and through unfamiliar environments
and landscapes. For other birds, it is foraging for food that poses the critical spatial
challenge. Non-storing birds need to locate a food source and immediately consume it
whereas food-storing birds not only need to locate and collect food items, but also to
relocate them to a variety of storage sites with the intention of returning to retrieve the
items at a later date, a time that may be months in the future. It is these additional for-
aging steps that may differentially influence how storing and non-storing birds encode
and rely upon spatial information within their respective environments.
In this chapter, we focused on a central aspect of avian spatial cognition, specifically
how birds use featural and geometric information to stay oriented to their surround-
ings. We have examined how birds use nearby and distant landmarks, how geometric
information is extracted from landmark arrays and surfaces, how features and geometry
interact depending on arena size and prior experience, as well as how view-matching
might provide birds with yet another strategy for encoding their spatial environment.
Research into avian spatial cognition has provided us with a wealth of understanding to
which we can only provide a limited overview. However, we have endeavored to raise
many central issues regarding how featural and geometric cues can be extracted from
the environment to build an accurate and reliable spatial representation.
Throughout this chapter we have attempted to provide a comparative approach to
avian spatial cognition, but with a bias to raising the idea that ecological and evolution-
ary pressures may have shaped how different species weigh the cues within their spatial
environment. A prime example is whether the spatial considerations involved in food-
storing might encourage a stronger reliance on relative geometric encoding in which
a general rule can be extracted from a few experiences and applied to new situations.
Research in this area is still too preliminary to draw strong conclusions, and indeed
the majority of the studies we have reviewed have been conducted on the pigeon and
the Clark’s nutcracker as representatives of non-storing and storing birds, respectively.
Although this approach is clearly restricted in terms of its ability to draw firm conclu-
sions, especially considering the number of closely related species and/or the number of
species that share similar ecological pressures. However, given that an increasing num-
ber of avian species are being used to address comparative questions of spatial cogni-
tion, we are both hopeful and confident that these ecologically- and evolutionary-driven
questions will be addressed in future research.
It should be noted that all of the research that has been addressed in this chapter thus
far has involved spatial cognition at the behavioral level, but of parallel importance is
what occurs at the neural level when an animal actively navigates its surroundings. How
animals extract spatial coordinates from their environment and how this information is
represented in the brain has received considerable attention over the years. One recent
and noteworthy sign of the importance of this field of research is the joint awarding
of the 2014 Nobel Prize in Physiology or Medicine to John O’Keefe, and May-Britt
Moser and Edvard Moser for their research on the functional properties of grid cells.
24 James F. Reichert, Sebastian Schwarz and Debbie M. Kelly
Located in the mammalian entorhinal cortex, grid cells provide the framework for a
metric coordinate system by which an animal can track its movements within a space
(Moser et al., 2008). Grid cells in turn send their information to the nearby hippocampus
where place cells are located and code for the construction of a more sophisticated
neural map of an environment that serves as the foundation of spatial memory. Together
with head direction cells which code for directional positioning of the animal (Knierim
et al., 1995) and boundary cells which code for environmental boundaries such as walls
or surfaces (Lever et al., 2009), a complete neural circuit dedicated to maintaining an
animal’s spatial awareness becomes evident, most of it located within the hippocampal
formation.
Although knowledge of the neural components of spatial cognition has been derived
chiefly from research with mammals (primarily rats), the mammalian and the avian
hippocampus are considered to be homologous structures and understanding how the
spatial map is represented in the avian brain is a worthy endeavor as birds are now
being recognized as important models of age-related spatial decline (Wilzeck & Kelly,
2013; Coppola et al., 2014; Coppola et al., 2015), since birds rely heavily on the visual
properties of the their environment to navigate, just as humans do. Understanding the
pressures of avian food storing has also lead to a clearer understanding of neurogenesis
and neural plasticity in general, most strikingly in the way that the hippocampus under-
goes seasonal fluctuations in size that correlate with seasonal spatial memory demands
(see Sherry & Hoshooley, 2010).
As a whole, the study of spatial cognition in birds allows for an unprecedented
breadth of knowledge given the range of different species involved and the varia-
tion in spatial demands required among those species. Homing pigeons and migratory
birds often travel hundreds of kilometers to reach a destination, and with advance-
ments in GPS data recording the cues that these birds use along their routes can be
inferred with ever greater precision. Comparisons between how food-storing and non-
storing birds use environmental cues provide an ideal framework from which to exam-
ine the ecological and evolutionary demands that help shape both short-term and long-
term spatial strategies. Although view-matching has thus far been studied primarily
in insects, the degree to which it contributes to both small-scale and large-scale nav-
igational strategies in birds offers a promising new area of study. By studying avian
behavior in the laboratory as well as in their natural environment, and correlating the
results with neuronal measures, the field of avian spatial cognition offers the promise
of a thorough understanding of how the brain stores, retrieves and processes spatial
information.
References
Able, K. (1982). Skylight polarization patterns at dusk influence migratory orientation in birds.
Nature, 299, 550–551.
Bingman, V. P. and Cheng, K. (2005). Mechanisms of animal global navigation: comparative
perspectives and enduring challenges. Ethology Ecology and Evolution, 17, 295–318.
2 Spatial Cognition in Birds 25
Biro, D., Meade, J. and Guilford, T. (2004). Familiar route loyalty implies visual pilotage in the
homing pigeon. Proceedings of the National Academy of Sciences of the United States of Amer-
ica, 101, 17440–17443.
