ArticlePDF Available

Abstract and Figures

Many prey species contain defensive chemicals that are described as tasting bitter. Bitter taste perception is, therefore, assumed to be important when predators are learning about prey defenses. However, it is not known how individuals differ in their response to bitter taste, and how this influences their foraging decisions. We conducted taste perception assays in which wild-caught great tits (Parus major) were given water with increasing concentrations of bitter-tasting chloroquine diphosphate until they showed an aver-sive response to bitter taste. This response threshold was found to vary considerably among individuals, ranging from chloroquine concentrations of 0.01 mmol/L to 8 mmol/L. We next investigated whether the response threshold influenced the consumption of defended prey during avoidance learning by presenting birds with novel palatable and defended prey in a random sequence until they refused to attack defended prey. We predicted that individuals with taste response thresholds at lower concentrations would consume fewer defended prey before rejecting them, but found that the response threshold had no effect on the birds' foraging choices. Instead, willingness to consume defended prey was influenced by the birds' body condition. This effect was age-and sex-dependent, with adult males attacking more of the defended prey when their body condition was poor, whereas body condition did not have an effect on the foraging choices of juveniles and females. Together, our results suggest that even though taste perception might be important for recognizing prey toxicity, other factors, such as predators' energetic state, drive the decisions to consume chemically defended prey.
Content may be subject to copyright.
The ocial journal of the
International Society for Behavioral Ecology
Address correspondence to L.Hämäläinen. E-mail:
© The Author(s) 2019. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology.
All rights reserved. For permissions, please e-mail:
Original Article
Predators’ consumption of unpalatable prey
does not vary as a function of bitter taste
Liisa Hämäläinen,a,b, Johanna Mappes,b, Rose Thorogood,a,c,d, JanneK. Valkonen,b,
Kaijamari Karttunen,b Tuuli Salmi,b and HannahM. Rowlanda,e,f,
aDepartment of Zoology, University of Cambridge, Downing Street, CB2 3EJ Cambridge, UK,
bDepartment of Biological and Environmental Sciences, PO Box 35, University of Jyväskylä, 40014
Jyväskylä, Finland, cHiLIFE Helsinki Institute of Life Science, PO Box 65, University of Helsinki, 00014
Helsinki, Finland, dResearch Programme in Organismal & Evolutionary Biology, Faculty of Biological
and Environmental Sciences, PO Box 65, University of Helsinki, 00014 Helsinki, Finland, eInstitute
of Zoology, Zoological Society of London, Regent’s Park, NW1 4RY London, UK, and fMax Planck
Institute for Chemical Ecology, Hans-Knöll-Straße 8, 07745, Jena, Germany
Received 28 June 2019; revised 29 October 2019; editorial decision 5 November 2019; accepted 12 November 2019.
Many prey species contain defensive chemicals that are described as tasting bitter. Bitter taste perception is, therefore, assumed to
be important when predators are learning about prey defenses. However, it is not known how individuals differ in their response to
bitter taste, and how this influences their foraging decisions. We conducted taste perception assays in which wild-caught great tits
(Parus major) were given water with increasing concentrations of bitter-tasting chloroquine diphosphate until they showed an aver-
sive response to bitter taste. This response threshold was found to vary considerably among individuals, ranging from chloroquine
concentrations of 0.01mmol/L to 8 mmol/L. We next investigated whether the response threshold influenced the consumption of de-
fended prey during avoidance learning by presenting birds with novel palatable and defended prey in a random sequence until they
refused to attack defended prey. We predicted that individuals with taste response thresholds at lower concentrations would consume
fewer defended prey before rejecting them, but found that the response threshold had no effect onthe birds’ foraging choices. Instead,
willingness to consume defended prey was influenced by thebirds’ body condition. This effect was age- and sex-dependent, with adult
males attacking more ofthedefended prey when their body condition was poor, whereas body condition did not have an effect on
the foraging choices of juveniles and females. Together, our results suggest that even though taste perception might be important for
recognizing prey toxicity, other factors, such as predators’ energetic state, drive the decisions to consume chemically defended prey.
Key words: aposematism, avoidance learning, bitter taste, chemical defense, great tits, toxins.
Aposematic prey species have evolved diverse chemical defenses,
including cardiac glycosides, alkaloids, and iridoid glycosides (Blum
1981), and often signal their defenses to predators with conspicuous
warning colors (Poulton 1890; Ruxton etal. 2018). These chem-
ical defenses are often described as bitter tasting (Brower and Fink
1985; Glendinning 1994) and they typically generate aversive re-
sponses in predators, including head shaking, bill/mouth cleaning,
and spitting out of food. Avian predators learn to avoid aposematic
prey based on bitter taste (e.g., Skelhorn and Rowe 2006; Skelhorn
and Rowe 2010) and birds’ responses to bitter tastes can also pro-
vide other predators with social information about prey quality
(Johnston et al. 1998; Skelhorn 2011; Thorogood et al. 2018;
Hämäläinen et al. 2019a). Bitter taste perception is therefore as-
sumed to be important when predators are gathering information
about prey profitability.
How predators respond to chemically defended prey, how-
ever, varies among predator species (Endler and Mappes 2004).
Some species may, for example, be more resistant to prey toxins
and consume chemically defended prey that are unpalatable and
toxic to other predators (Fink etal. 1983; Brodie and Brodie 1990;
applyparastyle "g//caption/p[1]" parastyle "FigCapt"
Behavioral Ecology (2019), XX(XX), 1–10. doi:10.1093/beheco/arz199
Downloaded from by Jyvaskylan yliopisto / Kirjasto - kausijulkaisut user on 04 December 2019
Behavioral Ecology
Glendinning etal. 1990). Some predator species have also evolved
apparently taste-guided behaviors to overcome chemical defenses,
with examples including black-backed orioles, black-headed gros-
beaks, and some Australian raptors that dissect prey and consume
only body parts that contain the least toxins (Calvert etal. 1979; Fink
and Brower 1981; Beckmann and Shine 2011). Individual predators
of the same species can also vary in their ability or motivation to
discriminate between defended prey (Halpin etal. 2012), but the
reasons for such individual dierences are poorly understood.
In humans, individual dierences in taste perception, linked to ge-
netic polymorphisms in bitter taste receptors (Tas2rs - Chandrashekar
etal. 2000; Lindemann 2001; Behrens and Meyerhof 2013), aect
food choice and dietary habits (Garcia-Bailo etal. 2009; Lipchock
etal. 2017). There is now growing molecular evidence for the poten-
tial for variation in taste perception in birds, both between (Behrens
etal. 2014; Wang and Zhao 2015; Zhao etal. 2015) and within spe-
cies (Davis etal. 2010; Su etal. 2016). For example, Su etal. (2016)
detected 3–13 short nucleotide polymorphisms (SNPs) in the family
of G-protein-coupled receptors responsible for bitter taste perception
in Sichuan domestic and Tibetan chicken populations. Genetic vari-
ation in bitter taste is also reported in white-throated sparrows (Davis
etal. 2010). Whether individual dierences in bitter taste perception
influence the food choice of avian predators, however, has so far not
been tested experimentally.
Although taste may be important for discriminative learning,
predators’ decisions to consume chemically defended prey may also
be shaped by other factors, such as their physiological state (Sherratt
2003; Sandre etal. 2010; Skelhorn etal. 2016). Rather than avoiding
all chemically defended prey, predators are assumed to make state-
dependent decisions to include toxic prey in their diet when the
benefits of gaining nutrients outweigh the costs of ingesting toxins
(Skelhorn etal. 2016; Marples etal. 2018). For example, European
starlings consume more chemically defended prey when their body
mass and fat stores are reduced (Barnett etal. 2007; Barnett et al.
2012), or when their energetic needs are higher because of lower
ambient temperature (Chatelain etal. 2013). Willingness to consume
chemically defended prey may also depend on predators’ previous
consumption of toxins (Skelhorn and Rowe 2007) or the nutrient
content of the defended prey (Halpin etal. 2014), and it is possible
that these physiological factors have a greater influence than bitter
taste perception on predators’ foraging decisions. Therefore, our
aim was to test whether predators’ bitter taste perception influences
the consumption of chemically defended prey, or whether foraging
decisions are driven more by predators’ physiological measures.
We investigated bitter taste perception and avoidance learning
in wild-caught great tits (Parus major). Great tits are generalist pred-
ators and during the summer their diet consists mainly of insects
and other invertebrates (Naef-Daenzer etal. 2000), so they are likely
to encounter chemically defended prey (Majerus and Majerus 1997).
We first tested whether individuals diered in their response to
bitter-tasting chloroquine diphosphate solution. In the taste percep-
tion assays, birds were oered solutions with increasing concentra-
tions of chloroquine until they showed an aversive response to bitter
taste. The test was repeated on thefollowing day to investigate the
repeatability of individuals’ responses. We then investigated avoid-
ance learning by the same birds by presenting them with a random
sequence of novel palatable and defended prey until they refused to
attack defended prey. We predicted that individuals that displayed
aversion at lower concentrations (i.e., had lower taste response
thresholds) would consume fewer defended prey before rejecting
them. If so, variation in bitter taste perception might introduce
more heterogeneity in the predator population and create varying
predation pressures for aposematic prey, therefore influencing the se-
lection for prey defenses and signaling (Endler and Mappes 2004).
Study species
The experiment was carried out at the Konnevesi Research Station
in Central Finland during the winter of 2014–2015. Great tits
(n= 59; 19 juvenile and 8 adult females, and 16 juvenile and 16
adult males) were captured from a feeding site, kept in captivity
for approximately 1 week for testing, and then released back to the
capture site. Birds were housed individually in plywood cages (80×
65× 50cm) with a daily light period of 12.5 h. Sunflower seeds,
peanuts, tallow, and fresh water were provided ad libitum, except
prior and during the experiments when food or water restriction
was necessary to motivate foraging or drinking. We recorded sex
and age of the individuals based on their plumage, and measured
their tarsus length (0.01cm). Birds were also weighed (0.25g) both
after capture and before release. Before release, all individuals were
ringed for identification purposes.
