Association for the Study of Animal Behaviour Winter Meeting, 2010

blaredsnottyAI and Robotics

Nov 15, 2013 (3 years and 6 months ago)



Association for the Study of
Animal Behaviour

Winter Meeting, 2010



Organised by Andy Radford & Innes Cuthill
(University of Bristol, UK)


Invited Talks

Signalling in cleaning mutualism

Redouan Bsh

University of Neuchâtel
, Switzerland)

Signalling in interspecific mutualism raises similar questions as intraspecific signalling. I will use
data on marine cleaning mutualism to illustrate several key issues. First, I will talk about signal
design, a
ddressing among others the question whether cleaner species have converged worldwide
towards rather uniform ‘guild’ colour patterns. In the second part I will address the issue of
honesty in signalling. I will present data on the question what signals may

allow cleaner wrasses
to enter the mouth of predators without being eaten. In addition, I will explore the validity of
reported cases of aggressive mimicry by sabre
tooth blennies that mimic the cleaner wrasse
Labroides dimidiatus.

Social learning acros
s species boundaries

Lars Chittka

(University of London, UK); given by Elli Leadbeater (University of
Sussex, UK)

The study of social learning has almost entirely focussed on information transfer between
members of the same species. Information about wate
r and food availability, food toxicity,
predation threats, etc. will often be of relevance for more than one species, and animals would do
well to use public information from members of other species. Indeed Darwin suggested that
pollinating insects might
learn food handling techniques by 'imitation' and 'understanding' of
heterospecifics' actions. There is now mounting evidence that such information transfer often
occurs, although sometimes, as in the case of avian alarm calls, probably by simple associati
learning. It turns out that honeybees can learn to interpret the species
specific dialect of the dance
language from interactions with other species. In some cases, habitat choice might in part be
guided by the type of information that can be gleaned fr
om attending to information provided by

Animal coloration as a form of communication

Leena Lindström

University of Jyväskylä
, Finland)

Aposematism is a defensive strategy found among many genera where conspicuous prey species

to predators that there is a cost in making an attack. These warningly coloured prey are
somehow unprofitable to the predator and a conspicuous pattern guarantees their survival, since
predators learn to avoid bold patterns more quickly than cryptic patte
rns. Therefore it has been
used as a classic example of a predator . prey interactions where the predator.s capacities (senses
and psychology) shape the prey species. However, although aposematism works well in the
laboratory it is relatively rare form of
defence. This might be explained by constraints those
conspicuously signalling prey species have. For instance conspicuous signalling is only possible if
there is unprofitable character to signal about. Thus although aposematic species have adapted to
toxins in their defence this strategy might not be a cost free, which can affect the signalling
intensity. Therefore this line of warning coloration research, which studies those other selection
pressures on aposematic prey species might also shed light to

its relative rarity as a defence. In
this talk I will try to evaluate the relative importance of the multiple selection pressures
aposematic prey experience and discuss how this research relates back to communication.


Vertebrate alarm calls and interspec
ific eavesdropping


(Australian National University, Australia)

Many birds and mammals produce alarm calls that potentially provide a rich source of
information for other prey species, even if the calls have evolved to communicate with
ecifics. I consider the extent of interspecific eavesdropping, the information gained, and the
issue of how individuals are able to recognize the alarm calls of other species. Recent playback
experiments reveal that many species do eavesdrop other species’

alarm calls, including those of
quite different taxa, such as birds eavesdropping on mammals and even lizards on birds.
Furthermore, at least some species respond appropriately to different alarm calls given by a single
species, such as those signalling d
ifferent predators or levels of danger. A remaining puzzle is
how individuals recognize other species’ alarm calls. Alarm calls of different species might share
acoustic properties, or individuals might learn to recognize individual alarm calls. Our work o
Australian birds suggests that both acoustic structure and learning are important, and that learning
can tailor behaviour to information available on a remarkably small geographic scale.
Heterospecifics therefore provide valuable information about danger
, and learning is likely to help
individuals harvest this information despite diversity and change.



and Poster Contributions

Competition and phylogeny determine community structure in Müllerian co


Markos Alexandrou

& Martin


(Bangor University, UK)

Until recently, the study of negative and antagonistic interactions (e.g. competition and predation) has
dominated our understanding of community structure, maintenance and assembly. Nevertheless, a recent
theoretical model sugges
ts that positive interactions (e.g. mutualisms) may counterbalance competition,
facilitating long
term coexistence even among ecolo
gically undifferentiated species
. Müllerian mimics are
mutualists that share the costs of predator education and are therefor
e ideally suited for the investigation of
positive and negative interactions in community dynamics. The sole empirical test of this model in a
Müllerian mimetic community supports the prediction that positive interactions outweigh the negative
effects of s
patial overlap (without quantifying resource acquisition). Understanding the role of trophic
niche partitioning in facilitating the evolution and stability of Müllerian mimetic communities is now of
critical importance, but has yet to be formally investiga
ted. Here we show that resource partitioning and
phylogeny determine community structure and outweigh the positive effects of Müllerian mimicry in a
species rich group of neotropical catfishes. From multiple independent reproductively isolated allopatric
ommunities displaying convergently evolved colour patterns, 92% consist of species that do not compete
for resources. Significant differences in phylogenetically conserved traits (snout morphology and body
size) were consistently linked to trait specific r
esource acquisition. Thus, we report the first evidence that
competition for trophic resources and phylogeny are pivotal factors in the stable evolution of Müllerian
mimicry rings. More generally, our work demonstrates that competition for resources is lik
ely to play a
dominant role in the structuring of communities that are simultaneously subject to the effects of both
positive and negative interactions.

Preliminary analysis of potential predator recognition and distinction of imminent threat by
backed uacari, Cacajao ouakary (Primates), in Brazilian Amazonia (poster)

Adrian Barnett

(Roehampton University, UK)

In Jaú National Park, Amazonian Brazil, golden
backed uacaris rarely associate with other primates, nor
are they regularly followed by b
irds. They are, however, known to be predated by large eagles, and trees
with a suite of predation
risk diminishing features appear to be preferred sleeping sites. Analysis of time
budgets indicate uacaris live in an environment where meeting daily energy
needs is a challenge, and so
avoiding incorrect responses to presence of real (but not currently threatening) predators, and also to non
predator species, is of prime importance. Here we report on the mechanisms by which golden
uacaris distinguish p
otential avian predators from non
raptoral species, the differences in response (vocal
and postural) made to known avian predators at different distances from the group, and responses to alarm
calls of other primate species. These mechanisms appear to be q
uite precise and indicate that

both recognise species likely to predate and can estimate their proximity and intent. The few
observed allocation ‘errors’ indicate
C. ouakary

uses the categories, ‘large’, dark’, ‘broad
winged’ to
delineate p
otential avian predators. Golden
backed uacaris ignore individual raptors further than 15m
away, and can distinguish between a large distant raptor and a closer one of a smaller species. Mammals
are almost universally ignored by diurnally
C. ouakary
, though trees selected as sleeping sites
possess a highly specific suite of characters that suggests these trees are chosen to minimize the possibility
of predation by nocturnal mammals. Direct observation of predation on
C. ouakary

suggests such events
re highly disruptive to the survivors’ daily time budget. Foraging in an environment where available food
is highly dispersed, and time to find it is highly restricted, the golden
backed uacari, may have developed
an acute system of distinguishing when pre
dation and from whom predation is likely to occur. They are, in
short, simply too busy to risk crying wolf.