Brodbeck, D. R. (1994). Memory for spatial and local cues: a comparison of a storing and non-
storing species. Animal Learning & Behavior, 22, 119–133.
Brown, A. A., Spetch, M. L., Hurd, P. L. (2007). Growing in circles: rearing environment alters
spatial navigation in fish. Psychological Science, 18, 569–573.
Cheng, K. (1986). A purely geometric module in the rat’s spatial representation. Cognition, 23,
149–178.
Cheng, K. (1988). Some psychophysics of the pigeon’s use of landmarks. Journal of Comparative
Physiology A, 162, 815–826.
Cheng, K. (1989). The vector sum model of pigeon landmark use. Journal of Experimental Psy-
chology: Animal Behavior Processes, 15, 366–375.
Cheng, K. (1992). Three psychophysical principles in the processing of spatial and temporal
information. In Cognitive aspects of stimulus control, eds. W. K. Honig and J. G. Fetterman,
Hillsdale, NJ: Erlbaum, pp. 69–88.
Cheng, K. (1994). The determination of direction in landmark-based spatial search in pigeons: a
further test of the vector sum model. Animal Learning & Behavior, 22, 291–301.
Cheng, K. (2012a). Arthropod navigation: ants, bees, crabs, spiders finding their way. In The
Oxford Handbook of Comparative Cognition, eds. E. A. Wasserman and T. R. Zentall, Oxford,
NY: Oxford University Press, pp. 189–209.
Cheng, K. (2012b). How to navigate without maps: the power of taxon-like navigation in ants.
Comparative Cognition & Behavior Reviews, 7, 1–22.
Cheng, K., Huttenlocher, J. and Newcombe, N. S. (2013). 25 years of research on the use of
geometry in spatial reorientation: a current theoretical perspective. Psychonomic Bulletin &
Review, 20, 1033–1054.
Cheng, K. and Newcombe, N. S. (2005). Is there a geometric module for spatial orientation?
Squaring theory and evidence. Psychonomic Bulletin & Review, 12, 1–23.
Cheung, A., Stürzl, W., Zeil, J. and Cheng, K. (2008). The information content of panoramic
images II: view-based navigation in nonrectangular experimental arenas. Journal of Experi-
mental Psychology: Animal Behavior Processes, 34, 15–30.
Chiandetti, C. and Vallortigara, G. (2010). Experience and geometry: controlled-rearing studies
with chicks. Animal Cognition, 13, 463–470.
Collett, T. S. (1996). Insect navigation en route to the goal: multiple strategies for the use of
landmarks. The Journal of Experimental Biology, 199, 227–235.
Collett, T. S. and Collett, M. (2002). Memory use in insect visual navigation. Nature Reviews
Neuroscience, 3, 542–552.
Collett, T. S., Graham, P., Harris, R. A. and Hempel de Ibarra, N. (2006). Navigational memories
in ants and bees: memory retrieval when selecting and following routes. Advances in the Study
of Behaviour, 36, 123–172.
Coppola, V. J., Flaim, M. E., Carney, S. N. and Bingman, V. P. (2015). An age-related deficit
in spatial-feature reference memory in homing pigeons (Columba livia). Behavioural Brain
Research, 280, 1–5.
Coppola, V. J., Hough, G. and Bingman, V. P. (2014). Age-related spatial working memory deficits
in homing pigeons (Columba livia). Behavioral Neuroscience, 128, 666–675.
Dittmar, L., Stürzl, W., Jetzschke, S., Mertes, M. and Boeddeker, N. (2014). Out of the box: how
bees orient in an ambiguous environment. Animal Behaviour, 89, 13–21.
26 James F. Reichert, Sebastian Schwarz and Debbie M. Kelly
Flores-Abreu, I. N., Hurly, T. A. and Healy, S. D. (2012). One-trial spatial learning: wild hum-
mingbirds relocate a reward after a single visit. Animal Cognition, 15, 631–637.
Gagliardo, A., Ioalè, P., Filannino, C. and Wikelski, M. (2011). Homing pigeons only navigate
in air with intact environmental odours: a test of the olfactory activation hypothesis with GPS
data loggers. PLoS ONE, 6, e22385.
Goodyear, A. J. and Kamil, A. C. (2004). Clark’s nutcrackers (Nucifraga columbiana) and the
effects of goal-landmark distance on overshadowing. Journal of Comparative Psychology, 118,
258–264.
Gould-Beierle, K. and Kamil, A. (1999). The effect of proximity on landmark use in Clark’s
nutcrackers. Animal Behaviour, 58, 477–488.
Gouteux, S., Thinus-Blanc, C. and Vauclair, J. (2001). Rhesus monkeys use geometric and nonge-
ometric information during a reorientation task. Journal of Experimental Psychology: General,
130, 505–519.
Graham, P. and Cheng, K. (2009). Ants use the panoramic skyline as a visual cue during naviga-
tion. Current Biology, 19, 935–937.
Gray, E. R., Spetch, M. L., Kelly, D. M. and Nguyen, A. (2004). Searching in the center: Pigeons
encode relative distances from walls of an enclosure. Journal of Comparative Psychology, 118,
113–117.
Healy, S. D. and Hurly, T. A. (1998). Rufous hummingbirds’ (Selasphorus rufus) memory for
flowers: patterns or actual spatial locations? Journal of Experimental Psychology: Animal
Behavior Processes, 24, 396–404.