Response to bittertaste
Bitter-tasting solutions were prepared by mixing chloroquine di-
phosphate salt with water to produce ten dierent concentrations
(mmol/L) 0.01; 0.05; 0.10; 0.50; 0.75; 1.0; 2.0; 3.0; 5.0; 8.0. Taste
assays were conducted in a 50× 66× 49 cm sized plywood cage
that had a front wall made of plexiglass, enabling us to observe and
film the birds during the assays. Birds were moved to the test cages
in the morning to start the taste assays. To increase their motivation
to drink, we moved birds before the automated light in their home
cages turned on (birds did not have light during the night) which en-
sured that they did not drink in their home cages before the assays.
Birds, therefore, did not have access to fresh water before they were
presented with the first test solution, and during the assays they only
had access to test solutions. Food was always freely available.
Birds were first presented with a white drinking bottle containing
fresh water (Figure 1a). After presenting a bottle, we waited for the
bird to drink and recorded its response with a video camera (Canon
Legria HF R37). If the bird did not drink in 15 min, we removed
the bottle for at least 15min (waiting for motivation to increase) be-
fore oering the same bottle again. This was repeated until the bird
drank from the bottle. We then removed the bottle and waited for
15min before presenting the bird with a bottle containing the lowest
concentration of chloroquine diphosphate solution. This procedure
continued, the bird being presented with solutions of increasing
concentration, until we observed a first aversive response, defined
as head shaking, bill-wiping, and/or spitting out of the solution
(Supplementary Video). Our criterion for the response threshold in-
cluded any of these responses, rather than a bird performing all three
behaviors. One of the main functions of bill-wiping is to clean the
beak (Cuthill etal. 1992), and some birds performed a couple of bill
wipes after drinking only fresh water. However, these responses were
much weaker than the aversive responses to the chloroquine solu-
tions. Other behaviors (head shaking and spitting out of the solution)
were never observed when birds were oered fresh water. The taste
assay was finished after birds showed the first aversive response.
We weighed each drinking bottle before and after presentation
to calculate the total amount of solution (mL) and chloroquine
(mmol) that birds consumed. This was important because, although
Page 2 of 10
Downloaded from by Jyvaskylan yliopisto / Kirjasto - kausijulkaisut user on 04 December 2019
Hämäläinen etal. • Predators’ consumption of unpalatable prey
all individuals were similarly water-deprived before the test, birds
diered in the amount of solution they drank from each bottle.
To investigate the repeatability of individual responses, 54 of 59 in-
dividuals were tested twice over two sequential days, with both tests
following the same protocol. In the beginning of the experiment,
we did not repeat the test with all individuals, therefore, five birds
were tested only once.
Avoidance of defendedprey
After the taste assays, we tested how many novel chemically defended
prey items each bird consumed before they rejected them. The avoid-
ance learning test was conducted in the same test cages as taste assays.
Prey items were chopped mealworm (Tenebrio molitor) pieces (approx-
imately 0.5 g). These were presented to birds in small plastic cups
(0.5mL) that contained either water (palatable prey) or 65mmol/L
chloroquine solution (defended prey). Our pilot tests indicated that
birds did not learn to avoid defended prey with lower doses of chlo-
roquine, possibly due to the low costs of consuming small quantities
of toxins. We, therefore, chose a 65 mmol/L concentration, even
though it was considerably higher than the taste response thresholds
in the taste assays (see Results). To make palatable and defended prey
visually distinct, we colored the solution by adding either green or
blue food dye, and placed the cup on Styrofoam cubes of the same
color (10cm3; Figure 1b). We counterbalanced the color associated
with unpalatability (blue palatable, green defended, n=29; reversed,
n= 27). Three birds (one juvenile female, and one juvenile and one
adult male) did not participate in the avoidance learning test, because
they refused to attack any prey (n=56).
Before the test birds were trained to eat mealworm pieces
from a cup that contained water and was presented on a white
Styrofoam cube. We then tested each bird’s initial color preference
by presenting them simultaneously with blue and green prey items
that were both palatable, and recording which prey birds attacked
first. Each bird was tested two times, alternating which prey item
was on the left and which on the right side of the cage. All birds,
therefore, had positive experience of both colors before the avoid-
ance learning test. We analyzed birds’ preferences using a gener-
alized linear mixed eects model (GLMM) with a binomial error
distribution (logit link). The model included the order in which prey
items were consumed as a response variable, prey color as an ex-
planatory variable and bird identity as a random eect. We found
that birds had a slight preference for green (estimate= −0.575±
0.270, Z=−2.131, P=0.03), with 59% of the individuals (33/56
birds) attacking the green prey first in the first preference test and
55% (31/56 birds) in the second preferencetest.
In the avoidance learning test, palatable and defended prey items
were presented to the birds sequentially in blocks of six (three palat-
able and three defended prey). Other food was restricted 90min prior
tothe test and blocks always started with a palatable prey item to en-
sure that the bird was motivated to forage. The following five prey
items were presented in a randomized order and birds were given
5min to attack each prey. If they attacked the prey within this time,
they were allowed to eat it before the next prey was presented. After
the bird had attacked all six prey items in the block, we paused for
10min before beginning the next block (next six prey items). Blocks
continued until thebirds refused to attack two consecutive defended
prey items (within 5min) but still continued to eat palatable prey pre-
sented immediately after a defended prey, indicating that they had
learned to recognize and avoid defended prey. However, if birds re-
fused to attack the palatable prey (within 5min), we concluded that
they had not learned to discriminate between prey items but instead
were not motivated to attack any prey. In this case, testing paused
for at least 10min before recommencing. To measure how fast birds
0 mmol/l
0.01 mmol/l 0.05 mmol/l
Experimental set-up. (a) We first conducted taste assays where birds (n = 59) were given water with increasing concentration of chloroquine diphosphate
until they showed an aversive response to bitter taste. (b) We then investigated how many defended prey birds (n=56) consumed before rejecting them by
presenting birds with novel palatable and defended prey. Prey items were mealworm pieces that were presented to birds in small plastic cups that were placed
on the colored cube.
Page 3 of 10
Downloaded from by Jyvaskylan yliopisto / Kirjasto - kausijulkaisut user on 04 December 2019
Behavioral Ecology
learned to discriminate between prey, we recorded the number of
defended prey attacked before birds reached the learning criterion.
Statistical analyses
We first calculated individuals’ body condition index to investigate
whether physiological state influenced the birds’ behavior in the
experiments. Body condition index was calculated based on tarsus
and weight measures, following the method described by Peig and
Green (2009). We used a weight measure that was taken after birds
were captured from the wild. However, calculating body condition
based on the weight at the release did not change any of the re-
sults. Because body condition had a dierent eect on avoidance
learning in juveniles and adults, and in males and females (see
Results), we also investigated whether body condition measures
diered between the age and sex groups. This was done using a
generalized linear model where body condition was modeled as a
function of birds’ age andsex.
We tested whether birds’ sex, age, or body condition influ-
enced their probability to respond to a chloroquine solution using
a GLMM with a binomial error distribution (using lme4 package;
Bates et al. 2015). Taste response thresholds (concentrations at
which birds showed the first aversive response) were first converted
to integers by multiplying them by 100, and then used as a bound
response variable, together with the probability of showing an aver-
sive response (i.e., 1). Explanatory variables in the model included
birds’ sex, age, body condition index and test day (first/second taste
assay), as well as bird identity as a random eect. We started model
selection with a model that included all possible two-way inter-
actions between sex, age, and body condition, and removed the
interaction terms based on their significance (see Supplementary
Material for model selection). Because consumption varied among
individuals, we also calculated the total amount of chloroquine
diphosphate consumed before the first aversive response, and con-
ducted the same analysis using this measure (mmol, converted to
integers) as a bound response variable, instead of the threshold
concentration. To investigate the consistency of birds’ responses
in the two taste assays, we calculated the repeatability in the taste
response thresholds between the assays by estimating which pro-
portion of the observed variance in the response thresholds was at-
tributed to bird identity. This was calculated from a GLMM with
a binomial error distribution, using the rptR package (Stoel etal.
2017). Bird identity was included as a random eect in themodel.
We next investigated how chloroquine concentration influenced
the volume of test solution consumed, and whether this depended
on the test day (first/second taste assay). We used a GLMM with a
negative binomial error distribution, as the data were right-skewed.
The volumes of solution consumed (mL) were converted to integers
and used as a response variable, and explanatory variables included
an interaction between chloroquine concentration (continuous var-
iable) and test day, and birds’ sex, age and body condition as fixed
eects, and bird identity as a random eect. Because most birds
responded to concentrations between 0.01 and 0.5mmol/L (Figure
2), we did not include higher concentrations in our analysis. We
further investigated whether consumption decreased when birds
were presented the first chloroquine solution using a paired sample
t-test where the consumption of fresh water (0 mmol/L) and the
first chloroquine solution (0.01mmol/L) were compared.
To investigate the relationship between taste perception and avoid-
ance learning, we used a generalized linear model with a Poisson error
distribution. The number of defended prey attacked before avoidance
was used as a response variable, and explanatory variables included
possible two-way interactions between birds’ taste response threshold
(in the first taste assay), sex, age, and body condition index. We re-
moved interaction terms from the model based on the significance of
Number of individuals
0.01 0.05 0.1 0.5 0.75
Taste response threshold to chloroquine (mmol/l)
1.0 3.0
Day 1
Day 2
5.0 8.0
Individual variation in bitter taste response thresholds (mmol/L) in the first (day 1; dark gray bars; n= 59) and second taste assay (day 2; light gray bars;
Page 4 of 10
Downloaded from by Jyvaskylan yliopisto / Kirjasto - kausijulkaisut user on 04 December 2019
Hämäläinen etal. • Predators’ consumption of unpalatable prey
the terms (see Supplementary Material for model selection). Because
birds showed an initial preference for green, the color that signaled
defended prey (green/blue) was included as a covariate in all models.