The role of receiver biases in the evolution of plant signal form (poster)

Michael Bentley

& Tim Benton (University of Leeds, UK)

The way in

which pollinators process and learn about plant floral signals sometimes leads to ‘receiver
biases’ which predispose pollinators to respond preferentially to some signals over others. Our
understanding of how these biases have affected the evolution of pl
ant signal form is relatively poor.
Biases resulting from pollinator sensory system physiology may allow the evolution of plant signals which
‘exploit’ pollinator preferences. Biases resulting from learning could lead to selection for diversification of
ant signals or possibly to the evolution of exaggerated signals, depending on the prevailing ecological
conditions. In this research, these hypotheses are tested theoretically by using an individual
modelling approach with novel integration of neural

networks to simulate the way in which pollinators
respond to signals.

The impact on organisms of transmitting secondary replicator systems (talk)

Joanna Bryson

& Sam Brown (University of Bath, UK)

Since Dawkins' (1976) first proposal of memetics, t
here has been considerable discussion of the theoretical
viability and consequences of two replicator systems running in parallel exploiting the same “carriers”.
However, microbial biologists have discovered a fully
documentable system of concurrent evolut
between host organisms and sub
organismic elements that transmit behaviour between hosts. Bacteria
share genetic behavioural instructions with unrelated individuals including those of different species.
These instruction mechanisms (phages and plasmid
s) are themselves subject to selection both
independently and as part of their hosts, and their hosts are subject to selection both with and without
“infection”. Contrary to some popular press accounts of memetics, these bacterial models indicate that
se communicated instructions are not necessarily virulent, but can transmit advantage between hosts
that compensate for their metabolic costs, e.g. defense against environmental poisons such as mercury.
They also appear to effectively increase inclusive fi
tness (Hamilton 1964) between host organisms, leading
to increases in altruistic behaviour between hosts (Rankin, Rocha & Brown 2010). In this talk we will
describe the nature of the interactions between the hosts and their instructive transmissions. We a
rgue this
opens the book on whether signals themselves are subject to evolution, and may co
evolve with their
signalling organisms.

Do alarm rattle calls in the common blackbird
Turdus merula
convey information about
individual identity of the calle
r? (poster)

Charline Couchoux

and Torben Dabelsteen (University of Copenhagen, Denmark)

Bird song’s ability to convey information about individual identity is highly adaptive in contexts of mate
choice and territory defence. Without this ability simultane
ously communicated information about the
quality of the singer would be meaningless. Many territorial song bird species have different calls which
are used both in territorial and alarm contexts. Obviously, it would be adaptive if such calls were also able

to communicate the identity of the caller. We investigated this in the common blackbird, a territorial
species which emits so
called alarm rattle calls in both territorial and alarm contexts. We used playback of
song inside territories to elicit and recor
d alarm rattles of male owners. An analysis of the alarm rattles
demonstrated a high individual variation both in basic frequency and temporal parameters and in higher
level structural. An alarm rattle typically starts with one type of element, continues w
ith another type
(internal elements) and ends with a third type. Especially three parameters defining the structure of the call
seem important for encoding information about individual identity: duration of the alarm rattle and number
of internal and endin
g elements. The functional implications of being able to communicate individual
identity also with bird calls will be discussed.


Why lyrebirds mimic (talk)

Anastasia Dalziell
, Robert Magrath and Alex Maisey (Australian National University, Australia)

ian vocal mimicry is a conspicuous but puzzling phenomenon and it is unclear whether mimicry
performs the same signal functions as non
mimetic vocalisations. One hypothesis suggests that mimics use
mimicry to manipulate the behaviour of other species. We i
nvestigated the function of mimicry in the male
superb lyrebird; a bird renowned for both the versatility and the accuracy of its mimicry. We found that
males compartmentalized their large repertoire of mimicked sounds into four distinct groups each
guished by both the types of mimicked calls and the social or ecological contexts in which lyrebirds
sang it. We examined in detail one set of mimicry associated with male’s display dance (“dance mimicry”).
We show that in contrast to other types of mimicr
y sung by lyrebirds, dance mimicry a) is an imitation of a
mixed species mobbing flock and b) attracts other passerines. Furthermore, dance mimicry is produced
when males are particularly vulnerable to predation. We argue that dance mimicry functions to ma
the behaviour of other species of birds to reduce the predation risk to the singer. We use these results to
show a pathway for the evolution of mimicry as a form of sexual display in the superb lyrebird.

The influence of conspecific and heterosp
ecific cues on bumblebee foraging behaviour (poster)

Erika Dawson

& Lars Chittka (Imperial College London, UK)

Using information provided by conspecifics is a widespread phenomenon found across many species.
However, this information transfer is not const
rained to within species boundaries, and in fact, where
resources or predators are shared, heterospecific information could prove to be just as valuable. The
learning mechanisms behind such information use can allow animals to be flexible in the cues that
used, which raises the question of whether conspecific cues are inherently more influential than cues
provided by heterospecifics, or whether animals can simply use any cue that predicts fitness enhancing

In this study, bumblebees,
Bombus t
, were used to establish whether the cues provided
by conspecifics are more salient than those provided by heterospecifics,
Apis mellifera
. Bees were trained
in a simple learning task to associate rewarding artificial flowers with a feeding consp
ecific or
heterospecific and the learning performance of this task was subsequently tested. When all cues predicted
equal rewards, bees showed a disproportionate degree of learning towards intraspecific cues, demonstrating
that conspecific information has
a stronger initial influence than information provided by a heterospecific
species. These findings expand the current understanding of social cue use in animals by illustrating that
bees, despite being able to use heterospecific information in a foraging c
ontext, show a stronger unlearnt
weighting towards conspecific information.

Prey pattern regularity and visual background complexity influence the detectability of
matching prey (poster)

Marina Dimitrova

& Sami Merilaita (Stockholm University,


In this study we test how aspects of prey colour pattern regularity affect crypsis and how visual complexity
of the background affects prey detection. We performed two predation experiments with artificial prey and
backgrounds, using blue tits (
yanistes caeruleus
) as predators. In the first experiment we found that a
prey pattern consisting of variable element shapes did not compensate for mismatching pattern elements,
because a variable pattern with some mismatching element shapes was easier to
detect than a variable
pattern with only background
matching element shapes. Contrary to a previous hypothesis, a pattern with
regular, background
matching element shapes was not easier to detect than the pattern with variable,
matching shapes.
All prey types were easier to detect on a simple than on a complex
background with more diverse and complex element shapes. In the second experiment we used prey
pattern consisting of regular, background
matching elements and tested how spatial regularity
of the
elements affects crypsis. We found that both on a simple and on a complex background the spatially
irregular prey with randomly placed pattern elements was more difficult to detect than the regular prey
with all elements aligned. Here background com
plexity was due to element shape complexity only. In
conclusion, our study shows that spatial regularity of prey pattern but not regularity due to invariable
pattern element shapes deteriorates crypsis. Visually complex habitat backgrounds and specifically

consisting of elements with complex shapes make detection of cryptic prey difficult.


Is Batesian mimicry effective against an invertebrate predator? (poster)

Rohanna Dow

and Tom Reader (University of Nottingham, UK)

It is common in nature for defe
nded prey species to possess signals advertising their defense to would
predators (aposematism). A Batesian mimic is a harmless prey species that has evolved signals similar to
those of aposematic species via exploitation of the signal
receiver relation
ship between predator and
defended prey (the “model”). Batesian mimicry theory predicts that there would always be strong selection
on the mimetic signals to resemble more accurately those of the model. There is, however, an abundance of
seemingly poor mim
icry. What can explain the evolution and maintenance of these conspicuous cheats?

a Batesian mimicry complex the predator is considered the main selecting agent shaping the honest signals
of the model and dishonest signals of the mimic. Different types
of predator with different sensory biases
and varying sensory acuity could therefore help to explain imperfect mimicry as perceived by humans
far the majority of work on Batesian mimicry has focused on the perception of vertebrate predators,
y birds. Using artificial and live prey I investigate which cues are important to the predatory crab
Symaema globossum
, when choosing prey. I show the importance of visual and possibly olfactory
cues as the spiders learn to avoid the mimetic “warni
ng” patterns of wasp
mimicking hoverflies following
previous experience with a wasp (
Polistes sp.
); and how the spiders appear to rank the hoverflies in order
of their mimetic accuracy in accordance with human volunteers.