Hodgson, Z. G. and Healy, S. D. (2005). Preference for spatial cues in a non-storing songbird
species. Animal Cognition, 8, 211–214.
Hurly, T. A., Fox, T. A. O, Zwueste, D. M. and Healy S. D. (2014). Wild hummingbirds rely on
landmarks not geometry when learning an array of flowers. Animal Cognition, 17, 1157–1165.
Hurly, T. A. and Healy S. D. (1996). Memory for flowers in rufous hummingbirds: location or
local visual cues? Animal Behaviour, 51, 1149–1157.
Ioalè, P., Nozzolini, M. and Papi, F. (1990). Homing pigeons do extract directional information
from olfactory stimuli. Behavioral Ecology and Sociobiology, 26, 301–305.
Jones, J. E., Antomiadis, E., Shettleworth, S. J. and Kamil, A. C. (2002). A comparative study of
geometric rule learning by nutcrackers (Nucifraga columbiana), pigeons (Columba livia), and
jackdaws (Corvus monedula). Journal of Comparative Psychology, 116, 350–356.
Kamil, A. C. and Cheng, K. (2001). Way-finding and landmarks: the multiple bearings hypothesis.
The Journal of Experimental Biology, 204, 103–113.
Kamil, A. C. and Jones, J. E. (1997). The seed-storing corvid Clark’s nutcracker learns geometric
relationships among landmarks. Nature, 390, 276–279.
Kamil, A. C. and Jones, J. E. (2000). Geometric rule learning by Clark’s nutcrackers (Nucifraga
columbiana). Journal of Experimental Psychology: Animal Behavior Processes, 26, 439–453.
Kelly, D. M., Chiandetti, C. and Vallortigara, G. (2011). Re-orienting in space: do animals use
global or local geometry strategies? Biology Letters, 7, 472–375.
Kelly, D. M., Kippenbrock, S., Templeton, J. and Kamil, A. C. (2008). Use of a geometric
rule or absolute vectors: Landmark use by Clark’s nutcrackers (Nucifraga columbiana). Brain
Research Bulletin, 76, 293–299.
Kelly, D. M. and Spetch, M. L. (2001). Pigeons encode relative geometry. Journal of Experimen-
tal Psychology: Animal Behavior Processes, 27, 417–422.
Kelly, D. M., Spetch, M. L. and Heth, C. D. (1998). Pigeons’ (Columba livia) encoding of geo-
metric and featural properties of a spatial environment. Journal of Comparative Psychology,
112, 259–269.
2 Spatial Cognition in Birds 27
Knierim, J. J., Kudrimoti, H. S. and McNaughton, B. L. (1995). Place cells, head direction cells,
and the learning of landmark stability. The Journal of Neuroscience, 15, 1648–1659.
Kohler, M. and Wehner, R. (2005). Idiosyncratic route-based memories in desert ants, Melopho-
rus bagoti: how do they interact with path-integration vectors? Neurobiology of Learning and
Memory, 83, 1–12.
Kreithen, M. and Keeton, W. (1974). Detection of polarized light by the homing pigeon, Columba
livia. Journal of Comparative Physiology, 89, 83–92.
LaDage, L. D., Roth II, T. C., Fox, R. A. and Pravosudov, V. V. (2009). Flexible cue use in food-
caching birds. Animal Cognition, 12, 419–426.
Lambinet, V., Wilzeck, C. and Kelly, D. M. (2014). Size does not matter, but features do: Clark’s
nutcrackers (Nucifraga columbiana) weigh features more heavily than geometry in large and
small enclosures. Behavioural Processes, 102, 3–11.
Learmonth, A. E., Nadel, L. and Newcombe, N. S. (2002). Children’s use of landmarks: implica-
tions for modularity theory. Psychological Science, 13, 337–341.
Learmonth, A. E., Newcombe, N. S. and Huttenlocher, J. (2001). Toddlers’ use of met-
ric information and landmarks to reorient. Journal of Experimental Child Psychology, 80,
225–244.
Lent, D., Graham, P. and Collett, T. S. (2013). Visual scene perception in navigating wood ants.
Current Biology, 23, 684–690.
Lever, C., Burton, S., Jeewajee, A., O’Keefe, J. and Burgess, N. (2009). Boundary vector cells in
the subiculum of the hippocampal formation. The Journal of Neuroscience, 29, 9771–9777.
Lipp, H. P., Vyssotski, A. L., Wolfer, D. P., Renaudineau, S., Savini, M., Tröster, G. and Dell’Omo,
G. (2004). Pigeon homing along highways and exits. Current Biology, 14, 1239–1249.
Mangan, M. and Webb, B. (2012). Spontaneous formation of multiple routes in individual desert
ants (Cataglyphis velox). Behavioral Ecology, 23, 944–954.
Moser, E. I., Kropff, E. and Moser, M. B. (2008). Place cells, grid cells, and the brain’s spatial
representation system. Annual Review of Neuroscience, 31, 69–89.
Pecchia, T. and Vallortigara, G. (2010). View-based strategy for reorientation by geometry. The
Journal of Experimental Biology, 213, 2987–2996.
Pecchia, T. and Vallortigara, G. (2011). Stable panoramic views facilitate snap-shot like memories
for spatial reorientation in homing pigeons. PLoS One, 6, 22657.