Finally, we investigated whether the amount of chloroquine consumed
1) during the first taste assay, or 2)in total during both taste assays
influenced the number of defended prey attacked before avoidance
by conducting the same analysis using the amount of chloroquine
consumed as an explanatory variable, instead of the taste response
threshold. Seven individuals (two juvenile and one adult females, and
four adult males) did not have a body condition measure (we did not
measure birds’ tarsus length in the beginning of the experiment) and
they were therefore excluded from the relevant analyses. In addition,
we excluded one bird that had a considerably higher body condition
index than others (body condition= 22.3), as this was likely due to
an error in tarsus measurement. All analyses were conducted using R
version 3.6.0 (R Core Team 2019).
Response to bittertaste
The concentration at which birds first showed an aversive re-
sponse varied considerably between individuals, ranging from
0.01 mmol/L to 8.0 mmol/L in the first taste assay, and from
0.01mmol/L to 0.75mmol/L in the second taste assay (Figure 2).
The repeatability in birds’ responses in the two taste assays was rel-
atively low: R=0.13 (95 % CI: 0.02–0.13). There was a significant
dierence between the two taste assays, with birds reacting to lower
concentrations during the second assay (Table 1). We found that ju-
veniles responded at lower concentrations than adults (adults versus
juveniles: estimate = 0.808 ± 0.367, Z = 2.200, P = 0.03), and
males at lower concentrations than females (females versus males:
estimate=0.714± 0.352, Z=2.030, P=0.04). However, these ef-
fects were driven by one adult female that reacted at a much higher
concentration than other birds (8 mmol/L); when this individual
was excluded from the analysis, the eects of sex and age were no
longer significant at alpha level 0.05 (Table 1). Body condition did
not influence birds’ response and this result did not change when
the outlier was excluded from the analysis (Table 1). Because birds
varied in the amount of solution consumed, we also tested whether
birds’ sex, age or body condition predicted the total amount of
chloroquine consumed before the first aversive response, but none
of these eects was significant (see Supplementary Material for the
full results). However, we found that birds responded to a smaller
amount of chloroquine in the second taste assay than the first (ef-
fect of test day: estimate=0.718± 0.227, Z=3.165, P = 0.002;
Supplementary Material).
Birds consumed less solution as the concentration of chloroquine
diphosphate increased (Table 2; Figure 3). This eect was the same
on both test days (concentration × test day: estimate= 0.158 ±
0.509, Z=0.310, P=0.76), and we therefore removed the interac-
tion between chloroquine concentration and test day from the final
model. The volume of solution consumed did not depend on birds’
age, sex, or body condition, but individuals consumed less oftheso-
lutions on the second day (Table 2), with the most noticeable de-
crease in the consumption of fresh water (0mmol/L; Figure 3). We
also found that in the first taste assay (day 1)the consumption of
test solution decreased significantly between the first (0 mmol/L)
and second (0.01 mmol/L) test solution (paired samples t-test:
t=6.833, df=56, P<0.001), whereas this dierence was not ob-
served on the second day (paired samples t-test: t=1.564, df=52,
Avoidance of defendedprey
Contrary to our prediction, birds’ bitter taste threshold did not
influence the number of defended prey that they attacked during
avoidance learning (Table 3). Similarly, the color that signaled
unpalatability (blue/green) did not influence avoidance learning
(Table 3). Instead, we found significant interactions between age
and body condition index (Figure 4a), and sex and body condition
index (Table 3; Figure 4b). Because there was no significant three-
way interaction between these variables (sex × age × body condi-
tion: estimate= 0.224± 0.167, Z =−1.343, P = 0.18), we next
investigated each interaction separately by conducting two models
that included only an interaction between age and body condition,
or sex and body condition.
We found that adult birds attacked more of the defended prey
when their body condition index was low (estimate=−0.274± 0.060,
Z= −4.545, P<0.001), whereas body condition did not have a sig-
nificant eect on the foraging choices of juveniles (estimate=0.088±
0.060, Z = 1.472, P= 0.14; Figure 4a). We also found that males
in poorer body condition attacked more ofthe defended prey (esti-
mate=−0.210 ± 0.056, Z = −3.743, P< 0.001), in contrast to fe-
males that did not attack defended prey as a function of their body
condition (estimate = 0.032 ± 0.058, Z = 0.550, P =0.58; Figure
4b). However, the number of birds in each age and sex category was
also not equal (adult females: n= 6; adult males: n=12; juvenile fe-
males: n = 15; juvenile males: n=15). Body condition measures did
not dier between the age (adults versus juveniles: estimate=0.284±
0.374, Z=0.761, P= 0.45) or sex categories (females versus males:
estimate=−0.076 ± 0.361, Z = −0.209, P= 0.84). A bird that re-
sponded only at the 8mmol/L concentration in the first taste assay
was excluded from the analysis (final n=48), as it was a clear outlier
Best-fit generalized linear mixed eects model explaining the variation in the taste response thresholds among individuals (n=50)
Terms in the model Estimate SE Z P
Intercept −3.100 2.551 −1.215 0.22
Sex (male) 0.543 0.338 1.606 0.11
Age (juvenile) 0.613 0.355 1.727 0.08
Body condition −0.029 0.138 −0.211 0.83
Test day (second taste assay) 0.469 0.230 2.042 0.04
The probability that birds responded to the test solution (within chloroquine concentration unit) was included as a response variable. Explanatory variables in
the best-fit model included birds’ age, sex, and body condition index, as well as test day (first/second taste assay). Bird identity was included as a random eect
(variance=0.708). Intercept gives the probability for adult females showing an aversive response in the first taste assay (day 1). One bird that responded at
much higher concentration than others (8mmol/L) was excluded from the analysis.
Page 5 of 10
Downloaded from by Jyvaskylan yliopisto / Kirjasto - kausijulkaisut user on 04 December 2019
Behavioral Ecology
and had a disproportionate eect on the results (see Supplementary
Material for the model including this individual). Finally, we con-
ducted similar models to test whether the amount of chloroquine con-
sumed during taste assays influenced avoidance learning, but found
no evidence that the amount of chloroquine consumed during the
first taste assay (estimate=55.364± 208.180, Z=0.266, P=0.79)
or in total during two taste assays (estimate = 20.847 ± 165.584,
Z=0.126, P = 0.90) influenced the consumption of defended prey
(see Supplementary Material for the full results).
Our study demonstrates that wild predators can dier greatly in
their response to bitter compounds that are structurally related to
real prey defenses. We found two orders of magnitude dierence
in the concentration at which great tits responded to chloroquine,
which is in line with the results of Warren and Vince (1963), who
also found individual variation in great tits’ responses to bitter-
tasting solutions. Contrary to our prediction, taste response did not
correlate with the number of defended prey items consumed in a
later avoidance learning test. Instead, we found that adult males
attacked more defended prey when their body condition index was
low. This supports the idea that individuals can strategically con-
sume defended prey when their energetic needs are higher (Sherratt
2003; Barnett etal. 2007, 2012; Skelhorn etal. 2016; Marples etal.
2018), although this may be age- and sex-dependent. It also indi-
cates that even though bitter taste perception may be important in
detecting toxic prey (Skelhorn and Rowe 2010), other factors, such
as physiological state, have a larger influence on predators’ foraging
Predators’ reactions to chemically defended prey can vary within
and between species (Exnerová etal. 2003; Endler and Mappes
2004). We found high individual variability in great tits’ responses
to bitter taste, with the taste response threshold varying from chlo-
roquine concentrations of 0.01 mmol/L to 8 mmol/L. Previous
studies in humans indicate that bitter taste sensitivity might decline
with age (Cowart et al. 1994; Fukunaga etal. 2005), but the ev-
idence of the eects of aging on taste in other animals is scarce
Volume consumed (ml)
0 0.01
Chloroquine concentration (mmol/l)
Day 1
Day 2
0.1 0.5
Mean (±SE) volume of test solutions that birds consumed at each chloroquine concentration in the first (day 1; dark gray bars) and second taste assay (day 2;
light gray bars). Because the test was finished when an individual responded to bitter taste, the number of observations at each concentration varies from 59
(0mmol/L; day 1)to 11 (0.5mmol/L; day 2).
Best-fit generalized linear mixed eects model explaining the volume of test solution that birds (n=51) consumed during the first
five chloroquine concentrations
Terms in the model Estimate SE Z P
Intercept 7.674 0.694 11.057 <0.001
Sex (male) 0.034 0.091 0.372 0.71
Age (juvenile) −0.019 0.094 −0.202 0.84
Body condition 0.022 0.037 0.574 0.57
Chloroquine concentration −2.087 0.247 −8.439 <0.001
Test day (second taste assay) −0.264 0.071 −3.703 <0.001
Explanatory variables in the best-fit model included birds’ age, sex, and body condition index, as well as chloroquine concentration and test day (first/second
taste assay). Bird identity was included as a random eect (variance=0.04). Intercept gives the estimate for the volume of the first test solution (0mmol/L) that
adult females consumed in the first taste assay (day 1).
Page 6 of 10
Downloaded from by Jyvaskylan yliopisto / Kirjasto - kausijulkaisut user on 04 December 2019
Hämäläinen etal. • Predators’ consumption of unpalatable prey
(but see Shin etal. 2012). We found some evidence that juveniles
responded to lower bitter taste concentrations than adults, but this
eect seemed to be driven by three adult birds that had high taste
response thresholds (3, 5, and 8mmol/L). It is possible that these
individuals were significantly older than other adult birds in our
study, which could explain their lower taste sensitivity, but we do
not have more accurate records of their age and cannot verify this.
We also found no evidence that sex or body condition influenced
taste responses, indicating that the observed dierences might be
genetically determined. In humans, polymorphism in bitter taste
receptor genes aects the ability to sense the intensity of bitterness,
as exemplified by the gene variants of TAS2R38 and sensitivity
to phenylthiocarbamide (PTC) and propylthiouracil (PROP; Bufe
etal. 2005). The varied perception of quinine intensity is also asso-
ciated with genetic variants in human TAS2r19 genes (Reed etal.