Comparative study of the respon
ses of captive black howler monkeys (Alouatta caraya) to
playback experiments


Holly Farmer
, Lisa Leaver and Amy Plowman (University of Exeter, UK)

The performance of vocalisations is an important aspect of primate communication. The howler monkey


species) is characterised by the production of loud howl calls, which are not regularly heard in
In situ

studies have determined the functions of howls to include regulation of space use,
territorial demarcation, opponent assessment a
nd resource defence. We carried out playback experiments to
investigate responses of black howlers (
Alouatta caraya
) to four acoustic stimuli; howls of conspecific (
howls of a non
sympatric howler species (
A. palliata
), the roars of a different

primate species, the
red ruffed lemur (
Varecia rubra
and the low frequency sound of a chainsaw. We found that conspecific
playbacks were found to elicit howling in previously non
vocal males and the majority of howling was in
response to conspecific cal
ls, supporting the role of howling in inter
group competition. The fact that
howlers are clearly able to identify conspecific calls reinforces the importance of howls in the
communication system of howler monkeys. Thus, with the application of auditory sti
muli, we can mimic
the natural conditions of wild howler monkeys, encouraging the performance of natural (vocal) behaviours
in the captive environment.

Deceptive vocal mimicry by the fork
tailed drongo: adaptive benefits of a variable



Tom Flower

(University of Cambridge, UK)

Despite the prevalence of vocal mimicry in birds, its functions remain elusive. Several hypotheses

a function in deceptive communication. Vocal mimicry could be particularly advantageous in

communication, since receivers typically cease responding to a deceptive signal when made

too frequently,
and by changing the deceptive signal, a species could maintain deception. Here I

investigate the use of
vocal mimicry in the false alarm calling syste
m of

the fork
tailed drongo

I use
observational and experimental data to show that drongos use deceptive mimicked

alarm calls in addition
to their own alarm calls to scare two target species, meerkats
(Suricata suricatta)

and pied bab
enabling drongos to steal their food. Furthermore, I demonstrate

that pied babblers are
more likely to be deceived by mimicked alarm calls than drongo
specific alarm

calls and that by
employing vocal mimicry to vary their false a
larm call type, drongos could maintain

deception when pied
babblers cease responding to other, more frequently made false alarm calls. This

work demonstrates a
novel function for vocal mimicry in deception, and illustrates how in

communication systems, sig
may benefit from varying their strategy depending upon the response of receivers.


Interspecific communication
: how can it evolve? (poster)

Nathalie Gontier

(American Museum of Natural History, USA)

parasite interactions, plant
pollinator inte
ractions, and symbiotic processes pose a theoretical
difficulty for gene
centered evolutionary thought and the debate over the units and levels of selection,
according to which an organism’s fitness depends upon the exclusive survival and transmission of t
organism’s own genes. Inter
specific interaction and communication systems that occur over many
generations, demonstrate that genes alone are not the exclusive unit of evolution. Rather, different, non
genetic factors can also be the target of selection

or other evolutionary mechanisms such as symbiogenesis.
A consequence of this fact is that we have to conceptualize evolution as something that occurs at many
different levels of biological organization and according to many different units. This paper t
pleads for the recognition of unit plurality, multi
level evolution and mechanism plurality. A
methodological framework shall be presented that allows one to adequately search for the myriad of units,
levels and mechanism of evolution involved in
interspecific interactions.

Interspecific communication in chickadees: do black
capped and mountain chickadees
differentiate between each others chick
dee calls?(talk)

Angélique Grava
, Thibault Grava & Ken Otter (University of Northern British Columb
ia, Canada)

capped chickadees (
Poecile atricapillus
) and mountain chickadees (
P. gambeli
) share a similar vocal
repertoire, and previous research suggests that, when their population overlap, both species respond weakly
to heterospecific vocalizatio
ns. Black
capped and mountain chickadees recognise, but can discriminate
calls from their own versus from the other species. In this study we focused on the responsiveness to
conspecific vs heterospecific


calls, a complex vocalisation used in v
arious contexts such as
flock coordination in both species. We conducted a playback experiment where black
capped and
mountain chickadees in a sympatric population were tested in aviary trials. All birds were individually
exposed to playback composed of 9
calls from either species

these calls were recorded in the
context of flock members locating food sources. Responses of tested birds were measured by recording
vocalisations and behaviour. Black
capped chickadees responded significantly more
to conspecific stimuli
than to heterospecific stimuli; mountain chickadees did not show differential responses, replying to both
mountain and black
capped chickadee calls. As black
capped chickadees are the more dominant species in
this overlapping populat
ion, they may perceive intraspecific competition as greater than interspecific, and
thus differentially respond to black
capped calls. By contrast, interspecific competition with the more
dominant species may results in mountain chickadees not differentia
lly responding between species calls,
despite their known ability to discriminate heterospecific vocalizations.

The role of visual status signals in contests between a paper wasp and its social parasite (talk)

Jonathan Green

& Jeremy Field (University
of Sussex, UK)

The use of status signals is thought to be important in the establishment of dominance hierarchies in a
range of species. In the paper wasp
Polistes dominulus
, patterns of black pigment on the clypeus have been
shown to signal status to con
specific rivals. In its European range,
P. dominulus

is a host to the social
P. semenowi
, which invades host colonies and assumes the role of dominant reproductive.
P. semenowi

also sports patterns on the clypeus, similar to those se
en on the hosts, though
with a greater amount of pigment. In this talk, I present the results of an experiment which tested whether
parasite facial patterns signal parasite strength during nest usurpation. Parasite facial patterns were
manipulated and para
sites were then given the opportunity to usurp host nests. The importance of the facial
pattern, and also of parasite and host body size, for the dynamics and outcomes of the usurpation contests
is discussed.


Educated predators and the evolutionary dynam
ics of unequally defended co
mimics: A test of
Müllerian mutualism (talk)

Christina Halpin
, John Skelhorn & Candy Rowe (Newcastle University, UK)

If two species of defended prey living in the same environment share the same toxin, both species may

from the presence of the other by saturating predators’ detoxification processes and reducing
predation rates. This ‘toxic mutualism’ should occur even when the toxin content of each prey species is
unequal, and when they are visually distinct. However, t
he relative benefits to unequally defended prey are
predicted to depend upon the visual similarity of the prey: the more toxic species should benefit more when
prey are visually dissimilar compared to when they share the same coloration, i.e. when they are

In the first direct test of toxic mutualism, we used wild
caught European starlings (
Sturnus vulgaris
foraging on sequentially presented mealworms (
Tenebrio molitor
). Defended mealworms were injected
with one of two concentrations of quinine sol
ution to produce mildly and moderately defended prey. Birds
received either one or both types of defended prey, and their visual similarity was manipulated using
coloration. We found that predators regulate their toxin intake, and that toxic mutualism can
exist between
unequally defended prey. However, we failed to find any significant additional effect of visual mimicry on
the predation rates of either defended prey type. Intriguingly, we also found that birds ingested more toxin
when the mildly and modera
tely defended prey were presented together compared to when either prey type
was encountered alone. This suggests that the amount of toxin that predators eat depends upon the types of
defended prey available, and that simulating the degree of mutualism bet
ween defended species may be
more complex than previously thought.