Pecchia, T. and Vallortigara, G. (2012). Spatial reorientation by geometry with freestanding
objects and extended surfaces: a unifying view. Proceedings of the Royal Society B: Biological
Sciences, 279, 2228–2236.
Philippides, A., Baddeley, B., Cheng, K. and Graham, P. (2011). How might ants use panoramic
views for route navigation? The Journal of Experimental Biology, 214, 445–451.
Ratliff, K. R. and Newcombe, N. S. (2008). Reorienting when cues conflict. Psychological Sci-
ence, 19, 1301–1307.
Reichert, J. F. and Kelly, D. M. (2015). How Clark’s nutcrackers (Nucifraga columbiana) weigh
geometric cues depends on their previous experience. Animal Cognition, 18, 953–968.
Reid, S. F., Narendra, A., Hemmi, J. M. and Zeil, J. (2011). Polarised skylight and the landmark
panorama provide night-active bull ants with compass information during route following. The
Journal of Experimental Biology, 214, 363–370.
Schmidt-Koenig, K. (1958). Experimentelle Einflußnahme auf die 24-Stunden-Periodik bei Brief-
tauben und deren Auswirkungen unter besonderer Berücksichtigung des Heimfindevermögens.
Zeitschrift für Tierpsychologie, 15, 301–331.
Schwarz, S., Julle-Daniere, E., Morin, L., et al. (2014). Desert ants (Melophorus bagoti) navi-
gating with robustness to distortions of the natural panorama. Insectes Sociaux, 61, 371–383.
28 James F. Reichert, Sebastian Schwarz and Debbie M. Kelly
Schwarz, S., Narendra, A. and Zeil, J. (2011). The properties of the visual system in the Australian
desert ant Melophorus bagoti. Arthropod Structure and Development, 40, 128–134.
Sherry, D. F. and Hoshooley, J. S. (2010). Seasonal hippocampal plasticity in food-storing birds.
Proceedings of the Royal Society B: Biological Sciences, 365, 933–943.
Sovrano, V. A., Bisazza, A. and Vallortigara, G. (2005). Animals’ use of landmarks and metric
information to reorient: effects of the size of the experimental space. Cognition, 97, 121–133.
Sovrano, V. A., Potrich, D. and Vallortigara, G. (2013). Learning of geometry and features in
bumblebees (Bombus terrestris). Journal of Comparative Psychology, 127, 312–318.
Sovrano, V. A., Rigosi, E. and Vallortigara, G. (2012). Spatial reorientation by geometry in bum-
blebees. PLoS One, 7, e37449.
Spetch, M. L. (1995). Overshadowing in landmark learning: Touch-screen studies with pigeons
and humans. Journal of Experimental Psychology: Animal Behavior Processes, 21, 166–181.
Spetch, M. L., Cheng, K., MacDonald, S. E., et al. (1997). Use of landmark configuration in
pigeons and humans: II. Generality across search tasks. Journal of Comparative Psychology,
111, 14–24.
Spetch, M. L. and Edwards, C. A., (1986). Spatial memory in pigeons (Columba livia) in an
open-field feeding environment. Journal of Comparative Psychology, 100, 266–278.
Spetch, M. L., Rust, T. B., Kamil, A. C. and Jones, J. E. (2003). Search by rules: pigeons’
(Columba livia) landmark-based search according to constant bearing or constant distance.
Journal of Comparative Psychology, 117, 123–132.
Stürzl, W., Cheung, A., Cheng, K. and Zeil, J. (2008). The information content of panoramic
images I: the rotational errors and the similarity of views in rectangular experimental arenas.
Journal of Experimental Pschology: Animal Behavior Processes, 34, 1–14.
Tomback, D. F. (1978). Foraging strategies of Clark’s nutcracker. Living Bird, 16, 123–161.
Tomback, D. F. (1980). How nutcrackers find their seed stores. Condor, 82, 10–19.
Tommasi, L. and Vallortigara, G. (2000). Searching for the center: spatial cognition in the domes-
tic chick (Gallus gallus). Journal of Experimental Psychology: Animal Behavior Processes, 26,
477–486.
Twyman, A. D., Newcombe, N. S. and Gould, T. J. (2013). Malleability in the development of
spatial representation. Developmental Psychobiology, 55, 243–255.
Vallortigara G., Feruglio M. and Sovrano V. A. (2005). Reorientation by geometric and landmark
information in environments of different spatial size. Developmental Science, 8, 393–401.
Vander Wall, S. (1982). An experimental analysis of cache recovery in Clark’s nutcrackers. Animal
Behaviour, 30, 84–94.
Wiltschko, R. and Wiltschko, W. (1978). Evidence for the use of magnetic outward-journey infor-
mation in homing pigeons. Nature, 65, 112–113.
Wiltschko, W. and Wiltschko, R. (1996). Magnetic orientation in birds. The Journal of Experi-
mental Biology, 199, 29–38.
Wilzeck, C. and Kelly, D. M. (2013). Avian visual pseudoneglect: the effect of age and sex on
visuospatial side biases. In Behavioral lateralization in vertebrates, eds. D. Csermely and L.
Reolin. New York: Springer, pp. 55–70.
Wystrach, A. and Beugnon, G. (2009). Ants learn geometry and features. Current Biology, 19,
61–66.
Wystrach, A., Beugnon, G. and Cheng, K. (2011a). Landmarks or panoramas: what do navigating
ants attend to for guidance? Frontiers in Zoology, 8, 21.