2010). Although genetic variants in bitter taste receptor genes has
been demonstrated in chickens (Su etal. 2016) these have not yet
been linked to taste behavior. While we did not genotype our indi-
vidual great tits, we hypothesize that similar variation is likely to be
found, though one or more bitter receptor or salivary proline-rich
protein genes, or receptor expression levels could be responsible for
the observed dierences (Lipchock etal. 2013), and this warrants
further investigation.
Determining taste sensitivity thresholds in animals is chal-
lenging (Rowland etal. 2015). Acommon method is to use two-
bottle choice tests in which individuals are given a choice of test
solution and pure water, and the consumption of the solutions is
compared (e.g., Warren and Vince 1963; Matson etal. 2004). In
our experiment, we investigated when birds showed a first behav-
ioral response to bitter-tasting solutions. Rather than testing taste
detection, our experiment, therefore, measured the threshold to
respond to bitter taste, and it is possible that individuals were able
to detect chloroquine at lower concentrations. In support of this
idea, we found that birds decreased their consumption behavior
significantly between 0 mmol/L and 0.01 mmol/L, indicating
that they might have already detected the lowest concentration
of chloroquine. However, it is also possible that birds were simply
thirstier in the beginning of the experiment. To disentangle
whether the decrease in consumption was due to birds detecting
chloroquine or being less thirsty, we would need a control group
that receives only water at each step. In our study, individuals also
varied in the volume of solution they drank, which might have
influenced their aversive responses, and further studies should
aim to control this by presenting birds with set amounts of chlo-
roquine solution. Furthermore, our results suggest that previous
experience may influence how birds respond to bitter taste. We
aimed to minimize any eects of learning during the assays, but
the lower taste response thresholds in the second taste assay in-
dicate that the birds did learn to associate drinking bottles with
some post-ingestive consequences of consuming chloroquine
during the first day. This is further supported by our finding of
the birds decreasing their consumption of fresh water from the
first to the second assay, which indicates that they were more hes-
itant to drink during the second day. The consumption of other
test solutions tended to similarly decrease between the assays, but
because their consumption was already low on the first day, this
decrease was less prominent compared to fresh water (Figure 3).
Variability in bitter taste sensitivity can influence food choice in
humans (Garcia-Bailo etal. 2009; Lipchock etal. 2017), but whether
it aects predators’ decisions to attack chemically defended prey has
until now not been tested. We did not find evidence that birds’ bitter
taste response threshold influenced the consumption of defended
prey during the avoidance learning test. Instead, birds’ physiolog-
ical state seemed to aect their foraging behavior, with adult males
attacking more defended prey when their body condition was poor.
This is consistent with previous work showing that starlings are more
likely to attack defended prey when their body mass and fat stores
are experimentally reduced (Barnett etal. 2007, 2012), supporting
the idea that educated predators attack toxic prey depending on
nutrient-toxin trade-os (Skelhorn etal. 2016). Indeed, our exper-
iment might have measured birds’ willingness to consume defended
prey, rather than how well they learned to discriminate the prey
items, as birds only had a choice to attack or reject defended prey
with no alternative prey present. Giving birds a simultaneous choice
between palatable and defended prey, or comparing hesitation times
to attack each prey type might, therefore, provide better estimates of
discriminative learning. Studies in a more complex foraging environ-
ment with palatable and defended prey (e.g., in the “novel world,”
Alatalo and Mappes 1996) would also allow us to investigate the ef-
fects of variation in predator taste sensitivity on the mortality of de-
fended prey, and how this influences the evolution of prey defenses.
For example, prey might evolve higher levels of chemical defense
and reduced visual conspicuousness when the predator community
consists of less sensitive predators, although this is likely to depend
on the costs to prey of producing chemical defenses (Longson and
Joss 2006). Similar to other intra- and interspecific variation among
predators (Endler and Mappes 2004), variation in taste sensitivity
might, therefore, have important consequences for prey, but further
studies are required to understand its eects on prey rejection and
avoidance learning.
Best-fit generalized linear model explaining the number of defended prey that birds (n=48) attacked during avoidance learning
Terms in the model Estimate SE Z P
Intercept 4.991 1.368 3.648 <0.001
Taste threshold 0.078 0.101 0.774 0.44
Sex (male) 4.318 1.445 2.989 0.003
Age (juvenile) −6.602 1.566 −4.215 <0.001
Body condition −0.151 0.074 −2.025 0.04
Defended prey color (green) 0.060 0.099 0.602 0.55
Body condition * Age (juvenile) 0.348 0.085 4.086 <0.001
Body condition * Sex (male) −0.226 0.078 −2.885 0.004
Explanatory variables in the best-fit model included interactions between age and body condition index, and sex and body condition index, as well as taste
response threshold in the first taste assay (mmol/l) and the color of defended prey (green/blue). Intercept gives the estimate for the number of defended prey
that adult females attacked when defended prey were blue. One bird that responded at much higher concentration than others (8mmol/L) was excluded from
the analysis.
Page 7 of 10
Downloaded from by Jyvaskylan yliopisto / Kirjasto - kausijulkaisut user on 04 December 2019
Behavioral Ecology
Disentangling the eects of taste and toxicity on predators’
foraging decisions is often dicult. Stronger taste might not be al-
ways linked to a higher toxin concentration (Ruxton and Kennedy
2006; Holen 2013; Nissim etal. 2017; Marples etal. 2018) or vice
versa (Lawrence etal. 2019), and predator species may vary in what
they find unpalatable (Rojas et al. 2017). Our study indicates that
after detecting bitter taste, birds make their foraging decisions based
on the post-ingestive feedback of consuming toxins. This distinction
between distastefulness and unprofitability was recently highlighted
by Marples etal. (2018) who suggested that these two prey qualities
should be treated as separate phenomena, with predators avoiding
unprofitable (but not only distasteful) prey. Our results support the
idea that prey toxicity is important for predator learning (Brower
1969; Skelhorn and Rowe 2010), and defense based solely on dis-
tastefulness may not protect prey (but see Skelhorn and Rowe
2009). To investigate the eects of taste and toxicity separately, fur-
ther studies should aim to manipulate predators’ ability to taste. An
antagonist for chicken taste receptors was recently described (Dey
etal. 2017), making taste perception manipulation possible, similar
to studies in which birds’ sense of smell is blocked to test how olfac-
tory cues influence their navigation (Gagliardo etal. 2013). Further
work is also needed to investigate the eects of taste on learning
about weaker or more variable prey defenses (Ihalainen et al.
2007, 2008). In our experiment, the concentration of chloroquine
Number of defended prey attacked
16 17 18
Body condition index
19 20 21
16 17 18
Body condition index
19 20 21
Number of defended prey attacked
Number of defended prey that birds (n=48) attacked during avoidance learning. (a) Adults (open circles and the dashed line) attacked more defended prey
when their body condition was low, whereas there was no eect of body condition in juveniles (filled circles and the solid line). (b) Males (open triangles and
the dashed line) attacked more defended prey when their body condition was low, whereas there was no eect of body condition in females (filled triangles
and the solid line).
Page 8 of 10
Downloaded from by Jyvaskylan yliopisto / Kirjasto - kausijulkaisut user on 04 December 2019
Hämäläinen etal. • Predators’ consumption of unpalatable prey
diphosphate solution was very high to ensure that birds acquired
avoidance to defended prey. Even though high doses of quinine
can be emetic to birds (Alcock 1970), we do not know how costly
it is to ingest at lower concentrations, and our pilot tests suggested
that birds did not avoid prey with weaker defense. However, the use
of high chloroquine concentration in the avoidance learning test
means that even less sensitive birds were likely to detect prey unpal-
atability easily, which could explain why we did not find a correla-
tion between the taste response threshold and learning.
In conclusion, our study suggests that great tits dier in their
responses to bitter taste, but this does not influence the number
of defended prey that they are willing to consume. This indicates
that other factors, such as visual cues, might be more important
than taste for influencing predators’ initial decision to consume
prey (Marples et al. 1994; Ihalainen et al. 2007), whereas prey
toxicity and its physiological eects might drive later foraging be-
havior (Skelhorn et al. 2016). Nevertheless, the ability to detect
bitter taste might be important when predators are sampling prey
with weaker chemical defenses or when defenses are more var-
iable, which can increase the risk of ingesting toxins (Skelhorn
and Rowe 2005; Barnett etal. 2014). Furthermore, predators can
gather social information about prey unpalatability by observing
the disgust responses of other predators (Mason and Reidinger
1982; Johnston et al. 1998; Skelhorn 2011; Thorogood et al.
2018; Hämäläinen etal. 2019a). Our study indicates that indi-
viduals vary in how likely they are to show these responses, which
could create heterogeneity in social information that is available
for observing predators. For example, more sensitive individuals
might have a larger role in providing information for others, and
this could influence how social information spreads in the pred-
ator population. We did not quantify the strength of aversive
response in our study, but this is also likely to vary among indi-
viduals, with some birds performing more beak wiping and head
shaking than others. Stronger responses might be a more salient
signal of unpalatability for the observing individuals (Skelhorn
2011), but how the strength of the aversive response influences
social avoidance learning remains untested.
Supplementary material can be found at Behavioral Ecology online.
This work was supported by the Academy of Finland (grant number
284666). L.H. was supported by the Finnish Cultural Foundation and Emil
Aaltonen Foundation. R.T. is supported by an Independent Research
Fellowship from the Natural Environment Research Council UK (grant
number NE/K00929X/1) and a start-up grant from the Helsinki Institute
of Life Science (HiLIFE), University of Helsinki. H.M.R. was supported by
a research fellowship from the Institute of Zoology, Zoological Society of
London, and is currently supported by the Max Planck Society.
We have no conflict of interest to declare.