Do you think I ate it? Owner perceptions and behavioral assessment of the “guilty look” in
dogs (poster)

Julie Hecht

and Márta Gácsi (University of Edinburgh, UK)

Using a questionnaire a
nd experiment, we investigated owner perceptions and pet dog behavior in relation
to an owner
reported anecdote. It is claimed that upon an owner returning home, a dog greets the owner
and sometimes displays “guilty” behavior, thereby alerting the owner to

a misdeed performed in the
owner’s absence. The questionnaire examined owners’ perceptions of dog “guilt”. The experiment
explored (1) whether dogs that were disobedient in owners’ absence show behaviors associated with
“guilt” upon owners’ return to a ro
om and (2) whether owners can determine dog disobedience based on
dog greeting behavior.

Questionnaire results from sixty
four owners revealed the following: dogs are said to display “guilty”
behavior in certain situations (87.5%); dogs display “guilty” b
ehavior before owners have discovered a
dog’s transgression (50%); “guilty” behavior implies dogs know that they have committed a disapproved
act (91%) and dog presentation of “guilty” behavior leads owners to scold dogs less (59%).

The experiment took
place at the Family Dog Project in Budapest, Hungary and included pet dogs (N=58),
their owners and an experimenter. The humans in the experiment aimed to communicate and establish the
social rule that food placed on a table was for humans and not dogs. Do
gs had the opportunity to
misbehave and eat the food when the owner and experimenter were in the room. If the dog ate, the owner
scolded, thereby explicitly communicating that this rule was in place. Dogs then had the opportunity to eat
the food after the
owner and experimenter left the room. The owner then returned to the room, where a
barrier blocked the owner’s view of the table and plate. The owner observed the dog’s greeting behavior
and tried to determine whether the dog ate or not. To examine dog gre
eting behavior when this social rule
was in place, the experimental procedure included two additional dog
owner greetings, one before the rule
had been established and another after the rule had been established but was no longer in place.

Dog behavioral
analysis revealed no significant difference in “guilty” behavior when greeting an owner
between dogs that ate and dogs that did not eat in the owner’s absence. Although owners appeared to
correctly determine whether or not their dog ate in their absence, t
hey seemed unable to base their answers
solely on dog greeting behavior and could have been relying on other contextual cues.

Although “guilty” behaviors could not be reliably observed in the present experiment, by owners or
behavior coding, owners do as
cribe “guilt” and “guilty” behavior to dogs. We propose these behavioral
displays could have an adaptive function in many dog
human social contexts. Further context
investigation of these behaviors is needed.


off between melanisation and
warning signal efficiency in the wood tiger moth
(Parasemia plantaginis

Robert Hegna
, Ossi Nokelainen, Carita Lindstedt and Johanna Mappes (University of Jyväskylä,

Evolution of conspicuous signals may be constrained if animal coloratio
n has nonsignaling as well as
signaling functions.
The wood tiger moth (
Parasemia plantaginis

is an aposematic species with wide
geographic distribution and distinct colour polymorhism. In addition to the different hindwing colours of
yellow and white,

bears three distinct hindwing melanisation patterns within Europe that
range from very little melanisation to nearly the entire hindwing being black in populations from the alps.

In this experiment we used artificial moths in the field to test

whether there is a possible cost to individuals
bearing more melanistic patterns (which are presumably better for thermoregulation purposes) via greater
predation risk due to decreased warning signal size. The experiment was conducted in the alps where a
three melanisation morphs co
exist and then replicated in Finland where two melanisation morphs co
We found that more melanistic individuals suffered increased predation, suggesting that warning signal
efficiency is traded
off with thermoregulat
ion. More melanic individuals are generally predicted to be
better able absorb radiation at a heightened rate compared to less melanic individuals which is likely to be
beneficial during takeoffs in low temperatures.
We conclude that the expression of apo
sematic display is
not only defined by its efficacy against predators; temperature may constrain evolution of a conspicuous
warning signal and maintain variation in it.

A preliminary investigation into the effects of owl calls on the behaviour of wild
rodents (poster)

Colin Hendrie

and Emma Story (University of Leeds, UK)

Laboratory studies have shown that several rodent species react to tape
recordings of owl calls (e.g.
Hendrie 1991, Eilam et al, 1999). However, it is not known whether si
milar responses are seen in the wild
and so the following studies were conducted in order to investigate this.

Animal activity was assessed within a 50 x 60 m

rectangle on part of a disused farm near Carlisle,
Cumbria by examining the marks left on tracki
ng plates that had been coated with lampblack (Quy et al,
1993). This and other evidence indicated that rats (Rattus norvegicus), mice (Mus musculus) and voles
Microtus agrestis) were active on this site.
Recorded Tawny owl (Strix aluco) calls were obtain
ed from the
British National Sound Archive and speakers (set to produce 50 dBA at thirty metres) were placed high on
the roof of a Dutch barn near the centre of the study site. A repeating loop of 6 minutes of owl calls,
containing both ‘hoots’ and ‘kee
cks’, followed by 30 mins of recorded silence or 36 mins of recorded
silence was played on study nights and activity assessed in and around 8 bait trays laden with 500g
sand/10g of millet seed placed at strategic locations around the site. Observations wer
e made over a total of
16 nights spread over a 3 month period. Three factor analysis of variance used to examine the effects of
stimuli (Owl call/no owl call), species (rats/voles;mice) and bait site (1
8) on total activity scores revealed
there to be sign
ificant main effects of stimuli (F(1,224) = 32.28, p = 0.00001), species (F(1,224) = 23.71, p
= 0.00001) and bait site (F(7,224) =17.79, p = 0.00001). Follow up tests revealed these effects to be due to
a reduction in the activity of the smaller rodents at

those sites most distant from overhead cover provided
by the buildings or hedgerow during those nights when owl calls were played.

living desert rodents have similarly been shown to concentrate their activity under bushes/overhangs
etc following the

flight of a trained owl overhead (Abramsky et al, 1996). The territorial calls of owls are
however, only signals of potential threat to animals that have developed the cognitive ability to interpret
them. Hence, the lack of effect of owl calls on the beha
viour of rats in this study was in keeping with
evidence to suggest that Tawny owls do not prey upon this species.


Social networks as models to represent contacts between species (poster)

Karlo Hock

and Nina Fefferman (The State University of New Jersey
, USA)

Social networks can be used to represent group structure as a network of interacting individuals, and are
growing in popularity as analytical tools for studying interactions in animal groups. The flexibility of such
frameworks is particularly attra
ctive, as it can be used to analyze, and also simulate, a wide variety of
interaction patterns and social contexts. To this end, we designed a series of simulation experiments based
on dynamic social networks to investigate how this versatile paradigm can
be used to represent contacts
between species. Using networks that feature social dynamics originating from individuals selecting social
partners according to a particular preference rule, we are able to explore the emergence, and the
significance, of cont
acts among distinct types of individuals. While the characteristics of particular
network components underlie the ability of such framework to represent individuals of different species, at
the same time the characteristics of links between the components
can represent different interaction
modalities that occur between these individuals. We discuss the results of our models from the aspects of
evolution of social behavior, ecology, conservation and communication patterns.