Wystrach, A., Beugnon, G. and Cheng, K. (2012). Ants might use different view-matching strate-
gies on and off the route. The Journal of Experimental Biology, 215, 44–55.
2 Spatial Cognition in Birds 29
Wystrach, A. and Graham, P. (2012a). View-based matching can be more than image matching:
the importance of considering an animal’s perspective. i-Perception, 3, 547–549.
Wystrach, A. and Graham, P. (2012b). What can we learn from studies of insect navigation?
Animal Behaviour, 84, 13–20.
Wystrach, A., Mangan, M., Philippides, A. and Graham, P. (2013). Snapshots in ants? New inter-
pretations of paradigmatic experiments. The Journal of Experimental Biology, 216, 1766–1770.
Wystrach, A., Schwarz, S., Schultheiss, P., Beugnon, G. and Cheng, K. (2011b). Views, land-
marks, and routes: how do desert ants negotiate an obstacle course? Journal of Comparative
Physiology A, 197, 167–179.
Wystrach, A., Sosa, S., Beugnon, G. and Cheng, K. (2011c). Geometry, features, and panoramic
views: ants in rectangular arenas. Journal of Experimental Pschology: Animal Behavior Pro-
cesses, 37, 420–435.
Zeil, J. (2003). Catchment areas of panoramic snapshots in outdoor scenes. Journal of the Optical
Society of America A, 20, 450–469.
Zeil, J. (2012). Visual homing: an insect perspective. Current Opinion in Neurobiology, 22, 285–
293.
Zollikofer, C. P. E., Wehner, R. and Fukushi, T. (1995). Optical scaling in conspecific Cataglyphis
ants. The Journal of Experimental Biology, 198, 1637–1646.
3 Spatial Cognition and Ecology:
Hummingbirds as a Case Study
Susan D. Healy and T. Andrew Hurly
Classically investigations of avian cognition have been addressed in the laboratory using
species that work well in those housing and training conditions. It is also typical that
the questions that have been addressed have concerned capability, such as whether
birds can count (Chapter 10, Pepperberg), whether they can discriminate one item from
another (Chapter 14, McMillan et al.), whether they can categorise objects (Chapter 12,
Castro and Wasserman) and so on. The rich tradition of variation in the choice of species
tested, for the most part, was based on the assumption that differences between species
would be quantitative rather than qualitative. Until, that is, the discovery of long-delay
taste aversion learning in rats (Kalat & Rozin, 1973). The discovery that not all infor-
mation was learned as readily as any other and that immediate temporal contiguity was
not necessary for learning an association set the scene for the later suggestion that there
might be adaptive specialisations in cognitive abilities (Shettleworth, 1990). An adaptive
specialisation is an ability that has been shaped over evolutionary time by an animal’s
ecology and may be a modification of how that animal perceives its environment, learns
and/or remembers pertinent information. The modification might be qualitative or quan-
titative although, in practice, it may be difficult to determine which best describes the
modification (Sherry & Schacter, 1987; Biegler et al., 2001).
The iconic example of a possible adaptive specialisation is the relationship between food
storing, spatial memory and hippocampal function seen in a range of bird and mammal
species (e.g. Krebs et al., 1989; Jacobs et al., 1990; Biegler et al., 2001; Pravosudov &
Clayton, 2002). Food-storing birds (and mammals) hide many pieces of food, often in
a large number of sites, and use memory to retrieve that food. As that body of work
is described in much more detail in chapters 2 (Reichert et al.) and 4 (Sherry), here it
will suffice to say that the investigation began with wild birds retrieving hidden food in
the wild (Cowie et al., 1981; Sherry et al., 1981) although the logistics of determining
where birds hid food and when they did (or did not) retrieve their stores soon defeated
the technology of the day. Given the sheer number of items stored (100s–1,000s), range
of locations (throughout a woodland territory), the possible distances between locations
of stores (several metres at least) and the time before retrieval (hours–months), even
now it would be far from easy to investigate the memory required for successful food
3 Hummingbird Spatial Cognition 31
(a) (b)
Figure 3.1 (a) A photograph of a male rufous hummingbird feeding from an experimental
‘flower’. (b) A photograph of a typical experiment showing the proximity to the experimental
set-up at which a human can observe the bird’s behaviour.
storing in the field. As for most other systems in which cognitive abilities have been
investigated, then, work on the cognitive abilities that underpin food storing has been
almost entirely conducted in the laboratory (e.g. Krebs et al., 1990; Healy & Krebs,
1992; Feeney et al., 2011; Barrett & Sherry, 2012).
With the (considerable) exception of work done on the role that cognition plays in
avian navigation, which is largely on homing (almost all work using homing pigeons
Columba livia: e.g. Nardi et al., 2012; Coppola et al., 2014; Flack et al., 2014; Mann
et al., 2014; Mora et al., 2014) and rather little on long-distance migration (e.g. Healy
et al., 1996; Mettke-Hofmann & Gwinner, 2003), there has been little investigation of
the role that cognition plays in the everyday lives of birds in the field. A rare exception
is the work that has addressed the nature of the information that free-living, foraging
hummingbirds attend to, learn and remember.