We are grateful to Helinä Nisu for taking care of the birds, the sta at
Konnevesi Research Station for providing facilities for the experiment,
Victoria Franks for providing illustrations, and Robert Burriss for his
comments on the manuscript. We also thank two anonymous referees for
helpful comments. Wild birds were used with permission from the Central
Finland Centre for Economic Development, Transport and Environment
and license from the National Animal Experiment Board (ESAVI/9114/
04.10.07/2014) and the Central Finland Regional Environmental Centre
Data accessibility: Analyses reported in this article can be reproduced using
the data provided by Hämäläinen etal. (2019b).
Handling editor: Marc Naguib
Alatalo RV, Mappes J. 1996. Tracking the evolution of warning signals.
Nature. 382:708–710.
AlcockJ. 1970. Punishment levels and the response of black-capped chicka-
dees (Parus atricapillus) to three kinds of artificial seeds. Anim Behav.
BarnettCA, BatesonM, RoweC. 2007. State-dependent decision making:
educated predators strategically trade o the costs and benefits of con-
suming aposematic prey. Behav Ecol. 18:645–651.
Barnett CA, Bateson M, Rowe C. 2014. Better the devil you know:
avian predators find variation in prey toxicity aversive. Biol Lett.
Barnett CA, Skelhorn J, Bateson M, Rowe C. 2012. Educated predators
make strategic decisions to eat defended prey according to their toxin
content. Behav Ecol. 23:418–424.
BatesD, MächlerM, BolkerB, WalkerS. 2015. Fitting linear mixed-eects
models using lme4. J Stat Softw. 67:1–48.
BeckmannC, ShineR. 2011. Toad’s tongue for breakfast: exploitation of a
novel prey type, the invasive cane toad, by scavenging raptors in tropical
Australia. Biol Invasions. 13:1447–1455.
BehrensM, KorschingSI, MeyerhofW. 2014. Tuning properties of avian
and frog bitter taste receptors dynamically fit gene repertoire sizes. Mol
Biol Evol. 31:3216–3227.
BehrensM, MeyerhofW. 2013. Bitter taste receptor research comes of age:
from characterization to modulation of TAS2Rs. Semin Cell Dev Biol.
Blum MS. 1981. Chemical defenses of arthropods. New York: Academic
BrodieED III, BrodieED Jr. 1990. Tetrodotoxin resistance in garter snakes:
an evolutionary response of predators to dangerous prey. Evolution.
BrowerLP. 1969. Ecological chemistry. Sci Am. 220:22–29.
BrowerLP, Fink LS. 1985. A natural toxic defense system: cardenolides in
butterflies versus birds. Ann N Y Acad Sci. 443:171–188.
Bufe B, Breslin PA, Kuhn C, Reed DR, Tharp CD, Slack JP, KimUK,
DraynaD, MeyerhofW. 2005. The molecular basis of individual dier-
ences in phenylthiocarbamide and propylthiouracil bitterness perception.
Curr Biol. 15:322–327.
CalvertWH, HedrickLE, BrowerLP. 1979. Mortality of the monarch but-
terfly (Danaus plexippus L.): avian predation at five overwintering sites in
Mexico. Science. 204:847–851.
Chandrashekar J, Mueller KL, Hoon MA, Adler E, Feng L, Guo W,
ZukerCS, RybaNJ. 2000. T2Rs function as bitter taste receptors. Cell.
ChatelainM, Halpin CG, RoweC. 2013. Ambient temperature influences
birds’ decisions to eat toxic prey. Anim Behav. 86:733–740.
Cowart BJ, Yokomukai Y, Beauchamp GK. 1994. Bitter taste in aging:
compound-specific decline in sensitivity. Physiol Behav. 56:1237–1241.
Cuthill I, Witter M, Clarke L. 1992. The function of bill-wiping. Anim
Behav. 43:103–115.
DavisJK, LowmanJJ, ThomasPJ, tenHallersBF, KoriabineM, HuynhLY,
Maney DL, de Jong PJ, Martin CL, Thomas JW; NISC Comparative
Sequencing Program. 2010. Evolution of a bitter taste receptor gene
cluster in a New World sparrow. Genome Biol Evol. 2:358–370.
DeyB, KawabataF, KawabataY, YoshidaY, NishimuraS, TabataS. 2017.
Identification of functional bitter taste receptors and their antagonist in
chickens. Biochem Biophys Res Commun. 482:693–699.
EndlerJA, MappesJ. 2004. Predator mixes and the conspicuousness of ap-
osematic signals. Am Nat. 163:532–547.
ExnerováA, Landová E, Štys P, Fuchs R, ProkopováM, Cehláriková P.
2003. Reactions of passerine birds to aposematic and non-aposematic
firebugs (Pyrrhocoris apterus; Heteroptera). Biol J Linn Soc. 78:517–525.
Page 9 of 10
Downloaded from by Jyvaskylan yliopisto / Kirjasto - kausijulkaisut user on 04 December 2019
Behavioral Ecology
FinkLS, BrowerLP. 1981. Birds can overcome the cardenolide defence of
monarch butterflies in Mexico. Nature. 291:67–70.
Fink LS, Brower LP, WaideRB, Spitzer PR. 1983. Overwintering mon-
arch butterflies as food for insectivorous birds in Mexico. Biotropica.
FukunagaA, UematsuH, Sugimoto K. 2005. Influences of aging on taste
perception and oral somatic sensation. J Gerontol ABiol Sci Med Sci.
Gagliardo A, Bried J, Lambardi P, LuschiP, Wikelski M, Bonadonna F.
2013. Oceanic navigation in Cory’s shearwaters: evidence for a cru-
cial role of olfactory cues for homing after displacement. J Exp Biol.
Garcia-BailoB, ToguriC, EnyKM, El-SohemyA. 2009. Genetic variation
in taste and its influence on food selection. OMICS. 13:69–80.
Glendinning JI. 1994. Is the bitter rejection response always adaptive?
Physiol Behav. 56:1217–1227.
Glendinning JI, Brower LP, Montgomery CA. 1990. Responses of
three mouse species to deterrent chemicals in the monarch butterfly.
I. Taste and toxicity tests using artificial diets laced with digitoxin or
monocrotaline. Chemoecology. 3–4:114–123.
HalpinCG, SkelhornJ, RoweC. 2012. The relationship between sympatric
defended species depends upon predators’ discriminatory behaviour.
PLoS One. 7:e44895.
Halpin CG, Skelhorn J, RoweC. 2014. Increased predation of nutrient-
enriched aposematic prey. Proc Biol Sci. 281:20133255.
Hämäläinen L, Mappes J, Rowland HM, Thorogood R. 2019a.
Social information use about novel aposematic prey is not influ-
enced by a predator’s previous experience with toxins. Funct Ecol.
Hämäläinen L, Mappes J, Thorogood R, Valkonen JK, Karttunen K,
Salmi T, Rowland HM. 2019b. Data from: predators’ consumption of
unpalatable prey does not vary as a function of bitter taste perception.
Dryad Digital Repository.
HolenØH. 2013. Disentangling taste and toxicity in aposematic prey. Proc
Biol Sci. 280:20122588.
Ihalainen E, Lindström L, Mappes J. 2007. Investigating Müllerian mim-
icry: predator learning and variation in prey defences. J Evol Biol.
IhalainenE, LindströmL, Mappes J, PuolakainenS. 2008. Butterfly eects
in mimicry? Unpredictable relationship between Müllerian co-mimics.
Behav Ecol Sociobiol. 62:1267–1276.
JohnstonANB, BurneTHJ, RoseSPR. 1998. Observation learning in day-
old chicks using a one-trial passive avoidance learning paradigm. Anim
Behav. 56:1347–1353.
LawrenceJP, RojasB, FouquetA, Mappes J, BlanchetteA, Saporito RA,
Bosque RJ, Courtois EA, Noonan BP. 2019. Weak warning sig-
nals can persist in the absence of gene flow. Proc Natl Acad Sci USA.
Lindemann B. 2001. Receptors and transduction in taste. Nature.
LipchockSV, Mennella JA, Spielman AI, Reed DR. 2013. Human bitter
perception correlates with bitter receptor messenger RNA expression in
taste cells. Am J Clin Nutr. 98:1136–1143.
Lipchock SV, Spielman AI, Mennella JA, Mansfield CJ, Hwang LD,
DouglasJE, ReedDR. 2017. Caeine bitterness is related to daily caf-
feine intake and bitter receptor mRNA abundance in human taste tissue.
Perception. 46:245–256.
Longson CG, JossJMP. 2006. Optimal toxicity in animals: predicting the
optimal level of chemical defences. Funct Ecol. 20:731–735.
MajerusMEN, MajerusTMO. 1997. Predation of ladybirds by birds in the
wild. Entomol Mon Mag. 133:55–61.
Marples NM, Speed MP, Thomas RJ. 2018. An individual-based profita-
bility spectrum for understanding interactions between predators and
their prey. Biol J Linn Soc. 125:1–13.
Marples NM, van Veelen W, Brakefield PM. 1994. The relative impor-
tance of colour, taste and smell in the protection of an aposematic insect
Coccinella septempunctata. Anim Behav. 48:967–974.
MasonJR, ReidingerRF. 1982. Observational learning of food aversions in
red-winged blackbirds (Agelaius phoeniceus). Auk. 99:548–554.
MatsonKD, MillamJR, KlasingKC. 2004. Cockatiels (Nymphicus hollandicus)
reject very low levels of plant secondary compounds. Appl Anim Behav
Sci. 85:141–156.
Naef-Daenzer L, Naef-Daenzer B, Nager RG. 2000. Prey selection and
foraging performance of breeding Great Tits Parus major in relation to
food availability. J Avian Biol. 31:206–214.
NissimI, Dagan-WienerA, NivMY. 2017. The taste of toxicity: a quanti-
tative analysis of bitter and toxic molecules. IUBMB Life. 69:938–946.
Peig J, Green AJ. 2009. New perspectives for estimating body condition
from mass/length data: the scaled mass index as an alternative method.
Oikos. 118:1883–1891.
PoultonEB. 1890. The colours of animals: their meaning and use especially
considered in the case of insects. London: Kegan Paul, Trench, Trübner
& Co.