Bird predators do not select f
or polyphenism in the European map butterfly (poster)

Eira Ihalainen

and Carita Lindstedt
(University of Jyväskylä, Finland)

The European map butterfly (
Araschnia levana
) is known for its striking seasonal polyphenism which is
under photoperiodic control:

the spring generation is orange with black patterning whereas the summer
generation is black with white bands across the wings. Physiology of the map butterfly polyphenism is
better known than its ecology and evolutionary basis. Since predation is one of
the major selective forces
for insect colouration, we studied whether avian predators attack the spring and summer forms at different
rates. We first established in a laboratory experiment with blue tits (
Cyanistes caeruleus
) as predators that
the map butt
erflies are palatable and their colouration has no warning function, and also that birds are not
wary of either of the colour forms
per se
. We then pinned fake butterflies (paper wings with modelling clay
body) of both colour forms onto vegetation during t
heir peak flying seasons in the spring and summer. The
pinning experiment was conducted outside the range of the map butterfly in Finland, so the local birds had
no prior experience of the species. After two days we checked the specimens for beak marks to
mortality from bird predation. There was no difference in survival of the colour forms in either spring or
summer suggesting that avoiding predation e.g. by more efficient camouflage according to season does not
explain polyphenism in the map butte

Selection for Müllerian mimicry in simple and complex prey communities (poster)

Eira Ihalainen
, Hannah Rowland, Michael Speed, Graeme Ruxton and Johanna Mappes
(University of Jyväskylä, Finland)

Müllerian mimicry where two or more unrelated apose
matic prey species resemble each other is assumed
to evolve because sharing a warning signal decreases per capita mortality that is due to sampling by naïve
predators learning to avoid the signal. Despite that Müllerian mimicry is one of the celebrated exa
mples of
natural selection, the mechanism behind its evolution is still debated. It has been suggested that mimicry
evolves more readily in a complex rather than a simple prey community. We show that, contrary to
expectation, accurate signal mimicry is mor
e likely to evolve in a simple prey community. We studied the
selection for mimicry in an experiment with wild great tits (Parus major) as predators of artificial prey in a
laboratory. The birds first learnt to discriminate between aposematic and edible pr
ey by foraging in a prey
community that was either simple (few prey appearances) or complex (several prey appearances). Despite
that discrimination between edible and defended prey was more difficult for the birds foraging from
complex prey communities, on
ly birds that foraged from simple communities subsequently selected against
imperfect mimics and for accurate mimics of a specific model signal they had encountered during learning.
Although aposematic prey in complex communities would benefit of mimicry w
hen facing naïve predators,
the evolution of mimicry seems to require predators that are knowledgeable about their prey either through
experience, specialization or a simple prey community.


The ability to individual discrimination via heterospecific uri
ne odor in Djungarian hamsters
(Phodopus sungorus Pallas, 1773) depends on donor species (poster)

Anastasia Khruschova

and Vasilievab Yu (University of Moscow, Russia)

We may assume that heterospesific discrimination plays an important role in multi
ies communities,
provide an effective
predator defense and

pave the way for domestication. However such ability is not
well documented. The aim of our study was to
demonstrate that Djungarian hamster males can
discriminate individual urine odors from
donors of different taxa varied in the level of relatedness
(including humans and a potential predator,
Felis catus
) with the use of habituation
paradigm. Our results indicate that hamsters may process the chemosensory signatures of individu
als from
not related species with the only exception

cat. The lack of dishabituation to

odor may indicate
the lack of discrimination and permanent v
igilance/risk assessment toward it. In its turn this may represent
species specific anti

behavioral strategy of Djungarian hamsters.

Eyespots as anti
predator device: Both intimidating

and deflective effect found against
predation by fish (talk)

Karin Kjernsmo
, Sami Merilaita & Annika Wiksten (Åbo Akademi University, Finland)

Eyespots (
colour patterns consisting of concentric rings) are found in a wide range of animal taxa and are
often assumed to have anti
predator function. Previous experiments with birds as predators have shown
that some eyespots have an intimidating effect. It has
also been hypothesized that in some species eyespots
could have a deflective function. These eyespots would divert attacks towards less vital parts of prey body
and hence increase the probability of the prey surviving an attack. There exist, however, no
firm studies
about their effect in aquatic environments, although eyespots have indeed initially evolved there. Our aim
was to investigate the protective function of eyespots against fish predators. We used different artificial
prey items to test both th
e intimidation and deflective hypothesis. We used three
spined sticklebacks
Gasterosteus aculeatus
) as predators. Our results show that large eyespots may cause hesitation in
attacking fish. Interestingly, small eyespots caused a strong deflective effe
ct in fish.

Male moths provide pollination benefits in the Silene latifolia

Hadena bicruris nursery
pollination system (poster)

Marie Labouche

and Giorgina Bernasconi (University of Neuchâtel, Switzerland)

Evolutionary conflicts of interest under
lie mutualisms, including “nursery pollination”, in which the
pollinators lay eggs inside the flowers and offspring consume the developing seeds. Low benefit to cost
ratios could destabilize such associations towards parasitism. Although in most of the wel
known cases
pollen transfer is associated with oviposition, in some systems the males of the seed predator may increase
the pollination benefits, and thus affect the outcome of the interaction. We investigated the dioecious plant
Silene latifolia

hyllaceae) and its nursery pollinator,
Hadena bicruris

(Noctuidae). Data on
visitation behaviour and pollination efficiency in experimental plant patches demonstrate that (i) male
moths are equally or even more efficient pollinators than females, fertilizi
ng on average 45% of the ovules
in one visit, (ii) female and male moths do not preferentially visit one plant sex, and (iii) feeding behaviour
is sufficient to ensure pollen transfer. This indicates that male moths contribute to seed production,
g the net benefits of this interaction, even before accounting for seed predation costs. Altogether
this suggests that both moth sexes provide a pollination benefit to the plant with no differences in
pollination efficiency but that female moths, before se
ed predation costs are accounted for, seem to provide
greater benefits owing to their increased activity. That male moths contribute to seed production likely
decreases the plant’s dependency on ovipositing moths for pollination.


Sounds produced by loco
motion (SOL) in inter
specific communication (talk)

Matz Larsson

(Universitetsjukhuset Örebro, Sweden)

Sounds produced by locomotion (SOL) and the impact of synchronized movements on animal perception
and behaviour have been little discussed. Fish (i.e. b
ird ancestors) may benefit from synchronized
movements: (A) It may facilitate detection of vital signals due to silent intervals and enhanced auditory
grouping. (B) Schooling may confuse predators’ octavolateralis system due to overlapping hydrodynamic
nals. (C) Prey fish that swims close may blur predators’ electrosensory system. Bird flight may produce
significant noise. Many birds produce wingbeats that are readily heard on the ground. The highly
synchronized formations that characterize the flight of

some bird groups have been explained as reducing
energy expenditure. An alternative (or complimentary) hypothesis is that synchronization of movements in
bird flocks may facilitate detection of vital signals. Acoustical advantages may also be achieved by
locomotor coupling and intermittent flight. Interactions between locomotion and vocal calls,
and known and potential roles of SOL in fish and bird communication will also be discussed.

Optimal toxicity and anti
predator signalling in model/m
imic populations (poster)

Thomas Lee
, Michael Speed, Hannah Rowland and Graeme Ruxton (University of Liverpool,

Defensive mimicry between prey species has served for over a century as key evolutionary case studies in
parasitism and arms races (Batesia
n mimicry) and mutualism (Mullerian mimicry). All existing models of
mimicry focus on the evolution of the signal, for example predicting whether parasitic mimics chase the
aposematic signal of their model, but they do not consider that the edibility and t
oxicity of the prey may
also coevolve with prey signals.

In this talk I will outline a new theoretical framework within which prey species can evolve both in their
appearance and in their level of a toxin secondary defence. Simulations predict that over
the course of
evolution, mimetic relationships and the effects mimetic signals can change profoundly, for example being
mutualistic for one phase, but parasitic for another phase. For example, once engaged as a mimic,
individuals of an ancestrally aposemat
ic defended species can reduce all toxin investment to become an
undefended Batesian mimic, switching from initial mutualism to total parasitism in the process. We show
also that mutualistic Mullerian mimicry between two defended prey forms not only reduce
s predation, but
allows both species to reduce their investment in costly defences. Hence Mullerian mimicry evolves not
only to prevent attack, but also to make the fitness costs of defence much cheaper.