Several hummingbird species (and especially the males of those species) have proved
useful for investigating cognitive abilities in the field. This is in no small part because
of their logistic advantages. For example, these birds will learn how to feed from any
manner of artificial feeders containing sucrose within an hour or two, they feed about
every ten minutes and they are not readily disturbed by the presence of humans. Indeed,
they will forage within a metre or so of a seated human (Figure 3.1). Furthermore, as
territorial hummingbirds like the rufous hummingbird Selasphorus rufus will largely
exclude other animals from the experimental apparatus it becomes easier to ensure that
only the relevant individuals are trained and tested.
As a result of such characteristics, these birds have provided some of the most com-
pelling data on the spatial learning and memory capabilities, in the context of foraging,
of ‘real’ birds (i.e. birds in the wild). Here, we review that work and in so doing show
that field tests of avian cognition do not need very complex designs or equipment and
that they can yield useful data on the cognitive abilities of wild animals that are trained
and tested while they go about their everyday business.
32 Susan D. Healy and T. Andrew Hurly
Because the ways in which hummingbirds forage have been observed in some detail,
one can formulate adaptive specialisation hypotheses concerning their cognitive abili-
ties relatively easily. An early example of this comes from an experiment in which the
researchers investigated how the birds learned to forage on resources that varied in their
depletion rate. The rationale for this experiment was that foragers such as hummingbirds
and other nectarivores that often feed on resource types that are depleted in a single visit
would benefit from being able to remember which flowers they have recently emptied
in order to avoid them. Such a strategy is known as a win-shift strategy. Those species
that feed on resources that are not depleted in a single visit, however, might benefit
from using the opposite strategy: return to a recently visited but not depleted resource,
a strategy termed win-stay.
In the experiment, wild-caught hummingbirds were presented with a single feeder
containing reward, which was removed once they had fed from it (Cole et al., 1982).
Within 10–12 seconds the birds were offered a choice of two feeders, one containing
reward and the other empty. In the stay version of the task the birds were rewarded if
they returned to the feeder in the location at which they had just fed, while in the shift
version, the birds had to go to the feeder in the alternative location. In the second phase
of each trial, the birds were allowed to visit only one feeder. Because all birds were
trained to both tasks, it was possible to determine, even with a small sample of birds
(n = 8), that all the birds took much longer to learn the win-stay task than they did to
learn the win-shift task and the slowest bird to learn to shift still did so sooner than the
fastest bird learned to stay. This effect came about in two ways: (1) birds had an initial
preference to shift, and (2) improvement in performance across trials was greater in the
shift trials.
Although this difference in task acquisition is consistent with the expectation of
an adaptation in learning strategy, it is not actually quite so easy to demonstrate an
adaptive specialisation. This is because the birds’ speedier learning of the win-shift
strategy might also have been due to their familiarity with foraging using a win-
shift strategy. As the hummingbirds in this experiment were wild-caught it is likely
that these birds had experienced win-shift as an appropriate foraging strategy. Other
experiments that have demonstrated the readiness of nectarivores to learn the appro-
priate foraging strategy also cannot differentiate between experience and adaptation
as the preferred explanation. For example, rainbow lorikeets Trichoglossus hemato-
dus, which are facultative rather than obligate nectarivores, learned to win-stay just
as readily as they learned to win-shift (Sulikowski & Burke, 2011), and noisy min-
ers Manorina melanocephala, which are omnivorous honeyeaters, learned to win-shift
faster when the reward was nectar but not when rewarded with invertebrates. Just
as with the earlier Cole et al. experiment, these data are consistent with an adap-
tive specialisation interpretation but because both the lorikeets and the noisy miners
were wild-caught, they could have learned prior to testing which was the appropriate
strategy.
3 Hummingbird Spatial Cognition 33
Just as there may be advantages for foragers such as hummingbirds to learn some asso-
ciations such as win-shift more readily than others such as win-stay, it also seems likely
that some information may be more important to them than others. For example, flower
colour is a key component of the relationship between pollinators and the plants they
visit (Shrestha et al., 2013) and variation in colour across flower taxa is often corre-
lated with the identity of typical pollinators. For example, flowers that are pollinated
by the Hymenoptera tend to be blue or white, colours to which hymenopteran visual
systems are more sensitive, while hummingbird-pollinated flowers are very often red.
This correlation in flower colouration with pollinator vision leads to a common predic-
tion about hummingbird cognition, which is that they should prefer flowers that are red
(Pickens, 1930; Grant, 1966). But, in fact, red does not necessarily offer a significant
contrast against the background (Chittka & Menzel, 1992; Siitari et al., 1999; Shrestha
et al., 2013) and all of the investigations into colour preferences in hummingbirds have
tended to show that hummingbirds choose rewarded feeders based on feeder location,
rather than on feeder colour. The frequent anecdotal reports of birds preferentially visit-
ing red flowers may, in fact, be due to the hummingbirds having learned that red objects
typically contain reward: in addition to the large number of hummingbird-pollinated
flowers that are red, all commercial hummingbird feeders are red to some extent, and
hummingbirds use such feeders avidly.
Nonetheless it seems plausible that any bird that does use colour to generalise from
a known food resource to a novel resource of a similar colour would have an advan-
tage over conspecifics that do not use colour in this way. Not only would such a gen-
eraliser reduce the amount of information it had to learn and remember, it might also
save time in unfamiliar environments (e.g. on migration) when making foraging deci-
sions (Dukas & Waser, 1994; Zentall et al., 2008). One explanation for this apparent
mismatch between our expectations and the relative importance of different kinds of
information for other animals (the hummingbirds in this example) is that sometimes the
information that seems to us to be the most obviously useful to a foraging animal is not
be perceived that way by the animal. Rather, it may be more helpful to attempt to ‘see
the world through the eyes/ears/noses/or other relevant sensory system’ of that animal.