R Core Team. 2019. R: a language and environment for statistical com-
puting. Vienna (Austria): R Foundation for Statistical Computing.
Reed DR, Zhu G, Breslin PA, Duke FF, Henders AK, Campbell MJ,
MontgomeryGW, MedlandSE, MartinNG, WrightMJ. 2010. The percep-
tion of quinine taste intensity is associated with common genetic variants in a
bitter receptor cluster on chromosome 12. Hum Mol Genet. 19:4278–4285.
RojasB, Burdfield-SteelE, PakkanenH, Suisto K, Maczka M, Schulz S,
MappesJ. 2017. How to fight multiple enemies: target-specific chemical
defences in an aposematic moth. Proc Biol Sci. 284:20171424.
Rowland HM, Parker MR, Jiang P, Reed DR, Beauchamp GK. 2015.
Comparative taste biology with special focus on birds and reptiles. In:
DotyRL, editor. Handbook of olfaction and gustation New York: Wiley.
RuxtonGD, Kennedy MW. 2006. Peppers and poisons: the evolutionary
ecology of bad taste. J Anim Ecol. 75:1224–1226.
RuxtonGD, SherrattTN, SpeedMP. 2018. Avoiding attack. Evolutionary
ecology of crypsis, warning signals and mimicry. 2nd ed. Oxford: Oxford
University Press.
Sandre SL, Stevens M, Mappes J. 2010. The eect of predator appetite,
prey warning coloration and luminance on predator foraging decisions.
Behaviour. 147:1121–1143.
Sherratt TN. 2003. State-dependent risk-taking by predators in systems
with defended prey. Oikos. 103:93–100.
ShinYK, CongWN, CaiH, KimW, MaudsleyS, EganJM, MartinB. 2012.
Age-related changes in mouse taste bud morphology, hormone expression,
and taste responsivity. J Gerontol ABiol Sci Med Sci. 67:336–344.
SkelhornJ. 2011. Colour biases are a question of conspecifics taste. Anim
Behav. 81:825–829.
Skelhorn J, Halpin CG, RoweC. 2016. Learning about aposematic prey.
Behav Ecol. 27:955–964.
Skelhorn J, Rowe C. 2005. Tasting the dierence: do multiple defence
chemicals interact in Müllerian mimicry? Proc Biol Sci. 272:339–345.
Skelhorn J, Rowe C. 2006. Taste-rejection by predators and the evolution
of unpalatability in prey. Behav Ecol Sociobiol. 60:550–555.
SkelhornJ, RoweC. 2007. Predators’ toxin burdens influence their strategic
decisions to eat toxic prey. Curr Biol. 17:1479–1483.
Skelhorn J, Rowe C. 2009. Distastefulness as an antipredator defence
strategy. Anim Behav. 78:761–766.
SkelhornJ, Rowe C. 2010. Birds learn to use distastefulness as a signal of
toxicity. Proc Biol Sci. 277:1729–1734.
StoelMA, NakagawaS, SchielzethH. 2017. rptR: repeatability estimation
and variance decomposition by generalized linear mixed-eects models.
Methods Ecol Evol. 8:1639–1644.
SuY, LiD, GaurU, WangY, WuN, ChenB, XuZ, YinH, HuY, ZhuQ.
2016. Genetic diversity of bitter taste receptor gene family in Sichuan
domestic and Tibetan chicken populations. J Genet. 95:675–681.
Thorogood R, Kokko H, Mappes J. 2018. Social transmission of avoid-
ance among predators facilitates the spread of novel prey. Nat Ecol Evol.
WangK, ZhaoH. 2015. Birds generally carry a small repertoire of bitter
taste receptor genes. Genome Biol Evol. 7:2705–2715.
WarrenRP, Vince MA. 1963. Taste discrimination in the great tit (Parus
major). J Comp Physiol Psychol. 56:910–913.
ZhaoH, LiJ, ZhangJ. 2015. Molecular evidence for the loss of three basic
tastes in penguins. Curr Biol. 25:R141–R142.
Page 10 of 10
Downloaded from by Jyvaskylan yliopisto / Kirjasto - kausijulkaisut user on 04 December 2019
... Many species have evolved chemical defenses and aposematic prey advertise these to predators with conspicuous warning signals (Poulton 1890). Birds are common predators of many aposematic species, and often perform beak wiping and head shaking when tasting an unpalatable prey (Clark 1970;Rowland et al. 2015;Hämäläinen et al. 2020a). These visible displays ("distaste responses") can provide observers information about prey unprofitability, and several studies have now demonstrated that avian predators can learn to avoid unpalatable prey by observing negative foraging experiences of both conspecifics (Mason and Reidinger 1982;Johnston et al. 1998; Thorogood et al. 2018;Hämäläinen et al. 2019) and heterospecifics (Mason et al. 1984;Hämäläinen et al. 2020bHämäläinen et al. , 2021a. ...
... For example, predators might not observe the whole predation event and therefore see only part of the response to unpalatable prey (Hämäläinen et al. 2021b). There is also likely to be among-and within-species variation in bitter taste sensitivity, with some species or individuals responding to lower concentrations of toxins and showing stronger aversive reactions than others (Rowland et al. 2015;Hämäläinen et al. 2020a). Investigating how the intensity of observed cues influences predators' foraging decisions is important if we want to understand how predation pressures act on different prey types. ...
... It is possible that there is genetic variation in bitter taste sensitivity, as demonstrated in some avian species (Davis et al. 2010;Su et al. 2016). Behavioral responses might also depend on an individual's age, sex, or body condition, although there was no evidence of this in a similar experiment with great tits (Hämäläinen et al. 2020a), and here we found no effect of age on beak wiping behavior of blue tits. Furthermore, we do not know how consistent these individual differences are, and whether similar range of variation in responses is observed when predators are foraging in the wild. ...
Animals gather social information by observing the behavior of others, but how the intensity of observed cues influences decision-making is rarely investigated. This is crucial for understanding how social information influences ecological and evolutionary dynamics. For example, observing a predator’s distaste of unpalatable prey can reduce predation by naïve birds, and help explain the evolution and maintenance of aposematic warning signals. However, previous studies have only used demonstrators that responded vigorously, showing intense beak-wiping after tasting prey. Therefore, here we conducted an experiment with blue tits (Cyanistes caeruleus) informed by variation in predator responses. First, we found that the response to unpalatable food varies greatly, with only few individuals performing intensive beak-wiping. We then tested how the intensity of beak-wiping influences observers’ foraging choices using video-playback of a conspecific tasting a novel conspicuous prey item. Observers were provided social information from 1) no distaste response, 2) a weak distaste response, or 3) a strong distaste response, and were then allowed to forage on evolutionarily novel (artificial) prey. Consistent with previous studies, we found that birds consumed fewer aposematic prey after seeing a strong distaste response, however, a weak response did not influence foraging choices. Our results suggest that while beak-wiping is a salient cue, its information content may vary with cue intensity. Furthermore, the number of potential demonstrators in the predator population might be lower than previously thought, although determining how this influences social transmission of avoidance in the wild will require uncovering the effects of intermediate cue salience.
... There are a number of reasons birds might decide to consume toxic prey, even after they have been warned about it (Barnett et al., 2007(Barnett et al., , 2012. For example, Hämäläinen et al. (2020) found that great tits differ in taste perception, but that their decision to eat toxic prey depended on the bird's body condition, and not taste perception. Similarly, we found that bird body weight (as a proxy for condition), impacted behaviors across the predation sequence including dropping, handling, killing and eating the moth. ...
Full-text available
Aposematic organisms warn predators of their unprofitability using a combination of defenses, including visual warning signals, startling sounds, noxious odors, or aversive tastes. Using multiple lines of defense can help prey avoid predators by stimulating multiple senses and/or by acting at different stages of predation. We tested the efficacy of three lines of defense (color, smell, taste) during the predation sequence of aposematic wood tiger moths ( Arctia plantaginis ) using blue tit ( Cyanistes caeruleus ) predators. Moths with two hindwing phenotypes (genotypes: WW/Wy = white, yy = yellow) were manipulated to have defense fluid with aversive smell (methoxypyrazines), body tissues with aversive taste (pyrrolizidine alkaloids) or both. In early predation stages, moth color and smell had additive effects on bird approach latency and dropping the prey, with the strongest effect for moths of the white morph with defense fluids. Pyrrolizidine alkaloid sequestration was detrimental in early attack stages, suggesting a trade-off between pyrrolizidine alkaloid sequestration and investment in other defenses. In addition, pyrrolizidine alkaloid taste alone did not deter bird predators. Birds could only effectively discriminate toxic moths from non-toxic moths when neck fluids containing methoxypyrazines were present, at which point they abandoned attack at the consumption stage. As a result, moths of the white morph with an aversive methoxypyrazine smell and moths in the treatment with both chemical defenses had the greatest chance of survival. We suggest that methoxypyrazines act as context setting signals for warning colors and as attention alerting or “go-slow” signals for distasteful toxins, thereby mediating the relationship between warning signal and toxicity. Furthermore, we found that moths that were heterozygous for hindwing coloration had more effective defense fluids compared to other genotypes in terms of delaying approach and reducing the latency to drop the moth, suggesting a genetic link between coloration and defense that could help to explain the color polymorphism. Conclusively, these results indicate that color, smell, and taste constitute a multimodal warning signal that impedes predator attack and improves prey survival. This work highlights the importance of understanding the separate roles of color, smell and taste through the predation sequence and also within-species variation in chemical defenses.
... Progress in understanding the ecology of chemical defences as they relate to prey sampling and survival has been more gradual (though unabating, e.g. [88][89][90]), and some fundamental questions remain. The extent to which Daphnia spp. ...