Diet quality affects warning coloration indirec
tly: excretion costs in a generalist herbivore

Carita Lindstedt
, Talsma Reudler, Eira Ihalainen, Leena Lindström and Johanna Mappes
(University of Cambridge, UK)

Aposematic herbivores are under selection pressure from their host plants and predat
ors. Although many
aposematic herbivores exploit plant toxins in their own secondary defense, dealing with these harmful
compounds might underlay costs. We studied
whether the allocation of energy to detoxification and/or
sequestration of host plant defens
e chemicals trades off with warning signal expression. We used a
generalist aposematic herbivore
Parasemia plantaginis
(Arctiidae), whose adults and larvae show
extensive phenotypic and genetic variation in coloration. We reared larvae from selection lines

for small
and large larval warning signals on
Plantago lanceolata

with either low or high concentration of iridoid
glycosides (IGs). Larvae disposed of IGs effectively; their body IG content was low irrespective of their
diet. Detoxification was costly as

individuals reared on the high IG diet produced fewer offspring. The IG
concentration of the diet did not affect larval coloration (no tradeoff) but the wings of females were lighter
orange (vs. dark red) when reared on the high IG diet. Thus, the differe
nce in plant secondary chemicals
did not induce variation in the chemical defense efficacy of aposematic individuals but caused variation in
reproductive output and warning signals of females.


Behavioural coordination in rooks and jackdaws (poster)

a Logan
, Nathan Emery and Nicola Clayton (University of Cambridge, UK)

The ability to coordinate behaviours is advantageous for carrying out activities such as allied predator
defense, group food acquisition, and caring for young (Bird 2009). Bird (2009)
investigated coordinated
behaviours in rooks
(Corvus frugilegus)
and found that the occurrence of bill wiping, fluffing, and bowing
increased after one individual demonstrated the behaviour. Fluffing and bowing spread to the whole group
(though most stron
gly to the partner) while bill wiping spread to the partner. We hypothesized that,
because rooks and jackdaws
(C. monedula)
are social species it will be advantageous to coordinate
behaviours at the pair and group level, and since they often associate with

each other in the wild, they may
coordinate behaviours between species. We found that fluffing and bill wiping were coordinated and
spread among conspecifics, and, though not a coordinated behaviour, preening spread from rooks to
jackdaws. Rooks tend to
be the dominant species when rooks and jackdaws are together, which could
account for why jackdaws attend to specific social cues in rooks and not vice versa. While jackdaws do not
attend to coordinated behaviours in rooks, they do pay attention to preenin
g which is important for
maintaining health when individuals gather in groups. Examining those behaviours that are advantageous
to both species when together (
, foraging and alarm calling) may provide more examples of interspecies

effect of plant
pollinator interactions on morphological and neutral genetic divergence of
populations of the white campion Silene latifolia and its pollinator Hadena bicruris (poster)

Isabel Santos Magalhaes
, Anne
Marie Labouche, Gabriela Gleiser and Gior
gina Bernasconi
(University of Neuchâtel, Switzerland)

pollinator interactions have been suggested as key drivers of morphological divergence and even
speciation in plants. Here we tested this hypothesis in populations of the white campion Silene la
tifolia that
are pollinated by the specialist frugivore moth, Hadena bicruris. We examined the geographic variation
across eight European populations in traits potentially involved in this complex plant
insect interaction,
that entails both mutualistic (po
llination) and antagonistic components (frugivory). We compared
divergence in phenotypic traits and neutral genetic markers among co
occurring populations for both plants
and insects. Finding significant and strong divergence in traits involved in plant
llinator interactions
would support the hypothesis that divergent selection driven by the interaction of the two organisms and
acting on several independent traits is sufficiently strong to maintain differences despite potential gene
flow among populations

Interspecies cooperation

Jays solve a cooperative task with a human partner but not with
another jay (poster)

Ljerka Ostojic
, Nathan Emery and Nicola Clayton (University of Cambridge, UK)

Cooperation is one of the cornerstones of human psychology a
nd thought to be a critical component of our
complex social cognition. Even though it is controversial whether non
human animals possess joint
communicative communication (Moll & Tomasello, 2007), some primates have been shown to solve
cooperation tasks. I
ndeed there is good evidence that primates can understand when a partner is required to
solve the task, and how to coordinate actions among the partnership (chimpanzees: Melis et al., 2006a,
2006b; Hirata & Fuwa, 2007; bonobos: Hare et al., 2007; capuchin
monkeys: Mendres & de Waal, 2000;
top tamarins: Cronin et al., 2005). These abilities are not restricted to primates however, because
spontaneous cooperative problem solving abilities have also been shown in rooks, a colonial corvid species
(Seed et

al., 2008). At issue is whether this ability is common to all corvids or only colonial species with
high intraspecific tolerance and synchronicity between mated partners (Emery et al., 2007).

To address this question we tested the Eurasian jay (
), a non
colonial corvid, which
displays highly developed social skills in the context of caching: for example they keep track of who was
watching when (Dally et al., 2006) and go to great lengths to protect their caches of potential thieves
yton et al., 2007; Goodwin, 1955; 1956). As these birds are territorial with usually low levels of
intraspecific tolerance, we tested the jays under two conditions; either with a conspecific or with a human
partner. We found that the jays did not spontaneo
usly cooperate with a conspecific as we might predict
from their low levels of intraspecific tolerance. However, they did cooperate with a familiar human
experimenter, the implications of which will be discussed.


parrot interactions: how to commu
nicate? (poster)

F. Péron
, N. Giret, M. Kreutzer, L. Nagle and D. Bovet (Université Paris Ouest, France)

Psittacids, and grey parrots more precisely, are very popular as pets. Indeed, their cognitive and
communicative abilities represent something attract
ive for the owners (or breeders). Since the late
seventies several studies have been conducted on African grey parrots, in particular with Alex who change
the way to consider bird intelligence. In their daily life of captive animals, these social birds ten
d to
communicate with their owners. At the same time owners tend to attribute intentions to their parrot as their
imitation skills are amazing. Humans could also learn to associate specific sound to particular objects or
events and thus create a new refere
ntial communication system. Nevertheless vocal communication is not
the only way to exchange information between humans and parrots. I will present a short review of the last
findings regarding African grey parrots’ social cognition: how they used human cu
es such as gazing or
pointing, how they can have perspective taking or adapt their behaviour according the intentions of their
caretaker and so on.

Dogs’ vocalization reveals context and inner state for humans

a result of domestication?