One other, now well-known, example of this was the ‘discovery’ that birds can see into
the ultraviolet part of the visual spectrum (Cuthill et al., 2000). The recognition that
female birds can see attributes of a male’s plumage that are invisible to the researchers
investigating variation in mate choice helped to explain why in some cases females
seemed to have chosen unsuitable males.
In the case of hummingbirds, however, the fact that they see flowers in a different way
to humans does not explain their apparent disregard for this ostensibly very conspicu-
ous cue. Firstly, hummingbirds can use colour when making foraging decisions. For
example, tests in which rufous hummingbirds had learned which experimental ‘flow-
ers’ (each bearing a different colour pattern) contained reward, by both moving the
flowers’ locations and disrupting their relative positions, produced evidence that the
Another random document with
no related content on Scribd:
Hobrok of hawks, | and Garm of hounds.
[103]
[105]
[106]
King Geirröth sat and had his sword on his knee, half
drawn from its sheath. But when he heard that Othin
was come thither, then he rose up and sought to take
Othin from the fire. The sword slipped from his hand,
and fell with the hilt down. The king stumbled and fell
forward, and the sword pierced him through, and
slew him. Then Othin vanished, but Agnar long ruled
there as king. [84]
[Contents]
NOTES
Prose. The texts of the two manuscripts differ in many minor details.
Hrauthung: this mythical king is not mentioned elsewhere. Geirröth:
the manuscripts spell his name in various ways. [86]Frigg: Othin’s
wife. She and Othin nearly always disagreed in some such way as
the one outlined in this story. Hlithskjolf (“Gate-Shelf”): Othin’s watch-
tower in heaven, whence he can overlook all the nine worlds; cf.
Skirnismol, introductory prose. Grimnir: “the Hooded One.” [87]
2. In the original lines 2 and 4 are both too long for the meter, and
thus the true form of the stanza is doubtful. For line 4 both
manuscripts have “the land of the Goths” instead of simply “the
Goths.” The word “Goths” apparently was applied indiscriminately to
any South-Germanic people, including the Burgundians as well as
the actual Goths, and thus here has no specific application; cf.
Gripisspo, 35 and note. [88]
5. Ydalir (“Yew-Dales”): the home of Ull, the archer among the gods,
a son of Thor’s wife, Sif, by another marriage. The wood of the yew-
tree was used for bows in the North just as it was long afterwards in
England. Alfheim: the home of the elves. Freyr: cf. Skirnismol,
introductory prose and note. Tooth-gift: the custom of making a
present to a child when it cuts its first tooth is, according to
Vigfusson, still in vogue in Iceland.
17. Vithi: this land is not mentioned elsewhere. Vithar avenged his
father, Othin, by slaying the wolf Fenrir. [92]
18. Stanzas 18–20 appear also in Snorri’s Edda. Very possibly they
are an interpolation here. Eldhrimnir (“Sooty with Fire”): the great
kettle in Valhall, wherein the gods’ cook, Andhrimnir (“The Sooty-
Faced”) daily cooks the flesh of the boar Sæhrimnir (“The
Blackened”). His flesh suffices for all the heroes there gathered, and
each evening he becomes whole again, to be cooked the next
morning.
19. Freki (“The Greedy”) and Geri (“The Ravenous”): the two wolves
who sit by Othin’s side at the feast, and to whom he gives all the
food set before him, since wine is food and drink alike for him.
Heerfather: Othin.
20. Mithgarth (“The Middle Home”): the earth. Hugin (“Thought”) and
Munin (“Memory”): the two ravens who sit on Othin’s shoulders, and
fly forth daily to bring him news of the world. [93]
21. Thund (“The Swollen” or “The Roaring”): the river surrounding
Valhall. Thjothvitnir’s fish: presumably the sun, which was caught by
the wolf Skoll (cf. Voluspo, 40), Thjothvitnir meaning “the mighty
wolf.” Such a phrase, characteristic of all Skaldic poetry, is rather
rare in the Edda. The last two lines refer to the attack on Valhall by
the people of Hel; cf. Voluspo, 51.
23. This and the following stanza stand in reversed order in Regius.
Snorri quotes stanza 23 as a proof of the vast size of Valhall. The
last two lines refer to the final battle with Fenrir and the other
enemies.
25. The first line in the original is, as indicated in the translation, too
long, and various attempts to amend it have been made. Heithrun:
the she-goat who lives on the twigs of the tree Lærath (presumably
the ash Yggdrasil), and daily gives mead which, like the boar’s flesh,
suffices for all the heroes in Valhall. In Snorri’s Edda Gangleri
foolishly asks whether the heroes drink water, whereto Har replies,
“Do you imagine that Othin invites kings and earls and other noble
men, and then gives them water to drink?”
31. The first of these roots is the one referred to in stanza 26; the
second in stanza 29 (cf. notes). Of the third root there is nothing
noteworthy recorded. After this stanza it is more than possible that
one has been lost, paraphrased in the prose of Snorri’s Edda thus:
“An eagle sits in the branches of the ash-tree, and he is very wise;
and between his eyes sits the hawk who is called Vethrfolnir.”