Full-text available
The combined use of noxious chemical defences and conspicuous warning colours is a ubiquitous anti-predator strategy. That such signals advertise the presence of defences is inherent to their function, but their predicted potential for quantitative honesty-the positive scaling of signal salience with the strength of protection-is the subject of enduring debate. Here, we systematically synthesized the available evidence to test this prediction using meta-analysis. We found evidence for a positive correlation between warning colour expression and the extent of chemical defences across taxa. Notably, this relationship held at all scales; among individuals, populations and species, though substantial between-study heterogeneity remains unexplained. Consideration of the design of signals revealed that all visual features, from colour to contrast, were equally informative of the extent of prey defence. Our results affirm a central prediction of honesty-based models of signal function and narrow the scope of possible mechanisms shaping the evolution of aposematism. They suggest diverse pathways to the encoding and exchange of information, while highlighting the need for deeper knowledge of the ecology of chemical defences to enrich our understanding of this widespread anti-predator adaptation.
... Therefore, the prey used in this study had two levels of defense in this experiment (defended and undefended). We arbitrarily chose to use the term 'chemically defended' over 'unpalatable' because a prey's palatability is known to change in response to an individual bird's energy reserves and other factors (Barnett et al. 2007;Skelhorn and Rowe 2007;Chatelain et al. 2013;Vesely et al. 2017;Hämäläinen et al. 2019). Moreover, what may be palatable to one species may be unpalatable to another species (Marples et al. 2018) meaning that the term 'unpalatable' is condition and species-dependent. ...
Full-text available
Birds are important predators of insects and insects often incorporate chemical defenses that may make themselves distasteful or toxic to the predators. Predators can respond to chemically defended prey in multiple ways, the predator psychology approach to predation often treats predation as a general process despite the possibility for multiple responses among species. The effectiveness of a prey’s chemical defense at reducing predation might also vary depending on what predator is attacking the prey. Here, we compared the attack strategies of three different species of avian predators (Japanese bush warblers [Horornis diphone], narcissus flycatchers [Ficedula narcissina], and Japanese tits [Parus minor]) which are found in the temperate forests of Japan. We found overall, that undefended prey was preferred over the defended prey, but the different predator species had different preferences and handled prey differently from one another. This suggests that different predator species might exert different selection pressures on chemically defended prey and this adds to our growing appreciation that predator behavior can vary among predator species. Moreover, our findings emphasize the importance of understanding differences in behavior among free-living predator species in studies of aposematism and mimicry.
... Interestingly, we found that great tits consumed more unpalatable almonds when they were lighter, irrespective of which colour indicated unpalatability. This supports the idea that birds are more willing to consume chemically defended prey when they are in a poorer physiological condition (Barnett et al. 2007(Barnett et al. , 2012Skelhorn et al. 2016;Hämäläinen et al. 2020a). In contrast, we found the opposite effect in blue tits, with heavier individuals tending to attack more unpalatable food items than lighter birds. ...
Full-text available
Colours are commonly used as visual cues when measuring animals’ cognitive abilities. However, animals can have innate biases towards certain colours that depend on ecological and evolutionary contexts, therefore potentially influencing their performance in experiments. For example, when foraging, the colour red can advertise profitable fruits or act as a warning signal about chemically defended prey, and an individual’s propensity to take food of that colour may depend on experience, age or physical condition. Here, we investigate how these contexts influence blue tits’ (Cyanistes caeruleus) and great tits’ (Parus major) responses to red-coloured almond flakes. We found that juvenile birds preferred red both when it was presented simultaneously with green, and when it was presented with three alternative colours (orange, purple, green). Adult birds, however, only preferred red after a positive experience with the colour, or when it was presented with the three alternative colours. We then tested whether colour influenced avoidance learning about food unpalatability. Despite the prediction that red is a more salient warning signal than green, we found only weak evidence that birds discriminated red unpalatable almonds from a green palatable alternative more quickly than when the colours were reversed. Our results suggest that biases towards red food may depend on birds’ age and previous experience, and this might influence their performance in experiments that use red stimuli. Considering the ecological relevance of colours is, therefore, important when designing experiments that involve colour cues.
Full-text available
Urbanization is rapidly changing ecological niches. On the inhabited Galapagos Islands, Darwin's finches consume human-introduced foods preferentially; however, it remains unclear why. Here we presented pastry with flavour profiles typical of human foods (oily, salty, sweet) to small and medium ground finches to test if latent taste preferences might drive selection of human foods. If human-food flavours were consumed more than a neutral or bitter control only at sites with human foods, then we predicted tastes were acquired after urbanisation; however, if no site-differences were found then this would indicate latent taste preferences. Contrary to both predictions, we found no evidence that human-food flavours were preferred compared to control flavours. Instead, medium ground finches consumed the bitter control pastry most and wiped their beaks more frequently after feeding on oily and sweet pastry (post-ingestion beak-wiping can indicate aversions). Small ground finches showed no differences in consumption but wiped their beaks most after feeding on sweet pastry. We found no evidence that medium and small ground finches found bitter-tasting food aversive. Furthermore, we found no evidence that taste preferences could have played a major role in Darwin's finches responding to the presence of human foods during increased urbanization.
In a recent publication, Wisenden et al. (2020) examined responses of territorial male Red-winged Blackbirds (Agelaius phoeniceus) to models constructed with ultraviolet (UV)-reflective red epaulets for the purpose of determining if the addition of UV reflectance to epaulets ("UV+") changed the effectiveness of signals to receivers relative to "control" epaulets under field conditions. The authors hypothesized that "UV+epaulet coloration represents a visual signal with increased efficacy in territorial interactions." They presented behavioral data but no visual modeling data. Our aims in this commentary are to suggest alternative terms to those used by the authors, to express concern about the use of sunscreen to manipulate the UV condition of surfaces, and to make a plea for additional data collection in future studies of avian visual ecology. The terms UV+ and UV- should be reserved for studies that create environments free from UV radiation for comparison with environments that include UV radiation. We believe that commercial sunscreens are not an appropriate choice for altering the UV conditions of surfaces presented during behavioral trials because they potentially introduce confounding influences from other sensory inputs or irritation of peripheral nerves. Wisenden et al. altered the UV absorbance of their sunscreen-treated models but did not present absorbance spectra and may not have collected those data. We acknowledge that the lack of absorbance spectra is not unusual. We implore any such future studies to collect absorbance spectra of treated and control surfaces so that those data may be used to improve visual models for UV-sensitive animals. Los ec´ologos visuales de aves deben considerer la absorbencia UV y todas las modalidades sensoriales: respuesta a Wisenden et al. (2020) RESUMEN (Spanish)—En una publicaci´on reciente, Wisenden et al. (2020) examinaron las respuestas de machos territoriales del tordo Agelaius phoeniceus a modelos construidos con charreteras rojas reflejantes ultravioletas (UV) con el prop´osito de determinar si la adici´on de reflectancia a las charreteras (‘‘UVþ’’) cambiaba la efectividad de las se˜nales dirigidas a receptores en relaci´on con charreteras ‘‘control’’ bajo condiciones de campo. Los autores ten´ıan la hip´otesis de que ‘‘la coloraci´on UVþde las charreteras representa una se˜nal visual con eficacia aumentada en interacciones territoriales’’. Los autores presentaron datos conductuales pero no datos de modelado visual. Nuestra meta en este comentario es sugerir t´erminos alternativos a aquellos empleados por los autores, expresar preocupaci´on por el uso de bloqueador solar para manipular la condici´on UV de superficies y hacer un llamado para la colecta de datos adicionales en futuros estudios de ecolog´ıa visual aviar. Los t´erminos UVþy UV– deben estar reservados para comparaciones con ambientes que incluyan radiaci´on UV. Pensamos que el uso de bloqueadores solares comerciales no es una elecci´on adecuada para la alteraci´on de las condiciones UV de superficies durante pruebas de comportamiento porque podr´ıan introducir influencias confusas provenientes de otras se˜nales sensoriales o irritar nervios perif´ericos. Wisenden et al. alteraron la absorbencia UV de sus modelos tratados con bloqueador solar pero no presentaron espectros de absorbencia y podr´ıan no haber colectado esos datos. Reconocemos que la carencia de espectros de absorbencia no es inusual. Pedimos que cualquier estudio futuro colecte espectros de absorbencia en superficies tratadas y controles para que esos datos puedan ser usados para mejorar los modelos visuales de animales sensibles a UV.
Full-text available
Chemical defences against predators underlie the evolution of aposematic coloration and mimicry, which are classic examples of adaptive evolution. Surprisingly little is known about the roles of ecological and evolutionary processes maintaining defence variation, and how they may feedback to shape the evolutionary dynamics of species. Cyanogenic Heliconius butterflies exhibit diverse warning color patterns and mimicry, thus providing a useful framework for investigating these questions. We studied intraspecific variation in de novo biosynthesized cyanogenic toxicity and its potential ecological and evolutionary sources in wild populations of Heliconius erato along environmental gradients, in common-garden broods and with feeding treatments. Our results demonstrate substantial intraspecific variation, including detectable variation among broods reared in a common garden. The latter estimate suggests considerable evolutionary potential in this trait, although predicting the response to selection is likely complicated due to the observed skewed distribution of toxicity values and the signatures of maternal contributions to the inheritance of toxicity. Larval diet contributed little to toxicity variation. Furthermore, toxicity profiles were similar along steep rainfall and altitudinal gradients, providing little evidence for these factors explaining variation in biosynthesized toxicity in natural populations. In contrast, there were striking differences in the chemical profiles of H. erato from geographically distant populations, implying potential local adaptation in the acquisition mechanisms and levels of defensive compounds. The results highlight the extensive variation and potential for adaptive evolution in defense traits for aposematic and mimetic species, which may contribute to the high diversity often found in these systems.
Full-text available
Aposematic organisms couple conspicuous warning signals with a secondary defense to deter predators from attacking. Novel signals of aposematic prey are expected to be selected against due to positive frequency-dependent selection. How, then, can novel phenotypes persist after they arise, and why do so many aposematic species exhibit intrapopulation signal variability? Using a polytypic poison frog (Dendrobates tinctorius), we explored the forces of selection on variable aposematic signals using 2 phenotypically distinct (white, yellow) populations. Contrary to expectations, local phenotype was not always better protected compared to novel phenotypes in either population; in the white population, the novel phenotype evoked greater avoidance in natural predators. Despite having a lower quantity of alkaloids, the skin extracts from yellow frogs provoked higher aversive reactions by birds than white frogs in the laboratory, although both populations differed from controls. Similarly, predators learned to avoid the yellow signal faster than the white signal, and generalized their learned avoidance of yellow but not white. We propose that signals that are easily learned and broadly generalized can protect rare, novel signals, and weak warning signals (i.e., signals with poor efficacy and/or poor defense) can persist when gene flow among populations, as in this case, is limited. This provides a mechanism for the persistence of intrapopulation aposematic variation, a likely precursor to polytypism and driver of speciation. aposematism | frequency-dependent selection | polymorphism | unpalatability | secondary defenses
Full-text available
1.Aposematism is an effective antipredator strategy. However, the initial evolution and maintenance of aposematism is paradoxical because conspicuous prey are vulnerable to attack by naïve predators. Consequently, the evolution of aposematic signal mimicry is also difficult to explain. 2.The cost of conspicuousness can be reduced if predators learn about novel aposematic prey by observing another predator's response to that same prey. On the other hand, observing positive foraging events might also inform predators about the presence of undefended mimics, accelerating predation on both mimics and their defended models. 3.It is currently unknown, however, how personal and social information combine to affect the fitness of aposematic prey. For example, does social information become more useful when predators have already ingested toxins and need to minimise further consumption? 4.We investigated how toxin load influences great tits' (Parus major) likelihood to use social information about novel aposematic prey, and how it alters predation risk for undefended mimics. Birds were first provided with mealworms injected with bitter‐tasting chloroquine (or a water‐injected control), before information about a novel unpalatable prey phenotype was provided via video playback (either prey alone, or of a great tit tasting the noxious prey). 5.An experimentally‐increased toxin load made great tits warier to attack prey, but only if they lacked social information about unpalatable prey. Socially educated birds consumed fewer aposematic prey relative to a cryptic phenotype, regardless of toxin load. In contrast, after personally experiencing aposematic prey, birds ignored social information about palatable mimics and were hesitant to sample them. 6.Our results suggest that social information use by predators could be a powerful force in facilitating the evolution of aposematism as it reduces predation pressure on aposematic prey, regardless of a predator's toxin load. Nevertheless, observing foraging events of others is unlikely to alter frequency‐dependent dynamics among models and mimics, although this may depend on predators having recent personal experience of the model's unpalatability. This article is protected by copyright. All rights reserved.
Full-text available
There is confusion in the animal behaviour literature over the use of the terms 'toxicity' and 'unpalatability', which are commonly used interchangeably when describing the function of chemical compounds in prey, although these terms describe very different functions. Toxic chemicals cause fitness-reducing harm, whereas unpalatability provides aversive taste but no reduction in fitness. Furthermore, chemical defences are only one aspect of prey profitability. We argue that if predators are maximizing fitness, all prey can be described in terms of their costs and benefits to predators across all currencies, giving each prey item a positive or negative position on a 'profitability spectrum'. Adaptively foraging predators should be selected to eat only prey with a positive profitability. The context of each predator-prey encounter also alters the profitability of the prey. Given that profitability is a function of the current state of both the predator and the prey individuals, we explain why it should be considered to be an attribute of a particular encounter, in contrast to its present usage as an attribute of a prey species. This individual-centred perspective requires researchers to investigate, through both theoretical models and empirical studies, the complex conditions in which predators and prey meet in real life.
Full-text available
Warning signals are an effective defence strategy for aposematic prey, but only if they are recognized by potential predators. If predators must eat prey to associate novel warning signals with unpalatability, how can aposematic prey ever evolve? Using experiments with great tits (Parus major) as predators, we show that social transmission enhances the acquisition of avoidance by a predator population. Observing another predator's disgust towards tasting one novel conspicuous prey item led to fewer aposematic than cryptic prey being eaten for the predator population to learn. Despite reduced personal encounters with unpalatable prey, avoidance persisted and increased over subsequent trials. Next we use a mathematical model to show that social transmission can shift the evolutionary trajectory of prey populations from fixation of crypsis to fixation of aposematism more easily than was previously thought. Therefore, social information use by predators has the potential to have evolutionary consequences across ecological communities.
Full-text available
The role of bitter taste-one of the few basic taste modalities-is commonly assumed to signal toxicity and alert animals against consuming harmful compounds. However, it is known that some toxic compounds are not bitter and that many bitter compounds have negligible toxicity while having important health benefits. Here we apply a quantitative analysis of the chemical space to shed light on the bitterness-toxicity relationship. Using the BitterDB dataset of bitter molecules, The BitterPredict prediction tool, and datasets of toxic compounds, we quantify the identity and similarity between bitter and toxic compounds. About 60% of the bitter compounds have documented toxicity and only 56% of the toxic compounds are known or predicted to be bitter. The LD50 value distributions suggest that most of the bitter compounds are not very toxic, but there is a somewhat higher chance of toxicity for known bitter compounds compared to known nonbitter ones. Flavonoids and alpha acids are more common in the bitter dataset compared with the toxic dataset. In contrast, alkaloids are more common in the toxic datasets compared to the bitter dataset. Interestingly, no trend linking LD50 values with the number of activated bitter taste receptors (TAS2Rs) subtypes is apparent in the currently available data. This is in accord with the newly discovered expression of TAS2Rs in several extra-oral tissues, in which they might be activated by yet unknown endogenous ligands and play non-gustatory physiological roles. These results suggest that bitter taste is not a very reliable marker for toxicity, and is likely to have other physiological roles. © 2017 IUBMB Life, 2017.
Full-text available
Animals have evolved different defensive strategies to survive predation, among which chemical defences are particularly widespread and diverse. Here we investigate the function of chemical defence diversity, hypothesizing that such diversity has evolved as a response to multiple enemies. The aposematic wood tiger moth (Arctia plantaginis) displays conspicuous hindwing coloration and secretes distinct defensive fluids from its thoracic glands and abdomen.We presented the two defensive fluids from laboratoryreared moths to two biologically relevant predators, birds and ants, and measured their reaction in controlled bioassays (no information on colour was provided). We found that defensive fluids are target-specific: thoracic fluids, and particularly 2-sec-butyl-3-methoxypyrazine, which they contain, deterred birds, but caused no aversive response in ants. By contrast, abdominal fluids were particularly deterrent to ants, while birds did not find them repellent. Our study, to our knowledge, is the first to show evidence of a single species producing separate chemical defences targeted to different predator types, highlighting the importance of taking into account complex predator communities in studies on the evolution of prey defence diversity.
Full-text available
Intra-class correlations (ICC) and repeatabilities (R) are fundamental statistics for quantifying the reproducibility of measurements and for understanding the structure of biological variation. Linear mixed effects models offer a versatile framework for estimating ICC and R. However, while point estimation and significance testing by likelihood ratio tests is straightforward, the quantification of uncertainty is not as easily achieved. A further complication arises when the analysis is conducted on data with non-Gaussian distributions, because the separation of the mean and the variance is less clear-cut for non-Gaussian than for Gaussian models. Nonetheless, there are solutions to approximate repeatability for the most widely used families of generalized linear mixed models (GLMMs). Here we introduce the R package rptR for the estimation of ICC and R for Gaussian, binomial and Poisson-distributed data. Uncertainty in estimators is quantified by parametric bootstrapping and significance testing is implemented by likelihood ratio tests and through permutation of residuals. The package allows control for fixed effects and thus the estimation of adjusted repeatabilities (that remove fixed effect variance from the estimate) and enhanced agreement repeatabilities (that add fixed effect variance to the denominator). Furthermore, repeatability can be estimated from random-slope models. The package features convenient summary and plotting functions. Besides repeatabilities, the package also allows the quantification of coefficients of determination R2 as well as of raw variance components. We present an example analysis to demonstrate the core features and discuss some of the limitations of rptR. This article is protected by copyright. All rights reserved.
Avoiding Attack discusses the diversity of mechanisms by which prey avoid predator attacks and explores how such defensive mechanisms have evolved through natural selection. It considers how potential prey avoid detection, how they make themselves unprofitable to attack, how they communicate this status, and how other species have exploited these signals. Using carefully selected examples of camouflage, mimicry, and warning signals drawn from a wide range of species and ecosystems, the authors summarise the latest research into these fascinating adaptations, developing mathematical models where appropriate and making recommendations for future study. This second edition has been extensively rewritten, particularly in the application of modern genetic research techniques which have transformed our recent understanding of adaptations in evolutionary genomics and phylogenetics. The book also employs a more integrated and systematic approach, ensuring that each chapter has a broader focus on the evolutionary and ecological consequences of anti-predator adaptation. The field has grown and developed considerably over the last decade with an explosion of new research literature, making this new edition timely.
The use of the "arms race" analogy as a conceptualization of evolutionary predator-prey interactions has been criticized because of the lack of evidence that predators can and do adapt to increased antipredator ability of prey. We present evidence that the garter snake Thamnophis sirtalis has evolved resistance to tetrodotoxin (TTX) in response to the toxicity of the newt Taricha granulosa on which the snake feeds. A bioassay (locomotor performance before and after injection of TTX) was used to obtain repeated measures of resistance for individual snakes. We studied interpopulation and interspecific variation by comparing resistance in Thamnophis sirtalis from populations occurring sympatrically and allopatrically with Taricha granulosa, and in Thamnophis ordinoides (which does not feed on the newt) occurring sympatrically with Taricha granulosa. We also examined intrapopulation variation in TTX resistance using snakes from a population known to feed on Taricha granulosa. Resistance differed significantly among individuals and litters; repeatability and heritability estimates of the assay were significantly different from zero, demonstrating the potential for response to selection. The population of Thamnophis sirtalis that occurs with Taricha granulosa exhibited levels of resistance much greater than either of the other groups. These results suggest that the predator-prey arms race analogy may be applicable to this system.