(Eötvös Loránd University, Hungary)

Dogs have lived with humans for about 15,000 years. During this time these carnivores have developed the
most intricate coexistence with our species. Earlier research has shown that dogs are attentive to and c
understand several forms of human communication. Our current focus regards how informative vocal
signals of dogs can be for humans. In an extensive series of experiments, human participants were asked to
categorize pre
recorded sequences of dog barks an
d additionally evaluate the possible inner states of the
animals as well. The range of human participants included young children, adults with differing amounts of
previous dog experience and also congenitally blind adult humans. Our results have shown tha
t humans in
all age, dog experience and vision groups, can recognize over chance level the contexts of dog barks. There
was however a difference in how accurately they recognized the six possible contexts. The human
participants also had an unambiguous opi
nion about the inner states of dogs based on the individual bark
samples. An acoustical analysis found several correlations between particular acoustic parameters and
features of the reported inner states. As we found recently that dog growls carry context
information for other dogs, we can hypothesize that barks became more variable and ‘understandable’ for
humans during the domestication process, thus helping the effective coexistence of the two species.

making in the face of bats: de
fensive sound production in tiger moths


John Ratcliffe
(University of Southern Denmark, Denmark)

Like insectivorous birds, insectivorous echolocating bats remember the warning

signals of chemically
defended prey. And do so by readily associating th
e ultrasonic clicks of

tiger moths with these insects’
unpalatable taste. While bat
moth interactions are similar

those of birds and butterflies in some ways,
they are also remarkably different. First, birds use

multimodal information for prey discrimin
ation; bats
and moths communicate almost entirely

within an auditory world. Second, the warning sounds produced
by toxic tiger moths are elicited

only in response to successive bat echolocation calls arriving at the moth’s
ear at intensities and

rates indi
cative of a hunting bat. That is, tiger moths only answer nearby hunting bats.

active and acoustic warning display is in contrast to the passive and constantly displayed visual

warning signals of many butterflies. Tiger moths thus provide a simple sys
tem for the study of

assessment within a single sensory modality. I will discuss our on
going research into the

evolution of tiger
moth sound production and the mechanisms that underlie it.


Greater anis discriminate between predatory yellow
caracaras and the similar, non
predatory snail kite (poster)

Christina Riehl

(Princeton University, USA)

The Greater Ani (
Crotophaga major
) is a cooperatively breeding Neotropical cuckoo; breeding groups are
composed of up to four socially monogamous pair
s, all of which lay eggs in a single nest and participate in
incubation and parental care. Nest defence is one of the most important benefits of social nesting: group
members cooperate to mob predators at the nest, and the likelihood of successfully repell
ing a predator
increases with group size. Here I show that Greater Anis are able to discriminate between Yellow
Caracaras (
Milvago chimachima
), which frequently prey upon ani chicks, and Snail Kites (
), which specialize on snail
s and do not prey upon ani chicks. The two raptors are similar in
appearance in all plumages, and immature birds are difficult for human observers to tell apart without
binoculars and experience. In a four
year field study in Panama, adult Greater Anis gav
e alarm calls
significantly more often in response to fly
overs by caracaras than by kites. In addition, fledgling anis were
observed to give alarm
calls at caracaras when their parents were present and alarm
calling, but not when
their parents were absent
. These results indicate that the ability to discriminate between predators and non
predators may be learned during the nestling and early fledgling period.

Effects of weaver ants on the relationships between Melastoma malabathricum and its flower
rs (poster)

Miguel Rodriguez

and Francisco Gonzalvez (CSIC, Spain)

When foraging bees face a trade
off between maximising intake rate and minimising predation risk, they
can approach fitness maximisation following the strategy that maximises the r
atio between intake rate and
predation risk. In the face of competition for resources, and when co
existing bees differ in their
susceptibility to predation, models predict that susceptible bees should concentrate their foraging effort on
those flowers whe
re predation risk is minimal, while less susceptible bees should forage preferentially on
risky flowers, where competition for resources is reduced. We tested these predictions comparing the
behaviour of small and large bees collecting pollen at

flowers in a population
where some plants, but not others, had been colonised by weaver ants,
Oecophylla smaragdina
. Ants
tended aphids at the base of flowers and chased intruders away. Flowers were visited mainly by large
carpenter bees,
ocopa latipes

X. confusa
, and medium bees,
Nomia strigata
. Ants seem unable to
capture carpenter bees, but easily handle

and other bees of their same size. In agreement with model
predictions, carpenter bees foraged mainly at ant
colonised plants
, which were avoided by


bees are pollen thieves at
M. malabathricum
, the presence of territorial ants increased the
reproductive success of the plants.

Is the high variability in colour patterns of the dyeing poison a constrain
t for aposematism?

Bibiana Rojas

and Diana Pizano (Deakin University, Australia)

The bright colours of poison frogs are commonly assumed to advertise their toxicity. Although not much is
known about the predators of these frogs in the wild, field

studies with plasticine models suggest that birds
would be among their most important predators. For aposematism to work, predators should learn to
associate the colouration of the prey with a bad experience upon attack. High variability in prey
on patterns makes this learning process difficult and therefore aposematic prey are expected to
have a rather uniform colour pattern. In highly polymorphic populations individuals would be expected to
suffer morph
specific attack rates because morphs diffe
r in conspicuousness. Individuals might need to
complement their toxicity with some strategy that minimizes attacks while predators learn. There is
evidence that an increase in aposematic brightness enhances predator learning, therefore, one would expect
hat individuals with duller patterns would be more frequently found in places where they are less exposed
or closer to a hiding shelter. We tested this hypothesis in a highly polymorphic population of the dyeing
poison frog,
Dendrobates tinctorius
, in Fren
ch Guiana by recording the microhabitat of 170 frogs during
censuses carried out in a longitudinal transect. Preliminary results show that there is no relationship
between the morph and the habitat in which it is more frequently found, suggesting that brig
ht colouration,
despite its variability, is enough for aposematism to work in these frogs.


Mimicry between unequally defended prey can be parasitic: Evidence for Quasi


Hannah Rowland

Johanna Mappes, Graeme Ruxton &

Michael Speed (U
niversity of Liverpool)

Conventional views of aposematic signal mimicry hold that edible prey are parasitic Batesian mimics,
whereas chemically defended prey are mutualistic Müllerian mimics. When
prey are both defended, but to
unequal extents, a
ly defended species is hypothesised to dilute the protection of a better defended
Batesian mimicry). Definitive evidence for this is lacking. In th
is study
we show that
mimicry between unequally defended co
mimics can be parasitic and quasi
Batesian. We tested the
foraging behaviour of great tits (
Parus major
), in an experiment designed to simulate the evolution of

We kept the abundance of a highly defended Model and a moderately defended dimorphic
(perfect and distinct) Mimic const
ant while varying the relative frequency of the two Mimic forms
found a combination of parasitic and neutral mimicry; as the perfect Mimic increased in abundance,

predation on the Model
Mimic pair increased. However, when perfect Mimics were

rare there was
no effect on the survival of the Model: perfect Mimics gained protection from predation but imposed no
coevolutionary pressure on Models.
We found that the feeding decisions of the birds were affected by their
individual toxic burdens. Bird
s with high toxin loads retained their aversion to Models and Mimics even
when both became completely edible, whereas individuals with low toxin burdens did not. This is
consistent with t
he idea that predators make foraging decision which trade off
y and nutrition

conclude that the conventional distinction between Batesian and Müllerian mimicry requires extension to
incorporate parasitic defended mimic species.

Are warning signals honest?


Michael Speed

(University of Liverpool, UK)


animals signal their toxicity with bright, aposematic displays. This talk will examine whether
aposematic displays are likely to function as “honest signals”, in the sense that brighter displays reliably
signal the level of toxicity that an individual pos
esses. A number of recently published datasets, with frogs,
seaslugs, and beetles do show positive correlations between toxicity and signal brightness. In contrast
prevailing theory predicts the converse: that if there are costs to signalling, animals whic
h are highly toxic
should be more cryptic than those which are less well defended. In an attempt to reconcile theory and data,
I will discuss new explanations for honest signalling in aposematism. One of these explanations assumes a
direct physiological li
nkage between aposematic display and secondary defences; another focuses on the
capacity of predators to stablise signal honesty by filtering out “cheats” from the prey population. Though
the new models provide a good qualitative match to the new datasets,

I argue that positive correlations
between toxicity and signal brightness are not in themselves sufficient to assume that the signal has
evolved "honesty"; rather these correlations may arise as byproducts of other evolutionary outcomes. The
for future empirical research will be discussed.

Colour change in crab spiders: aggressive and protective crypsis, prey attraction, or avoidance
of UV photodamage?


Marc Théry
, Jérémy Defrize, Teresita Insausti & Jérôme Casas (Muséum National d’His
Naturelle, France)

Crab spiders constitute an excellent model to study interspecific visual communication because, depending
on the species, they have been shown to use colour change either to hide simultaneously from predator and
prey, or to attrac
t prey to the flowers they sit on. Early studies have been conducted on relatively small
samples of European spider and flower species. They showed either a chromatic or an achromatic match of
spiders on flowers. However, recent systematic field survey of
Misumena vatia
on all flower species on
which spiders were observed showed that although crypsis was always achieved through achromatic
contrast, spiders were mostly chromatically detectable in both bee and bird visions. Nevertheless, spiders
were always c
hromatically undetectable for fly prey. Even if European crab spiders achieve some degree of
crypsis, colour
matched spiders do not have a higher encounter rate or capture success compared to
conspicuous spiders. By comparing the pigmentation of cryptic an
d non
cryptic crab spiders, we suggest
that ommochromes used for colour change might be primarily used to protect the transparent cuticle from

UV photodamage. This is

supported by the fact that, contrary to European species, Australian crab spiders
ently reflect UV, which may be related to higher UV radiation in Australia compared to Europe.
Camouflage or prey attraction profit from ommochrome pigmentation, but may not be the driving forces.


Dietary conservatism and interspecific visual signals


Rob Thomas

& Nicola Marples (Cardiff University, UK)

Some individual birds and fish persistently avoid perfectly edible food, simply because its colour is novel
to them. It has recently been shown that this apparently maladaptive behaviour comprises t
wo avoidance
mechanisms; neophobia which is a general avoidance of contact with novel objects, and lasts a matter of
hours, and dietary conservatism (DC), which is a much longer
lived and robust to experience of novelty,
and is related specifically to eati
ng of prey. However DC is only present in a proportion of the members of
a population. The remainder of the population are Adventurous Consumers (AC), consuming novel prey
readily once their neophobia has waned.
These contrasting foraging strategies appear

to be a dichotomous
response rather than a continuum, with an underlying genetic basis in that individuals breed true for either
being DC or being AC. Dietary Conservatism has important implications for the origin and maintenance of
interspecific visual s
ignals between prey species and their predators, since prey can be protected from DC
predators simply by being novel in appearance.
In order to investigate the prevalence of dietary
conservatism among visually foraging animals, we tested for its presence a
nd frequency in a range of
visually foraging taxonomic groups, both among vertebrates and invertebrates. All vertebrate species tested
included DC and AC individuals in their population, while none of the invertebrates tested displayed DC
foraging strategi
es. The presence of a stable polymorphism of foraging strategies in every population of
such phylogenetically distant species as the vertebrates tested, living in completely different foraging
environments, is a very striking result. The implications of th
ese results will be discussed in relation both
to evolutionary processes giving rise to such foraging strategies, and the ecological conditions maintaining
them now.

Triplex relationship between bumblebees, flowers and flower
dwelling predators: shift
ing of
foraging behaviour in bees facing different flower rewards and predation risk (poster)

Yun Wang
, Thomas Ings and Lars Chittka (Queen Mary University, London, UK)

prey and pollinator
plant interactions have both played important roles in

animal communication
ecology. However few studies have discussed the tripartite relationship in the same scheme. Bees are
known to be able to differentiate flowers differing in reward levels, and can also detect camouflaged
predators that blend into the s
urroundings amongst flowers. Some species of the cryptic predators, such as
crab spiders, can change their body colour to fit the flower they sit on and also select the high quality
flowers. We therefore tested whether bees are able to continue discriminat
ing between flowers of different
quality, but similar appearance, when they also have to detect camouflaged spiders. We first trained the
bees to two rewarded flowers with two similar colours which are difficult, but possible for bees to
distinguish, and t
hen added cryptic and conspicuous artificial “crab spiders” on 25% of the high reward
flowers. Bees were attacked (held by the “arms” of the spiders for two seconds) when land on flowers with
spiders. The results showed that
when bees encountered flowers w
ith different reward but similar
appearance, they can choose the highly rewarding flowers, but take longer decision time. When the highly
rewarding flowers harbour spiders which were easy to detect, the bees can still choose highly rewarding
flowers and av
oid flowers with spiders at the same time. Most bees managed to entirely avoid dangerous
flowers at the end of the test. When the spiders were cryptic, the bees can decrease capture rate by spiders
after training, but still made some mistakes at the end of

the test. Moreover, bees switched to visit safe
highly and low rewarding flowers at random instead of choosing highly rewarding flowers.

Cryptic crab
spiders may have indirect effects on plant reproductive success by reducing visitation rates of pollinato
When high reward flowers harbouring cryptic predators, pollinators may switch to visit other plant species.


Death feigning, an anti
predator behaviour, in the context of sexual selection in the confused
flour beetle (poster)

Corin White
, Shelli

James, Cherise Washington, Isadora Porter and Aditi Pai (Spelman College,

Death feigning, also known as thanatosis and tonic immobility, is the mimicry of a “death
like” stance in
which the subject lacks motion. This is an anti
predator defensive be
havior. We observed that death
feigning varies across populations of the storage pest beetle,
Tribolium confusum.

Although, in general,
rates of thanatosis are similar for males and females, in one particular population, we found that females
feign death m
ore frequently in the presence of males. In addition, we also found that copulation success
correlated positively with female death feigning rates. From this we hypothesized that males manipulate
the females into feigning death, possibly via a chemical cu
e, to facilitate intromission and hence the
success of the copulation. We are presently investigating this hypothesis in our lab. Results from this study
will be discussed in the context of evolution of interspecific communication.

Novel cues produces
by the floral epidermis for insect pollinators (talk)

Heather Whitney

(University of Bristol, UK)

The evolution of many floral features has been driven by coevolution with, and adaptation to, animal
pollinators, and many of these features provide examples

of interspecific communication. One floral
feature that appears to be intricately linked to animal pollination is that of the structure of the petal
epidermis. The petal epidermis is very rarely smooth, but can show petal specific structures at both the

and the nano

scale. Structural features at both scales have been shown to influence pollinator
perception and preference and therefore to be of importance in flower
pollinator interactions. At the nano
scale cuticular striations have been found to b
e produced in both a sufficiently highly
ordered state and at
the right periodicity to act as a diffraction grating. These structures interact with light such that they
produce a form of structural colour known as iridescence. Floral structural colour char
acterised to date is
particularly strong in the ultra
violet wavelengths, which is visible to many animal pollinators.

Using lab
based bumblebee and biomimetics methods, the importance of epidermal patterning in
emphasising floral pigment patterning and en
hancing pollinator foraging efficiency can be investigated.
Biomimetic replicas of the petal surface of iridescent flowers can be used to establish if bumblebees can
use floral surface structure, or any of its associated non
chromatic phenomena, as a cue.

Do honeybees perceive floral scents invariantly with respect to concentration? (talk)

Geraldine Wright

& Amir Choudhary (Newcastle University, UK)

Reliable signal detection underlies the use of signals between species. Floral scents are signals that
pollinators such as honeybees learn to associate with reward. However, floral scents are highly variable in
concentration as they are emitted, and such variation could influence whether bees always recognize floral
signals. For example, humans perceive qua
litative changes in some odours as a function of their
concentration. Here, we explore the extent to which scent concentration influences how honeybees
perceive floral scents.
Using a peak
shift paradigm for examining generalization gradients, we present
vidence which indicates that honeybees do not perceive floral scent signals invariantly as a function of
their concentration. Our data suggest that the honeybee’s olfactory system cannot be both ‘general’ and
‘invariant’ and suggest that the olfactory syst
ems of animals may be subject to greater constraints than
those placed on other sensory systems.