33. Stanzas 33–34 may well be interpolated, and are certainly in bad
shape in the Mss. Bugge points out that they are probably of later
origin than those surrounding them. Snorri [98]closely paraphrases
stanza 33, but without elaboration, and nothing further is known of
the four harts. It may be guessed, however, that they are a late
multiplication of the single hart mentioned in stanza 26, just as the
list of dragons in stanza 34 seems to have been expanded out of
Nithhogg, the only authentic dragon under the root of the ash.
Highest twigs: a guess; the Mss. words are baffling. Something has
apparently been lost from lines 3–4, but there is no clue as to its
nature.
36. Snorri quotes this list of the Valkyries, concerning whom cf.
Voluspo, 31 and note, where a different list of names is given. Hrist:
“Shaker.” Mist: “Mist.” Skeggjold: “Ax-Time.” Skogul: “Raging” (?).
Hild: “Warrior.” Thruth: “Might.” Hlokk: “Shrieking.” Herfjotur: “Host-
Fetter.” Gol: “Screaming.” Geironul: “Spear-Bearer.” Randgrith:
“Shield-Bearer.” Rathgrith: Gering guesses “Plan-Destroyer.”
Reginleif: “Gods’-Kin.” Manuscripts and editions vary greatly in the
spelling of these names, and hence in their significance.
38. Svalin (“The Cooling”): the only other reference to this shield is in
Sigrdrifumol, 15.
39. Skoll and Hati: the wolves that devour respectively the sun and
moon. The latter is the son of Hrothvitnir (“The Mighty Wolf,” i.e.
Fenrir); cf. Voluspo, 40, and Vafthruthnismol, 46–47, in which Fenrir
appears as the thief. Ironwood: a conjectural emendation of an
obscure phrase; cf. Voluspo, 40.
40. This and the following stanza are quoted by Snorri. They seem to
have come from a different source from the others of this poem;
Edzardi suggests an older version of the Vafthruthnismol. This
stanza is closely parallel to Vafthruthnismol, 21, which see, as also
Voluspo, 3. Snorri, following this account, has a few details to add.
The stones were made out of Ymir’s teeth and such of his bones as
were broken. Mithgarth was a mountain-wall made out of Ymir’s
eyebrows, and set around the earth because of the enmity of the
giants. [101]
42. With this stanza Othin gets back to his immediate situation,
bound as he is between two fires. He calls down a blessing on the
man who will reach into the fire and pull aside the great kettle which,
in Icelandic houses, hung directly under the smoke-vent in the roof,
and thus kept any one above from looking down into the interior. On
Ull, the archer-god, cf. stanza 5 and note. He is specified here
apparently for no better reason than that his name fits the initial-
rhyme.
43. This and the following stanza are certainly interpolated, for they
have nothing to do with the context, and stanza 45 continues the
dramatic conclusion of the poem begun in stanza 42. This stanza is
quoted by Snorri. Ivaldi (“The Mighty”): he is known only as the
father of the craftsmen-dwarfs who made not only the ship
Skithblathnir, but also Othin’s spear Gungnir, and the golden hair for
Thor’s wife, Sif, after Loki had maliciously cut her own hair off.
Skithblathnir: this ship (“Wooden-Bladed”) always had a fair wind,
whenever the sail was set; it could be folded up at will and put in the
pocket. Freyr: concerning him and his father, see Voluspo, 21, note,
and Skirnismol, introductory prose and note. [102]
45. With this stanza the narrative current of the poem is resumed.
Ægir: the sea-god; cf. Lokasenna, introductory prose. [103]
51. Again the poem returns to the direct action, Othin addressing the
terrified Geirröth. The manuscripts show no lacuna. Some editors
supply a second line from paper manuscripts: “Greatly by me art
beguiled.”
53. Ygg: Othin (“The Terrible”). The maids: the three Norns.
54. Possibly out of place, and probably more or less corrupt. Thund:
“The Thunderer.” Vak: “The Wakeful.” Skilfing: “The Shaker.” Vofuth:
“The Wanderer.” Hroptatyr: “Crier of the Gods.” Gaut: “Father.” Ofnir
and Svafnir: cf. stanza 34. [107]
[Contents]
SKIRNISMOL
The Ballad of Skirnir
[Contents]
Introductory Note
The Skirnismol is found complete in the Codex Regius, and through
stanza 27 in the Arnamagnæan Codex. Snorri quotes the concluding
stanza. In Regius the poem is entitled “For Scirnis” (“Skirnir’s
Journey”).
The Skirnismol differs sharply from the poems preceding it, in that it
has a distinctly ballad quality. As a matter of fact, however, its verse
is altogether dialogue, the narrative being supplied in the prose
“links,” concerning which cf. introductory note to the Grimnismol. The
dramatic effectiveness and vivid characterization of the poem seem
to connect it with the Thrymskvitha, and the two may possibly have
been put into their present form by the same man. Bugge’s guess
that the Skirnismol was the work of the author of the Lokasenna is
also possible, though it has less to support it.
[Contents]
Skirnir spake:
[109]
Skirnir spake:
Freyr spake:
Skirnir spake:
Freyr spake:
[110]
7. “To me more dear | than in days of old
Was ever maiden to man;
But no one of gods | or elves will grant
That we both together should be.”
Skirnir spake:
Freyr spake:
[111]
Skirnir spake:
Gerth spake: