ArticlePDF Available

Distinguishing discrete polymorphism from continuous variation in throat colour of tree lizards, Urosaurus ornatus

Authors:

Abstract and Figures

Within population colour variation in animals is ubiquitous and can range from discrete polymorphism to continuous variation. Distinguishing discrete polymorphism from continuous variation can be challenging, and this hinders our ability to test hypotheses about colour variation. We tested whether throat colour variation in the ornate tree lizard [Urosaurus ornatus (U. ornatus); Baird & Girard, 1852) is discrete or continuous using photographs of 617 individuals from seven study sites in the Chiricahua Mountains of Arizona, USA. Using principal component analyses on ten colour variables derived from photographs, we found throat colour to be highly repeatable for both males and females. Cluster analyses suggested that there are different groups of individuals based on their throat colours in males and in females, but distinguishing between the groups was difficult due to significant overlap in colour. Therefore, it appears that there is a significant amount of continuous variation in both male and female tree lizard throat colours. We suggest quantifying trait variation before assuming a colour trait is discrete. By using numerical descriptors of colour, more information is retained than by using discrete groups. Quantifying individual variation in colour is important for linking colour with other traits such as reproductive strategy, immune function, and size and for testing hypotheses about the evolution and maintenance of colour polymorphism.
Content may be subject to copyright.
© 2017 The Linnean Society of London, Biological Journal of the Linnean Society, 2017, 121, 72–81 72
Biological Journal of the Linnean Society, 2017, 121, 72–81. With 3 figures.
Distinguishing discrete polymorphism from continuous
variation in throat colour of tree lizards, Urosaurus
ornatus
JAMES E. PATERSON* and GABRIEL BLOUIN-DEMERS
Department of Biology, University of Ottawa, Ottawa K1N 6N5, Canada
Received 17 August 2016; revised 6 October 2016; accepted for publication 1 November 2016
Within population colour variation in animals is ubiquitous and can range from discrete polymorphism to continu-
ous variation. Distinguishing discrete polymorphism from continuous variation can be challenging, and this hinders
our ability to test hypotheses about colour variation. We tested whether throat colour variation in the ornate tree
lizard [Urosaurus ornatus (U. ornatus); Baird & Girard, 1852) is discrete or continuous using photographs of 617
individuals from seven study sites in the Chiricahua Mountains of Arizona, USA. Using principal component analy-
ses on ten colour variables derived from photographs, we found throat colour to be highly repeatable for both males
and females. Cluster analyses suggested that there are different groups of individuals based on their throat colours
in males and in females, but distinguishing between the groups was difficult due to significant overlap in colour.
Therefore, it appears that there is a significant amount of continuous variation in both male and female tree lizard
throat colours. We suggest quantifying trait variation before assuming a colour trait is discrete. By using numerical
descriptors of colour, more information is retained than by using discrete groups. Quantifying individual variation in
colour is important for linking colour with other traits such as reproductive strategy, immune function, and size and
for testing hypotheses about the evolution and maintenance of colour polymorphism.
ADDITIONAL KEYWORDS: alternative strategies – colour morphs – photograph – reptile.
INTRODUCTION
Within population colour variation is widespread in
animals (Thompson & Moore, 1991; Hill et al., 1994;
Sinervo & Lively, 1996; Blouin-Demers et al., 2013)
and ranges from discrete polymorphism to continu-
ous variation. Discrete colour morphs with correlated
traits, such as behaviour, reproductive tactics, and
body size, have been reported in a wide variety of taxa
(Knapton & Falls, 1983; Sinervo & Lively, 1996; Knapp
et al., 2003; McKinnon & Pierotti, 2010). For exam-
ple, the Gouldian finch (Erythrura gouldiae) exhibits
two genetically determined head colour morphs (red
and black) that differ in aggression and in immune
response because the morphs differ in their endocrine
responses and in their sensitivity to conspecific den-
sity (Pryke et al., 2007). Other colour traits vary con-
tinuously between individuals of a population (Hill
& Montgomerie, 1994; Salvador et al., 1996; Weiss,
2006). For instance, female striped plateau lizards
(Sceloporus virgatus) display significant variation in
the size and intensity of their orange throat patch,
and patch size and colour are correlated with clutch
size and parasite load (Weiss, 2006). The cause and
evolutionary maintenance of colour variation between
individuals of a population have long been of inter-
est to ecologists and evolutionary biologists (Sinervo,
Bleay & Adamopoulou, 2001; Roulin, 2004; Gray &
McKinnon, 2007).
Interindividual colour variation can be under genetic
control, environmental control, or a combination of
both (Gray & McKinnon, 2007; Hubbard et al., 2010).
Discrete colour polymorphism is the presence of at
least two genetically determined colour morphs within
a population (Ford, 1945). Multiple colour morphs are a
source of genetic diversity in populations, and can have
several ecological consequences, including decreased
intraspecific competition (Smith & Skulason, 1996),
increased abundance, and increased resilience to
extinction (Forsman et al., 2008). Continuous colour
*Corresponding author. E-mail: james.earle.paterson@gmail.com
DISTINGUISHING DISCRETE POLYMORPHISM 73
© 2017 The Linnean Society of London, Biological Journal of the Linnean Society, 2017, 121, 72–81
variation between individuals can be under genetic
control. For example, plumage colour in tawny owls
(Strix aluco) is highly heritable, ranges from grey to
reddish brown, and has adaptive value (Brommer,
Ahola & Karstinen, 2005). Continuous colour variation
between individuals can also be under environmental
control, in which case colour could indicate between-
individual differences that are correlated with fitness
(Wilson & Nussey, 2010). For example, male house
finch (Haemorhous mexicanus) plumage brightness
signals nutritional status because brightness is a func-
tion of diet-derived carotenoids deposited in feathers
(Hill & Montgomerie, 1994; Hill et al., 1994). In sev-
eral other species, colour patch size and intensity are
correlated with body condition, clutch size, and immu-
nocompetence (Weiss, 2006). Colour patches may thus
provide honest signals of individual quality (McGraw
& Ardia, 2003).
For many species that display variation in colour, it
is unclear whether the colour trait is discrete or con-
tinuous. This is a serious shortcoming because test-
ing hypotheses about how other phenotypic traits are
related to colour, and about how colour variation is
maintained evolutionarily, requires knowing whether
the trait is constrained to a finite number of unordered
values (discrete) or whether it is continuous (Roulin,
2004; Cote et al., 2008). In instances where colour is a
true discrete polymorphism, an individual’s genotype
can be inferred from its phenotype (Sinervo, Svensson
& Comendant, 2000; Sinervo et al., 2001; Corl et al.,
2010). In instances where colour is a continuous trait
and in instances where colour is under environmental
control, however, an individual’s genotype cannot be
inferred from its phenotype (Cote et al., 2008). Several
mechanisms, such as frequency-dependent selection,
assortative mating, or disassortative mating, can
explain the evolutionary maintenance of discrete col-
our polymorphism (Gray & McKinnon, 2007). If indi-
viduals cannot be assigned reliably to a genotype based
on their phenotype, then these alternate mechanisms
cannot be tested without genetic data. The mainte-
nance of continuous colour variation often depends on
sexual selection or on variation in resources and the
traits related to resource acquisition (Salvador et al.,
1996). Testing whether colour variation is discrete
or continuous can inform whether individuals can be
accurately assigned a genotype based on their pheno-
type, and can help identify possible mechanisms that
maintain colour variation. A more complete under-
standing of the genes associated with colour polymor-
phism requires data from breeding experiments to
measure heritability and genome wide associations.
There are methodological challenges associated with
quantifying colour variation and with distinguishing
discrete from continuous variation. There are cases
where colour polymorphism was deemed discrete and
where phenotype was used to infer genotype (Vercken
et al., 2007; Vercken, Sinervo & Clobert, 2008) that
were later deemed to represent continuous colour vari-
ation (Cote et al., 2008). Colour is usually measured in
one of three ways: qualitative scoring by researchers
(Carpenter, 1995; Hews et al., 1997; Corl et al., 2010),
spectrophotometry (Cogliati, Corkum & Doucet, 2010;
Lattanzio, Metro & Miles, 2014), or analysis of photo-
graphs (Stevens et al., 2007; Teasdale, Stevens & Stuart-
Fox, 2013). Qualitative scoring is useful when there is
a marked discrete polymorphism easily detected with
the human visual range. However, this method is unre-
liable if there are subtle differences between colours or
if there are signals outside the human visual range (for
instance, flowers that reflect ultraviolet light; Chittka
et al., 1994). Spectrophotometry allows colours to be
measured quantitatively and within the spectrum of
wavelengths that the receivers perceive. One of the
criticisms of using spectrophotometry, however, is that
fine-scale transitions between patches of different col-
ours could be missed (Vercken et al., 2008); analyzing
high-quality photographs can overcome this difficulty
by examining differences in colour for each pixel. If
the colour trait of interest reflects wavelengths only
within the human visual range, then analyzing photo-
graphs proves to be an effective method for measuring
colour variation.
Lizards are ideal to quantify colour variation
because their mating and social systems often involve
visual communication (Fleishman et al., 1997), and
many species display marked intra- and inter-popu-
lation variation in colour (Hews et al., 1997; Teasdale
et al., 2013). However, lizards may vary continuously
in wavelength reflected, or they may occur in discrete
morphs as a result of different alleles in a single
gene or in a group of tightly linked genes (Sinervo
& Lively, 1996). The ornate tree lizard, U. ornatus,
displays a variety of throat colours that are used in
visual signaling between competitors and between
potential mates (Fig 1). Both males and females
display throat colour variation, which has been
assigned to discrete morphs (Zucker & Boecklen,
1990; Thompson & Moore, 1991; Lattanzio & Miles,
2014), but colour variation has not been quantified
to test this assertion formally. Spectrophotometry
indicated that throat colour patches of U. ornatus do
not reflect significant amounts of light in the ultra-
violet range (Lattanzio et al., 2014). Our objectives
were to develop a method for quantifying individual
variation in lizard throat colour using photographs,
to determine whether colour variation is repeatable,
and to test whether tree lizard throat colour varia-
tion is discrete (as assumed by previous authors) or
continuous.
74 J. E. PATERSON and G. BLOUIN-DEMERS
© 2017 The Linnean Society of London, Biological Journal of the Linnean Society, 2017, 121, 72–81
METHODS
PhotograPhs
U. ornatus (317 males and 300 females) were captured
by noose from seven study sites in the Chiricahua
Mountains in southeastern Arizona, USA from May to
August in 2014 and 2015. The study sites were in simi-
lar habitats along riparian corridors and were sepa-
rated by a maximum of 16 km. All work was approved
by the University of Ottawa Animal Care Committee
(protocol BL-2300) and was conducted with scientific
collecting permits from the Arizona Fish and Game
Department (SP674341 and SP713940). We used
a medical cauterizer to give lizards a unique mark
(Ekner et al., 2011), and we photographed their throat
colour patch in the field with a light background at
ambient temperature and lighting conditions (without
a flash) with a Canon Rebel EOS T3 DSLR camera and
a 18–55 mm lens. The lighting and temperature condi-
tions represent the range in which lizards are active
and engage in intraspecific signaling. Based on visual
inspection in the field, male lizards were classified
into discrete throat colour morphs: blue, green, orange,
orange-blue, or orange-green (Fig. 1AE). Morphs with
two colours had an outer perimeter of orange with a
central patch of either blue or green. Female lizards
were classified as: orange, orange-yellow, white, or yel-
low. Orange-yellow females had an outer perimeter of
orange with a central white patch (Fig. 1FI).
Linearizing and equaLizing coLour channeLs
To use photographs for the analysis of colour, a digital
camera’s three sensors (corresponding to red, green,
and blue wavelengths) should respond linearly and
equally to increases in light intensity (Stevens et al.,
2007). We tested this by photographing colour cards
with grey standards of varying reflectance (X-Rite
ColorChecker Passport) in the field under the same
lighting conditions as the lizard photographs. We con-
structed linearization equations (Stevens et al., 2007;
Bergman & Beehner, 2008; White et al., 2015) based
on 102 measurements of grey colour cards ranging in
reflectance from 3 to 90%:
Qab
r
r
=
()
11
* (1)
Qab
g
g
=
()
22
* (2)
Qab
b
b
=
()
33
* (3)
Figure 1. Urosaurus ornatus colour morphs from the
Chiricahua Mountains of Arizona, USA categorized into (A)
blue, (B) green, (C) orange, (D) orange-blue, and (E) orange-
green for males, and (F) orange, (G) orange-yellow, (H) yel-
low, and (I) white for females.
Figure 2. Schematic diagram of steps to extract the ten
quantitative variables used to analyze photographs of
Urosaurus ornatus throat colour from the Chiricahua
Mountains of Arizona, USA.
DISTINGUISHING DISCRETE POLYMORPHISM 75
© 2017 The Linnean Society of London, Biological Journal of the Linnean Society, 2017, 121, 72–81
Where Q is the known reflectance value from the grey
colour standard, a and b are constants (for each cam-
era sensor), and r, g, and b are the pixel scores from the
camera’s three sensors (red, green, and blue). Grey col-
our standards are designed to reflect all wavelengths
equally. The RGB colour scores were extracted from
photographs using the software ImageJ (Schneider,
Rasband & Eliceiri, 2012). Using these data, we fit
equations with the nls function in R (R Core Team,
2015) to estimate a and b for each colour sensor. We
confirmed that linearized pixel scores (Q) of grey col-
our standards were equal in each sensor using paired
t-tests. All photographs of lizards were then linearized
and equalized using eqn. [1–3] before further analyses.
segmentation and quantifying coLour variation
To quantify individual colour variation, we measured
the proportion of the throat of each lizard covered by
each colour and the intensity of colour for two zones of
the patch (Fig. 2). To obtain the proportion of different
colours in lizard throats, we converted linearized pho-
tographs into segmented black-and-white photographs
(Teasdale et al., 2013) that represent the colours dis-
played by our U. ornatus populations. The distinct col-
ours represented in our populations were: blue, green,
orange, yellow, and white. To determine threshold lim-
its for these colours, we used pixel scores from 25 to
35 measurements of each colour from linearized photo-
graphs of lizard throats. Instead of the raw linearized
pixel scores from each sensor, we used the propor-
tion of that sensor’s value relative to the other sen-
sors (Teasdale et al., 2013) to standardize for lighting
intensity [e.g., proportion blue: pb = b/(r + g + b) for
each pixel]. We created a classification tree using the
R package rpart (Therneau et al., 2015) to classify pix-
els into each colour using these proportions. The parti-
tioning was highly accurate at separating pixel colours
(classification error = 2.7%). The splits from this par-
titioning analysis were used to segment each lizard
photo into four black-and-white photos that repre-
sented blue, green, orange, and yellow pixels. The pro-
portion of white was not measured and was assumed
to comprise the remainder of the throat not assigned
to any of the other colours. In each segmented photo,
all pixels were assigned 1 or 0. For example, in the blue
segmented photo all blue pixels were 1, and all others
were 0. Finally, the proportion of a lizard’s throat com-
posed of each colour was determined using the soft-
ware ImageJ (Schneider et al., 2012) by counting the
number of pixels that were 1 for each segmented pho-
tograph and by dividing by the total number of pixels.
We also measured the colour intensity of throat
patches in two zones of 0.25 cm2 for every lizard. The
mean pixel scores of each colour channel (Qr, Qb and
Qg) on linearized photos were measured using ImageJ
at the center of the throat and at the posterior left
side of the throat. These two zones captured variation
in colour expression between individuals. Therefore,
for each lizard photograph, we had ten variables to
describe throat colour: four proportions to describe
the relative size of each colour, three intensity scores
from the center of the throat patch, and three intensity
scores from the periphery of the throat patch.
PrinciPaL comPonent anaLysis of coLour
variation
To derive a composite measure of colour variation, we
conducted a principal component analysis (PCA) on
the variables describing lizard throat colour varia-
tion. For males, we used all ten variables. For females,
we only used eight variables because no individuals
displayed significant blue (maximum 6% of throat) or
green (maximum 11% of throat). Variables were scaled
to have a mean of 0 and an SD of one to account for
uneven variance prior to conducting the PCA.
rePeatabiLity of coLour measurements
To assess repeatability of colour measurements, we
used a subset of photographs of lizards that were
recaptured between 2 weeks and 1 year apart. We
analyzed males (n = 77 lizards) and females (n = 71
lizards) separately because they display different col-
ours. We assessed repeatability of the throat patch col-
our variables with intraclass correlation coefficients
(ICC) that compared within individual variation to
between individual variation (Lessells & Boag, 1987).
Since we had ten colour variables for each photo, we
calculated ICCs using the first two axes of the PCA.
To test whether the assignment of individuals to
morphs was repeatable, we also assigned individuals
to morphs using photographs, unaware of their initial
morph assignment in the field. We compared whether
field and photograph morph assignments were con-
sistent and whether individuals were assigned to the
same morph in the field on their two capture occasions.
is coLour exPression simiLar between sites?
We ensured lizard throat colour variation was similar
at each of the seven study sites. We used ANOVAs with
PC1 and PC2 as dependent variables and colour morph,
site, and the interaction between colour morph and
site as independent variables. If throat colour expres-
sion differed between sites, we predicted there would
be a significant effect of site or an interaction between
site and colour morph on PC1 or PC2. We conducted
separate analyses for males and females because they
differed in the range of colours expressed. For the anal-
ysis of males, we excluded orange individuals (n = 5),
76 J. E. PATERSON and G. BLOUIN-DEMERS
© 2017 The Linnean Society of London, Biological Journal of the Linnean Society, 2017, 121, 72–81
green individuals (n = 21), and one study site (n = 8)
because there was inadequate replication to estimate
coefficients. For females, we excluded one study site
(n = 4) because there was inadequate replication to
estimate coefficients.
are the coLour morPhs discrete?
We used linear discriminant function analyses (DFA)
to determine whether the morphs identified in the
field could be distinguished based on quantitative
throat colour variables. Separate DFAs were con-
structed for males (ten colour variables) and females
(eight colour variables) because the morphs assigned
in the field differed by sex. In addition, we also per-
formed support vector machine (SVM) classification
using tuned parameters for cost and gamma (e1071
package; Meyer et al., 2014) to verify whether indi-
viduals could be classified into colour groups. We
again conducted separate analyses for males and
for females. We computed correct classification rates
for DFA and SVM as a measure of how well morphs
identified in the field could be distinguished based
on quantitative throat colour variables. Both SVM
and DFA maximize the distances between group cen-
troids based on a priori groupings, but identifying the
various morphs a priori is subjective. Therefore, we
also conducted cluster analyses on the numerical col-
our variables to identify morphs objectively a poste-
riori. We used the Mclust function (Fraley & Raftery,
2002) to create clusters using normal mixture mod-
els and the Bayesian Information Criterion (BIC) to
determine the optimal number of clusters and their
structure with the default models of variance and
covariance (Fraley & Raftery, 2002). The model with
the highest BIC was considered the most supported.
We again conducted separate cluster analyses for
males and females.
RESULTS
PrinciPaL comPonent anaLysis of coLour
variation
For males, the first two components of the PCA
explained 74% of the variance in throat colour
(Fig. 3A). The first component (PC1) had heavy vari-
able loadings for the proportion of the throat that was
blue, and for the pixel scores for the blue sensor (Qb) in
the center and at the periphery of the throat (Table 1).
Therefore, males with more blue in their throats had
higher PC1 scores. The second component (PC2) had
high variable loadings for the proportion of the throat
that was orange, pixel scores for the red sensor (Qr)
in the center and at the periphery of the throat, and
negative loadings for the proportion of the throat that
was green and the pixel scores for the green sensor
(Qg). Therefore, males with low PC2 scores had more
green throats, and males with high PC2 values had
more orange throats (Fig. 3A).
For females, the first two components of the PCA
explained 83% of the variance in throat colour. The
variables with the heaviest loadings for PC1 were
proportion of the throat that was orange and the pixel
scores for the red sensor (Qr) in the center and at the
periphery of the throat (Table 1). The variables with
the heaviest loadings for PC2 were proportion of the
throat that was yellow and the pixel scores for the
green sensor (Qg) in the center and at the periphery
of the throat (Table 1). Generally, females with higher
PC1 scores had more orange throats and females with
higher PC2 scores had more yellow throats (Fig. 3B).
rePeatabiLity of coLour measurements
Repeatability (ICC) estimates were high for princi-
pal components describing colour variation in throat
patches. For males, the estimates were 0.83 and 0.74
for PC1 and PC2, respectively. For females, ICC esti-
mates were 0.65 and 0.76 for PC1 and PC2, respec-
tively. Field and photograph morph assignment was
consistent in 86% (356/413) of cases for males and 77%
(298/388) of cases for females. For males, 89% (49/55)
of the inconsistent cases involved distinguishing blue
from orange-blue individuals. For females, 56% (55/90)
of inconsistent cases involved distinguishing orange-
yellow from orange or yellow. For individuals with at
least two captures, 4% (3/77) of males and 31% (22/71)
of females were classified as a different colour morph
when recaptured. Therefore, colour measurements
from photographs were repeatable between captures
for individual lizards, but categorical morph assign-
ment was not always consistent between field and pho-
tograph assessment, or between captures of the same
individual in the field.
is coLour exPression simiLar between sites?
In males, there was no significant effect of site (F
= 0.671, d.f. = 5, 361, P = 0.646) and no interaction
between site and colour morph (F = 0.543, d.f. = 10,
361, P = 0.859) on PC1. There was no effect of site
(F = 0.610, d.f. = 5, 361, P = 0.693), but there was
a significant interaction between colour morph and
site on PC2 (F = 5.768, d.f. = 10, 361, P < 0.001).
The interaction was driven by two individuals
with orange-green throats at one site. When these
two individuals were removed, the interaction was
no longer significant (F = 1.765, d.f. = 10, 359, P =
0.066). In females, there was a significant effect of
site (F = 6.185, d.f. = 5, 360, P < 0.001) and no sig-
nificant interaction between colour morph and site
DISTINGUISHING DISCRETE POLYMORPHISM 77
© 2017 The Linnean Society of London, Biological Journal of the Linnean Society, 2017, 121, 72–81
on PC1 (F = 1.374, d.f. = 15, 360, P = 0.157). The site
effect arose because one site had lower PC1 scores
than the other sites, indicating there may be dif-
ferent frequencies of throat colour morphs between
sites. The non-significant interaction indicates that
throat colour morphs had similar quantitative col-
our expression between sites. There was no signifi-
cant effect of site (F = 1.540, d.f. = 5, 360, P = 0.177)
and no significant interaction between throat colour
morph and site on PC2 (F = 1.374, d.f. = 15, 360, P
= 0.157). Overall, therefore, we believe that throat
patch colours were similar at each site for males and
for females, and that our conclusions drawn from
pooling individuals from different sites are thus
robust to variation between these sites.
are the coLour morPhs discrete?
The DFA for males correctly classified 88% (363/413)
of lizards to their colour morph designation in the
field. The SVM classification (using a cost of 0.5 and
a gamma of 0.125) for males correctly classified 91%
(375/413) of lizards to their colour morph designation
in the field. The morph with the highest misclassifi-
cation rate was orange-green (16% of cases misclassi-
fied with DFA). For the male cluster analysis, the best
Figure 3. Biplot of principal component analyses on throat colour variation in Urosaurus ornatus from the Chiricahua
Mountains of Arizona, USA based on photographs of (A) males (n = 413 captures of 317 individuals), and (B) females
(n = 388 captures of 300 individuals) and coded by their assignment to throat colour morphs in the field.
78 J. E. PATERSON and G. BLOUIN-DEMERS
© 2017 The Linnean Society of London, Biological Journal of the Linnean Society, 2017, 121, 72–81
model had three clusters with varying volume, shape,
and orientation (BIC = 5653.29). This model had con-
siderably more support than the best model with two
(BIC = 5314.19) or four clusters (BIC = 5551.21). The
three identified clusters in the best mixture model for
males corresponded to individuals with (1) mostly blue
throats, (2) blue and orange throats, and (3) orange,
orange-green, and green throats (Table 2).
The DFA for females correctly classified 79%
(310/388) of lizards to their colour morph designation
in the field. The SVM classification (using a cost of 1
and a gamma of 0.125) for females correctly classified
82% (318/388) of lizards to their colour morph desig-
nation in the field. For the female cluster analysis,
the best model had four clusters with varying vol-
ume, shape, and orientation (BIC = 4403.73). This
model had more support than the best model with two
(BIC = 4014.55) or three clusters (BIC = 4269.73). The
identified clusters corresponded to individuals with
throats that were (1) very orange, (2) pale orange or
white, (3) orange and yellow, and (4) yellow (Table 2).
DISCUSSION
Analyzing tree lizard throat colour with photographs
and PCA was very effective at quantifying interin-
dividual variation. Colour was highly repeatable
(ICC > 0.74 for males, > 0.65 for females) between cap-
tures of the same individual. These estimates of colour
repeatability were comparable to those of other spe-
cies, including an agamid lizard (Ctenophorus ornatus;
0.65–0.91) and Green Lizards (Lacerta viridis; 0.89)
using spectrophotometry (LeBas & Marshall, 2000;
Vaclav, Prokop & Fekiac, 2007). The high repeatabil-
ity of throat colour measurements, including captures
more than a year apart, suggests that colour traits are
fixed within individual U. ornatus. Using a numeri-
cal description of colour is objective and captures
more variation between individuals than just morph
assignment. Capturing more subtle variation between
individuals is important for examining genetic and
environmental influences on colour and for testing
hypotheses about how colour is related to other phe-
notypic traits.
The DFA, SVM, and cluster analysis suggested there
are discrete male morphs, but distinguishing them is
difficult due to significant overlap in colour between
morphs. The DFA (88% accurate) and SVM (91%
accurate) were relatively accurate at classifying indi-
viduals into pre-determined morphs, and the cluster
analyses suggested there was more than one morph
(three clusters). However, the three morphs identified
in the cluster analysis of males did not correspond to
pre-determined morphs because one cluster included
most green, orange, and orange-green individuals.
These three morphs may not have been distinguished
because of the limited number of individuals in each
morph. Cluster analyses are less likely to detect
morphs as group size decreases (de Craen et al., 2006).
There were only five individuals in the orange morph,
and these individuals ended up in a cluster with the
other less common morphs (green and orange-green).
Correctly identifying rare morphs with cluster analy-
ses will always be problematic.
Female colour displayed more continuous variation
between individuals than male colour (Fig. 3B). The
cluster analysis for females identified four morphs
Table 2. Cluster assignments of Urosaurus ornatus from
the Chiricahua Mountains of Arizona, USA based on the
best normal mixture models of throat colour variation for
males (three clusters; n = 413 captures of 317 individuals)
and females (four clusters; n = 388 captures of 300 individu-
als) compared to their assignment to throat colour morphs
in the field.
Throat colour Cluster
1234
Males Blue 125 0 64 –
Green 0 22 1
Orange 0 5 0 –
Orange-blue 83 2 61 –
Orange-green 0 45 5
Females Orange 145 27 3 17
Orange-yellow 6 9 1 47
White 0 57 0 3
Yellow 2 4 1 70
Table 1. Variable loadings for the first two components
from principal component analyses on variables describ-
ing throat colour patches for male (n = 413 captures of 317
individuals) and female (n = 388 captures of 300 individu-
als) Urosaurus ornatus from the Chiricahua Mountains of
Arizona, USA.
Variable Males Females
PC1 PC2 PC1 PC2
Proportion blue 0.39 0.13
Proportion green –0.30 –0.37
Proportion orange –0.30 0.36 0.85
Proportion yellow –0.32 –0.34 0.70
Center Qr–0.36 –0.14 0.42 –
Center Qg–0.14 –0.19 –0.30 0.49
Center Qb0.38 0.22 –0.38 –0.28
Periphery Qr–0.34 0.39 0.41 0.18
Periphery Qg0.20 –0.55 –0.24 0.23
Periphery Qb0.36 –0.21 –0.38 –0.32
DISTINGUISHING DISCRETE POLYMORPHISM 79
© 2017 The Linnean Society of London, Biological Journal of the Linnean Society, 2017, 121, 72–81
that broadly corresponded to the pre-determined
morphs of orange, orange-yellow, white, and yellow.
However, identifying the boundaries of female morphs
in parameter space was difficult and resulted in a
higher rate of misclassification in the DFA and SVM
classification for females. In the DFA, there was a 21%
misclassification rate overall, but a 52% misclassifica-
tion rate of orange-yellow females that lie in the center
of the parameter space (Fig. 3B). In the SVM, there
was an 18% misclassification rate overall, which dem-
onstrates that continuous variation between morphs
is real and not a consequence of the type of classifica-
tion analysis used to distinguish between morphs. The
relatively high number of females that were classi-
fied into different morphs when they were recaptured
(28%) also suggests that female colour displays sig-
nificant continuous variation, and that distinguishing
between supposedly discrete morphs in the field may
be unreliable.
Throat colour in male U. ornatus is related to several
other phenotypic traits. For example, colour influences
dominance (Carpenter, 1995), trophic niche (Lattanzio,
Miles & Lattanzio, 2016), and home range size (Moore,
Hewst & Knapp, 1998) in other populations. However,
those studies assumed throat colour to be discrete and
have not considered the possibility of continuous vari-
ation in colour. Our data indicate significant continu-
ous variation in several aspects of male throat colour,
including colour intensity and the proportion of the
throat patch of different colours. Therefore, we sug-
gest that predictions about how colour relates to other
traits would best be tested with numerical descrip-
tors rather than with discrete categories. For instance,
because of the known behavioural differences between
individuals with different colours, individuals with
different colours likely occupy different habitats. We
predict that males with more blue (higher PC1) or
green (lower PC2) and less orange should secure home
ranges in better habitats because they are more domi-
nant (Carpenter, 1995).
There is little known about how female tree lizard
throat colour relates to other phenotypic traits. Some
populations of tree lizards are monomorphic for female
throat colour (Zucker & Boecklen, 1990), but at all of
our study sites females varied from white to yellow
and orange. In a related species [Uta stansburiana
(U. stansburiana)], female throat colour is related to
alternative reproductive strategies where morphs
invest in offspring differently and their success fluctu-
ates with density (Sinervo et al., 2000, 2001). Recently,
behavioural trials have shown that orange and yel-
low throated female tree lizards respond differently
to courting males (Lattanzio et al., 2014). However, in
previous studies authors again assumed that colour
was discrete and scored the trait with a small number
of discrete values. The continuous variation between
white, orange, and yellow we observed in the present
study suggests that female U. ornatus throat colour
cannot be reliably assigned to one of a few discrete
morphs.
Quantifying variation in colour is difficult because it
may represent multiple phenotypic traits. The reflec-
tion of different wavelengths is caused by different cel-
lular mechanisms, so it is possible that different genes
or condition-dependent factors influence them. For
example, blue is typically a structural colour caused
by the scattering of light in iridiophores (Bagnara,
Fernandez & Fujii, 2007), while orange and yellow
are usually from carotenoid or pteridine pigments in
xanthophores (Evans & Sheldon, 2014), and green
is caused by a combination of cells that scatter light
and contain yellow pigment (Bagnara et al., 2007).
Typically, carotenoid pigments are acquired through
the diet and are thus influenced by the environment
(Hill et al., 1994), while pteridine pigments can be syn-
thesized and expression is thus not influenced by diet
(Hurst, 1980). Although the cellular basis of throat col-
our has not been examined in Urosaurus, both pteri-
dines and carotenoids have been found in the throat
colour patches of U. stansburiana, a closely related
species (Haisten et al., 2015). In addition, colour may
vary continuously in one aspect, such as the proportion
of blue on the throat, but vary discretely in another
aspect, such as the presence of green. Considering the
various mechanisms underlying colour expression,
we expect continuous variation in colour to be com-
mon. Therefore, we suspect that quantifying different
aspects of coloration as we did here will represent real-
ity better than classification in a few discrete colour
morphs, and will thus allow better tests of the various
hypotheses that have been proposed as explanations
of the presence and of the evolutionary maintenance of
colour polymorphisms.
There are several promising lines of research to
uncover the underlying causes of colour expression in
U. ornatus and other lizards. First, genetic data from
individual lizards could be used to test whether there
is a link with colour expression, ideally using nuclear
markers. Second, experimental crosses of different
colour phenotypes, and in which the parental and
offspring throat colours are measured quantitatively,
could be used to determine the heritability of the col-
our traits. Finally, histological sections of throat colour
patches could be used to determine which aspects of
colour are structural and which aspects are derived
from the environment.
We used photographs to demonstrate that both male
and female U. ornatus exhibit significant interindivid-
ual variation in throat colour that is not easily catego-
rized into discrete morphs. Our cluster analyses suggest
there could be discrete throat colour morphs in this spe-
cies, but that their parameter boundaries are not easily
80 J. E. PATERSON and G. BLOUIN-DEMERS
© 2017 The Linnean Society of London, Biological Journal of the Linnean Society, 2017, 121, 72–81
distinguished due to significant within morph varia-
tion in throat colour. Heritability estimates and genetic
mapping would provide additional evidence of whether
throat colour in tree lizards represents a discrete poly-
morphism or a continuous trait. Using PCA to summa-
rize colour expression can be used to test how colour
is related to other traits without having to collapse
interindividual variation into a few discrete morphs.
Considering the suite of factors that influence colour
expression, it seems unlikely that a single gene (or a
few tightly linked genes) is the only influence on colour
expression. It appears more likely that there is signifi-
cant continuous variation between individuals caused
by some combination of genetic and environmental fac-
tors. We encourage researchers to quantify formally
interindividual variation in colour expression before
assuming that it represents a discrete trait.
ACKNOWLEDGEMENTS
We would like to thank Valérie Bertrand, Lucy
Patterson, Dariya Quenneville, and Hannah Watkins
for assisting with data collection. This manuscript was
improved by the constructive comments provided by
three anonymous reviewers. We are grateful for logis-
tical support provided by the Southwestern Research
Station (American Museum of Natural History).
FUNDING
This research was funded by a Natural Sciences and
Engineering Research Council of Canada (NSERC)
grant to G.B.D. and an NSERC scholarship to J.E.P.
References
Bagnara JT, Fernandez PJ, Fujii R. 2007. On the blue col-
oration of vertebrates. Pigment Cell Research 20: 14–26.
Bergman TJ, Beehner JC. 2008. A simple method for meas-
uring colour in wild animals: validation and use on chest
patch colour in geladas (Theropithecus gelada). Biological
Journal of the Linnean Society 94: 231–240.
Blouin-Demers G, Lourdais O, Bouazza A, Verreault C,
Mouden H El, Slimani T. 2013. Patterns of throat colour
variation in Quedenfeldtia trachyblepharus, a high-altitude
gecko endemic to the High Atlas Mountains of Morocco.
Amphibia-Reptilia 34: 567–572.
Brommer JE, Ahola K, Karstinen T. 2005. The colour of fit-
ness: plumage coloration and lifetime reproductive success
in the tawny owl. Proceedings of the Royal Society of London
B 272: 935–940.
Carpenter GC. 1995. Modeling dominance: the influence of
size, coloration, and experience on dominance relations in
tree lizards (Urosaurus ornatus). Herpetological Monographs
9: 88–101.
Chittka L, Shmida A, Troje N, et al. 1994. Ultraviolet as a
component of flower reflections, and the colour perception of
hymenoptera. Vision Research 34: 1489–1508.
Cogliati KM, Corkum LD, Doucet SM. 2010. Bluegill col-
oration as a sexual ornament: evidence from ontogeny, sex-
ual dichromatism, and condition dependence. Ethology 116:
416–428.
Corl A, Davis AR, Kuchta SR, Comendant T, Sinervo
B. 2010. Alternative mating strategies and the evolution
of sexual size dimorphism in the side-blotched lizard, Uta
Stansburiana: a population-level comparative analysis.
Evolution 64: 79–96.
Cote J, Le Galliard JF, Rossi JM, Fitze PS. 2008.
Environmentally induced changes in carotenoid-based color-
ation of female lizards: a comment on Vercken et al. Journal
of Evolutionary Biology 21: 1165–1172.
de Craen S, Commandeur JF, Frank LE, Heiser WJ 2006.
Effects of group size and lack of sphericity on the recov-
ery of clusters in K-means cluster analysis. Multivariate
Behavioral Research 41: 127–145.
Ekner A, Sajkowska Z, Dudek K, Tryjanowski P. 2011.
Medical cautery units as a permanent and non-invasive
method of marking lizards. Acta Herpetologica 6: 229–236.
Evans SR, Sheldon BC. 2014. Colour in a new light: A spec-
tral perspective on the quantitative genetics of carotenoid
colouration. Functional Ecology 29: 96–103.
Fleishman LJ, Bowman M, Saunders D, Miller WE, Rury
MJ, Loew ER. 1997. The visual ecology of Puerto Rican ano-
line lizards: habitat light and spectral sensitivity. Journal of
Comparative Physiology – A Sensory, Neural, and Behavioral
Physiology 181: 446–460.
Ford EB. 1945. Polymorphism. Biological Reviews 20: 73–88.
Forsman A, Ahnesjo J, Caesar S, Karlsson M. 2008. A
model of ecological and evolutionary consequences of color
polymorphism. Ecology 89: 34–40.
Fraley C, Raftery AE. 2002. Model-based clustering, dis-
criminant analysis, and density estimation. Journal of the
American Statistical Association 97: 611–631.
Gray SM, McKinnon JS. 2007. Linking color polymorphism
maintenance and speciation. Trends in Ecology & Evolution
22: 71–79.
Haisten DC, Paranjpe D, Loveridge S, Sinervo B. 2015.
The cellular basis of polymorphic coloration in common
side-blotched lizards, Uta stansburiana. Herpetologica 71:
125–135.
Hews DK, Thompson CW, Moore IT, Moore MC. 1997.
Population frequencies of alternative male phenotypes in tree
lizards: rearing studies and common-garden variation geo-
graphic. Behavioral Ecology and Sociobiology 41: 371–380.
Hill GE, Montgomerie R, Inouye CY, Dale J. 1994.
Influence of dietary carotenoids on plasma and plumage col-
our in the house finch: intrasexual and intersexual variation.
Functional Ecology 8: 343–350.
Hill GE, Montgomerie R. 1994. Plumage colour signals
nutritional condition in the house finch. Proceedings of the
Royal Society of London B 258: 47–52.
DISTINGUISHING DISCRETE POLYMORPHISM 81
© 2017 The Linnean Society of London, Biological Journal of the Linnean Society, 2017, 121, 72–81
Hubbard JK, Uy JAC, Hauber ME, Hoekstra HE, Safran
RJ. 2010. Vertebrate pigmentation: from underlying genes
to adaptive function. Trends in Genetics 26: 231–239.
Hurst DT. 1980. An Introduction to the chemistry and bio-
chemistry of pyrimidines, purines, and pteridines. New York:
Wiley.
Knapp R, Hews DK, Thompson CW, Ray LE, Moore MC
2003. Environmental and endocrine correlates of tactic
switching by nonterritorial male tree lizards (Urosaurus
ornatus). Hormones and Behavior 43: 83–92.
Knapton RW, Falls JB. 1983. Differences in parental contri-
bution among pair types in the polymorphic white-throated
sparrow. Canadian Journal of Zoology 61: 1288–1292.
Lattanzio MS, Metro KJ, Miles DB. 2014. Preference for
male traits differ in two female morphs of the tree lizard,
Urosaurus ornatus. PloS One 9: e101515.
Lattanzio MS, Miles DB. 2014. Ecological divergence among
colour morphs mediated by changes in spatial network struc-
ture associated with disturbance. The Journal of Animal
Ecology 83: 1490–1500.
Lattanzio MS, Miles DB, Lattanzio MS. 2016. Trophic niche
divergence among colour morphs that exhibit alternative
mating tactics. Royal Society Open Science 3: 150531.
LeBas NR, Marshall NJ. 2000. The role of colour in sig-
nalling and male choice in the agamid lizard Ctenophorus
ornatus. Proceedings of the Royal Society of London Series
B-Biological Sciences 267: 445–452.
Lessells CM, Boag PT. 1987. Unrepeatable repeatabilities: a
common mistake. The Auk 104: 116–121.
McGraw KJ, Ardia DR. 2003. Carotenoids, immunocompe-
tence, and the information content of sexual colors: an exper-
imental test. The American Naturalist 162: 704–712.
McKinnon JS, Pierotti ME. 2010. Colour polymorphism and
correlated characters: genetic mechanisms and evolution.
Molecular Ecology 19: 5101–5125.
Meyer D, Dimitriadou E, Hornik K, Weingessel A, Leisch
F. 2014. e1071: Misc functions of the department of statis-
tics (e1071), TU Wien. R package version 1.6-2. Available at:
http://CRAN.R-project.org/package=e1071.
Moore MC, Hewst DK, Knapp R. 1998. Hormonal control
and evolution of alternative male phenotypes : generaliza-
tions of models for sexual differentiation. American Zoologist
151: 133–151.
Pryke SR, Astheimer LB, Buttemer WA, Griffith SC.
2007. Frequency-dependent physiological trade-offs between
competing colour morphs. Biology Letters 3: 494–497.
R Core Team. 2015. R: A Language and Environment for
Statistical Computing. Version 3.1.1. Available at: http://
www.R-project.org/
Roulin A. 2004. The evolution, maintenance and adaptive
function of genetic colour polymorphism in birds. Biological
Reviews of the Cambridge Philosophical Society 79: 815–848.
Salvador A, Veiga JP, Martin J, Lopez P, Abelenda M,
Puerta M 1996. The cost of producing a sexual signal:
testosterone increases the susceptibility of male lizards to
ectoparasitic infestation. Behavioral Ecology 7: 145–150.
Schneider CA, Rasband WS, Eliceiri KW. 2012. NIH Image to
ImageJ: 25 years of image analysis. Nature Methods 9: 671–675.
Sinervo B, Bleay C, Adamopoulou C. 2001. Social causes
of correlational selection and the resolution of a herit-
able throat color polymorphism in a lizard. Evolution 55:
2040–2052.
Sinervo B, Lively C. 1996. The rock-paper-scissors game and
the evolution of alternative male strategies. Nature 380:
240–243.
Sinervo B, Svensson E, Comendant T. 2000. Density cycles
and an offspring quantity and quality game driven by natu-
ral selection. Nature 406: 985–988.
Smith TB, Skulason S. 1996. Evolutionary significance of
resource polymorphisms in fishes, amphibians, and birds.
Annual Review of Ecology and Systematics 27: 111–133.
Stevens M, Párraga CA, Cuthill IC, Partridge JC,
Troscianko TOMS. 2007. Using digital photography to
study animal coloration. Biological Journal of the Linnean
Society 90: 211–237.
Teasdale LC, Stevens M, Stuart-Fox D. 2013. Discrete
colour polymorphism in the tawny dragon lizard
(Ctenophorus decresii) and differences in signal conspicu-
ousness among morphs. Journal of Evolutionary Biology
26: 1035–1046.
Therneau T, Atkinson B, Ripley B. 2015. rpart: recursive
partitioning and regression trees. Version 4.1-10. Available
at: https://cran.r-project.org/package=rpart
Thompson CW, Moore MC. 1991. Syntopic occurrence of mul-
tiple dewlap color morphs in male tree lizards, Urosaurus
ornatus. Copeia 1991: 493–503.
Vaclav R, Prokop P, Fekiac V. 2007. Expression of breed-
ing coloration in European Green Lizards (Lacerta viridis):
variation with morphology and tick infestation. Canadian
Journal of Zoology 85: 1199–1206.
Vercken E, Massot M, Sinervo B, Clobert J. 2007.
Colour variation and alternative reproductive strategies in
females of the common lizard Lacerta vivipara. Journal of
Evolutionary Biology 20: 221–232.
Vercken E, Sinervo B, Clobert J. 2008. Colour variation
in female common lizards: why we should speak of morphs,
a reply to Cote et al. Journal of Evolutionary Biology 21:
1160–1164.
Weiss SL. 2006. Female-specific color is a signal of quality in
the striped plateau lizard (Sceloporus virgatus). Behavioral
Ecology 17: 726–732.
White TE, Dalrymple RL, Noble DWA, O’Hanlon JC,
Zurek DB, Umbers KDL. 2015. Reproducible research in
the study of biological coloration. Animal Behaviour 106:
51–57.
Wilson AJ, Nussey DH. 2010. What is individual quality? An
evolutionary perspective. Trends in Ecology & Evolution 25:
207–214.
Zucker N, Boecklen W. 1990. Variation in female throat
coloration in the tree lizard (Urosaurus ornatus): rela-
tion to reproductive cycle and fecundity. Herpetologica 46:
387–394.
... As stated in Paterson and Blouin-Demers [7] , "To use photographs for the analysis of color, a digital camera's three sensors (corresponding to red, green, and blue wavelengths) should respond linearly and equally to increases in light intensity ". Contrary to Paterson and Blouin-Demers [7] , our observed reflectance values already appeared linearly related to expected reflectance values before any corrections because we used ImageJ's DCRaw package [ 3 , 10 ]. ...
... As stated in Paterson and Blouin-Demers [7] , "To use photographs for the analysis of color, a digital camera's three sensors (corresponding to red, green, and blue wavelengths) should respond linearly and equally to increases in light intensity ". Contrary to Paterson and Blouin-Demers [7] , our observed reflectance values already appeared linearly related to expected reflectance values before any corrections because we used ImageJ's DCRaw package [ 3 , 10 ]. The observed reflectance values, however, were always lower than the expected reflectance values. ...
... If the relationships had not been linear (as it was the case for Paterson and Blouin-Demers [7] ), we could have used the following equations instead, using the nls() function in R: ...
Article
Full-text available
Coloration is often used in biological studies, for example when studying social signaling or antipredator defense. Yet, few detailed and standardized methods are available to measure coloration using digital photography. Here we provide a step-by-step guide to help researchers quantify coloration from digital images. We first identify the do's and don'ts of taking pictures for coloration analysis. We then describe how to i) extract reflectance values with the software ImageJ; ii) fit and apply linearization equations to reflectance values; iii) scale and select the areas of interest in ImageJ; iv) standardize pictures; and v) binarize and measure the proportion of different colors in an area of interest. We apply our methodological protocol to digital pictures of painted turtles (Chrysemys picta), but the approach could be easily adapted to any species. More specifically, we wished to calculate the proportion of red and yellow on the neck and head of turtles. With this protocol, our main aims are to make coloration analyses with digital photography:•More accessible to researchers without a background in photography. •More consistent between studies.
... Based on these assumptions, color polymorphisms have long been used as visual markers of genetic variation, thereby providing a simple but effective tool to study a variety of evolutionary processes in a range of different organisms [8][9][10] . However, within what are considered discrete color morphs, there is mounting evidence for the presence of continuous variation in coloration 5,[11][12][13] , which is thought to be typically controlled by complex arrays of genes (reviewed in 14 ), and moreover can be sensitive to environmental variation -i.e. phenotypically plastic 15,16 . ...
... Despite the importance of genetic color polymorphisms for biological diversity 51,52 , explicit consideration of intramorph phenotypic variation remains limited in the study of evolutionary processes 5,12 . Based on our findings, we argue that quantifying intramorph phenotypic variation can increase our understanding of how polymorphisms are maintained 52,53 , but also how populations persist in space and time 22,54 . ...
Preprint
Full-text available
There is a long tradition of using heritable color polymorphisms in natural populations to study selection, gene flow, and other evolutionary processes. However, we still have only limited knowledge on how continuous color variation within genetically discrete morphs affects selective dynamics, which narrows our understanding of how color polymorphisms persist. Our comprehensive analysis of 43 years of plumage color scores from a bi-morphic Finnish population of tawny owls (Strix aluco) reveals that intra-morph variation is substantial, but also unexpectedly dynamic. We show that both morphs recently diverged in their plumage coloration: while the brown morph is on a steady trajectory toward more intense plumage pigmentation, the gray morph has recently shifted toward lighter plumage, following a series of extreme winters. Evidence suggests that this divergence is driven by the brown morph's higher reproductive success and greater plasticity in response to seasonality, while the gray morph, with its more genetically determined plumage color and limited standing variation, has a reduced capacity to respond to selection and track new phenotypic optima. Our study confirms the presence of substantial and dynamic phenotypic variation within genetically discrete color morphs, and demonstrates its relevance for the evolutionary potential of populations to respond to environmental change.
... Given the known differences between the visual systems of lizards and humans, establishing the existence of discrete colour morphs from the lizards' perspective is essential for many current hypotheses about the genetic underpinnings of the polymorphism and the evolutionary processes generating and maintaining it (e.g. Cote et al. 2008;Paterson and Blouin-Demers 2017). ...
... Given the known differences between the visual systems of lizards and humans, establishing the existence of discrete colour morphs from the lizards' perspective is essential for many current hypotheses about the genetic underpinnings of the polymorphism and the evolutionary processes generating and maintaining it (e.g. Cote et al., 2008;Paterson and Blouin-Demers, 2017). ...
Thesis
Full-text available
Colour polymorphisms, the coexistence of two or more colour morphs of a species within a population, have long fascinated evolutionary biologists interested in the mechanisms generating and maintaining phenotypic variation in nature. The functional significance of colour polymorphisms (i.e. their adaptive value) is often linked to the selective mechanisms responsible for their maintenance over time. In lizards, the hypothesis that colour morphs may reflect alternative reproductive strategies involving differential sociosexual behaviour and/or alternative reproductive strategies has come to dominate the field. Wall lizards (family Lacertidae, genus Podarcis), with several geographically distant species that exhibit two or more alternative ventral colour morphs, have often been identified as a group in which alternative reproductive strategies and frequency-dependent selection likely underpin colour polymorphisms. However, several key aspects regarding the link between behaviour, fitness, and lacertid colour polymorphism remain overlooked or inadequately studied. In this thesis we tried to fill these gaps by experimentally addressing some of the most commonly assumed ideas about the functional significance of colour polymorphism in the European common wall lizard (Podarcis muralis, Laurenti 1768). In some populations of this species (e.g. eastern Pyrenees), adult lizards of both sexes may show up to three “pure” alternative colour morph (orange, white, yellow), and two mixed morphs consisting of a mosaic of differently-coloured scales (orange-white, yellow-orange). Although colour morphs are frequently thought to function as social signals mediating intraspecific interactions, morph categorization has been never assessed from the viewpoint of the intended receivers (i.e. conspecifics). Here, using a discrimination experiment and visual modelling, we found that lizards likely perceive the alternative colour morphs as chromatically distinct and show evidence of discriminating among them based on hue, rather than luminance. To evaluate the role of male coloration (including alternative colour morphs) in intra-sexual competition), we conducted lab-staged dyadic contests among size-matched males. We found lower fighting ability in lizards showing a smaller fraction of their outer ventral scales covered with dark melanin-based spots, and in lizards with orange ventral colour (which could result from the existence of a subordinate non-territorial strategy in this morph). However, our results in later studies (with a free-ranging population and ten experimental mesocosm enclosures), strongly refute the hypothesis that alternative socio-sexual behaviour or space use strategies characterize colour morphs when lizards interact under more natural conditions. In the free-ranging populations, colour morphs did not differ in inter-annual site-fidelity, home-range size, or male-female overlap. In the mesocosm enclosures, spatial dominance was the prime determinant of male fitness across colour morphs. In a later experiment, we conducted controlled matings among pure colour morphs and found no overall effect of female morph on clutch size or juvenile mass, and no effect of morph combination on offspring viability or prospective fitness. These results refute the existence of alternative reproductive strategies in female morphs and are also in disagreement with predictions from both correlational selection and heterosis. Lastly, by keeping the juveniles of known crosses in outdoor enclosures for a year, we studied the inheritance and ontogeny of P. muralis colour polymorphism. Specifically, our results confirmed that orange and yellow colour expression depends on two recessive alleles located at two separate autosomal loci and revealed that the whitish coloration exhibited by newborn lizards is likely perceived by conspecifics as a chromatically distinct colour different from any of the morph colours expressed by adult lizards. Overall, in this thesis we have hopefully presented compelling arguments to revise our perspective on the functional significance of lacertid colour polymorphisms, suggested promising lines of research for future work, and generally contributed to our understanding of the processes maintaining intra-specific variation in natural populations at large.
... Colour polymorphism implies the co-occurrence of discrete morphs by definition, but the extent to which variation between individuals can fit into discrete classes or is continuous with a strongly bimodal (or multimodal) distribution has recently been a subject of debate (Cote et al., 2008;Davison et al., 2019;Paterson & Blouin-Demers, 2017;Vercken et al., 2008). Studies of colour polymorphic systems have often focused on specific colour patches that best separate morphs to classify individuals into discrete categories rather than integrating colour variation across the entire body (but see Chaine & Lyon, 2008Greene et al., 2000). ...
... Studies of colour polymorphic systems have often focused on specific colour patches that best separate morphs to classify individuals into discrete categories rather than integrating colour variation across the entire body (but see Chaine & Lyon, 2008Greene et al., 2000). Although some species seem to show a clear pattern of discrete polymorphism (Pryke & Griffith, 2006;Teasdale et al., 2013), some recent studies on classic cases of colour polymorphism have shown that variation among individuals may be better described by a continuous distribution owing to extensive within-morph variation or the presence of intermediate types (Brommer et al., 2005;Davison et al., 2019;Parejo et al., 2018;Paterson & Blouin-Demers, 2017;Rankin et al., 2016). ...
Article
Categorizing individuals into discrete forms in colour polymorphic species can overlook more subtle patterns in coloration that can be of functional significance. Thus, quantifying inter‐individual variation in these species at both within‐ and between‐morph levels is critical to understand the evolution of colour polymorphisms. Here we present analyses of inter‐individual colour variation in the Reunion grey white‐eye (Zosterops borbonicus), a colour polymorphic wild bird endemic to the island of Reunion in which all highland populations contain two sympatric colour morphs, with birds showing predominantly grey or brown plumage, respectively. We first quantified colour variation across multiple body areas by using a continuous plumage colour score to assess variation in brown‐grey coloration as well as smaller scale variation in light patches. To examine the possible causes of among‐individual variation, we tested if colour variation in plumage component elements could be explained by genotypes at two markers near a major‐effect locus previously related to back coloration in this species, and by other factors such as age, sex and body condition. Overall, grey‐brown coloration was largely determined by genetic factors and was best described by three distinct clusters that were associated to genotypic classes (homozygotes and heterozygote), with no effect of age or sex, whereas variation in smaller light patches was primarily related to age and sex. Our results highlight the importance of characterizing subtle plumage variation beyond morph categories that are readily observable since multiple patterns of colour variation may be driven by different mechanisms, have different functions and will likely respond in different ways to selection. Inter‐individual color variation in a polymorphic wild bird endemic to the island of Reunion (Zosterops borbonicus). Our results highlight the importance of characterizing subtle plumage variation beyond morph categories.
... Lively 1996, Roulin 2004). But although many examples of colour variation can be found across lizard species (Olsson et al. 2013, Stuart-Fox et al. 2020, it is not always so straightforward to define whether colour varies continuously or whether it corresponds to discrete colour morphs (Paterson and Blouin-Demers 2017). For example, ventral colour variation in Zootoca vivipara has been classified as polymorphic by some authors (Vercken et al. 2007(Vercken et al. , 2008, whereas other studies suggest that Z. vivipara coloration varies continuously (Cote et al. 2008). ...
Article
Animal coloration is often shaped by a myriad of factors that lead to differences in colour through changes in the chromatophores. Depending on how this variation is partitioned, coloration is often categorized as continuous or polymorphic. However, the boundaries between these two categories are not always clear. Here, we investigated whether the ventral coloration of the Madeiran wall lizard (Teira dugesii) varies continuously or corresponds to discrete colour morphs, via by-eye colour classification and visual modelling. By combining these two approaches, we show that T. dugesii coloration varies continuously and that colour classifications based on anthropomorphic approaches alone are ill suited to describe animal coloration. We also tested the influence of size, body condition, and sex as possible factors that might explain differences in T. dugesii coloration. We found that body condition, and especially size and sex, explain a great proportion of the variability observed in this species. These differences point to an effect of ontogeny, which might play a major role in colour development owing to the longevity of this species. Moreover, the sexual dichromatism that this species shows is indicative of an effect of sexual selection on coloration, perhaps explained by differences in circulating hormones.
... Whilst several studies use reflectance spectrophotometry to categorise colour, or to determine whether variation is continuous or discrete Reflectance spectrophotometry is revealing that variation of colour in several species of lizard is continuous, rather than discrete as previously assumed (Cote et al. 2008, Vercken et al. 2008, Paterson and Blouin-Demers 2017. However, even within populations which do possess wholly discrete colour morphs, it is likely that there is some variation within these discrete categories. ...
Article
The study of both animal colouration and patterning across the natural world has been imperative for understanding some of the key principles of biology throughout the past century, particularly with respect to evolution and genetics. Generally, colour and pattern have each been described qualitatively, often being binned into discrete groups relying on human perception of colour or pattern, rather than being considered in a biologically relevant context. ‘Binning’ traits into discrete groups has the consequence that variation within discrete morphs is often overlooked. Terrestrial gastropods, such as the selected study genera Cepaea and Auriculella, provide an ideal system for the study of polymorphisms and colour variation due to the extreme variety of morphs present across the taxa, as well as the nature of shell growth providing a complete ontogeny of an individual. The aims of this thesis are threefold; firstly, I aimed to understand finescale variation within and between established banding pattern morphs in Cepaea, to allow inferences to be made regarding the genetic mechanisms responsible for this variation. The implementation of two quantitative methods for measuring variation in band position and width in individual shells found that individual band absence has a minor but significant effect on the position of the remaining bands, implying that the locus controlling band presence/absence acts mainly after the position of bands is determined. I establish a method which is useful for comparative studies of quantitative banding variation in snail shells, and for extraction of growth parameters and morphometrics, highlighting the importance and usefulness of gastropod shells in the understanding of how variation is established and maintained in a population. Secondly, I aimed to understand the shell colour variation present in both Cepaea nemoralis and Cepaea hortensis by using spectrophotometry and psychophysical modelling in tandem. It was revealed that colour variation in Cepaea hortensis is continuous, with no detectable effects of geographic location with the exception of an association of the paleness of yellow shells with latitude. Differences between the colour of Cepaea hortensis and Cepaea nemoralis, both in terms of exact shade and overall colour were revealed; Cepaea hortensis are generally paler, and less pink-toned, but slightly more brown-toned. Precise shade variation of yellow individuals from genetically diverse lineages of Cepaea nemoralis were also detected. The results presented have significance in furthering the understanding of the precise nature of the colour polymorphism displayed in Cepaea spp., and the nature of the selection which acts upon it, as well as highlighting the importance of considering colour as a continuous trait, rather than binning it into discrete groups. Thirdly, I aimed to investigate colour variation across a number of scales in the Hawaiian land snail genus Auriculella, to allow inferences about the genetic architecture responsible for the variation, and to highlight the usefulness of museum collections of gastropod shells in understanding variation in extinct or endangered species. I demonstrated that there are differences in colour within single shells of Auriculella, similar to variation displayed by other Pacific Island snails. I described significant variation between isolated populations of the same species, and determined that there is no difference in colour variation between shells on the islands of Maui and Oahu. Finally, I demonstrated that there is no difference in colour between shell chiralities, suggesting that interchiral mating is not uncommon, and that the loci responsible for colour variation and chirality are not closely linked. By describing the variation present in Auriculella in a biologically relevant context, inference regarding genetic mechanisms of variation becomes possible in a taxa of conservation concern. By achieving these aims, and synthesising conclusions drawn from their achievement, I have highlighted the importance of accurately defining phenotypes for the purposes of evolutionary ecology and genetics. Defining phenotypes and investigating variation present within morphs has allowed inferences to be made regarding the underpinning genetic mechanisms which control variation in two gastropod genera, although the principles are applicable to other taxonomic groups. Finally, and more broadly, I have demonstrated the usefulness of gastropods as study systems, particularly where large collections of shells are available.
Article
Few studies have explored eye (iris) color evolution beyond humans and domesticated animals. Felids exhibit significant eye color diversity, unlike their brown-eyed relatives, making them an ideal model to study the evolution of eye color in natural populations. Through machine learning analysis of public photographs, five felid eye colors were identified: brown, green, yellow, gray, and blue. The presence or absence of these colors was reconstructed on a phylogeny, as well as their specific quantitative shades. The ancestral felid population likely had brown-eyed and gray-eyed individuals, the latter color being pivotal for the diversification of eye color seen in modern felids. Additionally, yellow eyes are highly associated with and may be necessary for, the evolution of round pupils in felids. These findings enhance the understanding of eye color evolution, and the methods presented in this work are widely applicable and will facilitate future research into the phylogenetic reconstruction of color beyond irises.
Preprint
Colour polymorphism is a pervasive phenomenon in both animal and plant kingdoms and understanding its evolution and maintenance is of great interest. Among the lizards of Sri Lanka, the endemic skink Lankascincus fallax shows throat colour polymorphism in which, the underlying basis is not clearly known. In this study, we examined the relationship of the three different throat colour morphs observed in this species with the sex, body size and the geographic location of L. fallax . Live skinks were sampled from two locations in Sri Lanka and sex and the throat colour was categorized visually and the snout to vent length (SVL) was measured. Tail tips of some selected individuals from the two locations were taken and a fragment of the 12S rRNA gene was sequenced in representative individuals having the different throat colour morphs. Pairwise genetic distance of the three colour morphs ranged between 0.4–0.5% confirming that the three colour morphs were the same species. Three colour morphs (red, black and white) were observed in males in both locations, while only the white morph was observed in females, suggesting that the throat color polymorphism was confined to males. There was a significant difference between the mean SVL of males with red and black throat colours (39.35 mm) and males with white throat colours (30.31 mm). Thus, the study suggests that the throat colour in these skinks is highly associated with sex and the body size in males. The study further suggests that L. fallax is sexually dichromatic and that the males show throat colour polymorphism. However, future studies are necessary to understand the underlying drivers for the presence and maintenance of sexual dichromatism and throat colour polymorphism in L. fallax .
Article
Northern Pacific Rattlesnakes (Crotalus oreganus) are ambush-hunting predators that rely on crypsis to forage and to avoid encounters with predators, yet little is known about color variation in this ubiquitous rattlesnake species. This study addressed how coloration and contrast in tail banding in C. oreganus vary among habitat types in Washington State. We also explored whether C. oreganus exhibit sexual dimorphism in coloration or tail band contrast (sexual dichromatism). We sampled 127 rattlesnakes from seven different populations across northern and central Washington state during spring emergence from overwintering hibernacula in 2017 and 2018. We characterized snake coloration as red/blue color ratios from standardized photographs, and used GIS supervised classification schemes of satellite imagery to characterize habitat. We used generalized linear models to assess relationships among snake color and tail bands, habitat, and sex. We found that coloration (red/blue color ratios) in C. oreganus varied greatly across the landscape, both within and among populations, likely a reflection of their variable and heterogenous shrub-steppe and forest ecotone habitats. In 20% of 21 pairwise comparisons, populations differed in body color. Rattlesnake coloration was not associated with habitat (amount of forested land within 0.5 km of the snake den), but male and female rattlesnakes showed different associations between color and percent forested habitat. Male rattlesnakes did not differ in body coloration from females, but males showed greater contrast than females in the black and white banding present on the tail. We discuss several, nonmutually exclusive, hypotheses for sexual dichromatism in tail band contrast, including the possibility that tail banding constitutes warning coloration in rattlesnakes. Our results suggest that sexual dichromatism, and the role of tail banding in rattlesnake ecology, are topics worthy of further investigation.
Article
Full-text available
Discrete colour morphs associated with alternative mating tactics are assumed to be ecologically equivalent. Yet suites of behaviours linked with reproduction can also favour habitat segregation and exploitation of different prey among morphs. By contrast, trophic polymorphisms are usually attributed to morphs exhibiting habitat or prey selectivity. An alternative hypothesis is that habitat variation generates a trophic polymorphism driven by differences in morph reproductive behaviour, the spatial dispersion of morphs in a landscape and their exposure to different prey types. In this scenario, morphs are allowed to vary in habitat or diet selectivity (e.g. specialist or generalist) as they do in behaviour, rather than being assumed to exhibit equivalent levels of ecological specialization. We test this hypothesis using male Urosaurus ornatus lizards that exhibit a discrete dewlap colour polymorphism that reflects alternative mating tactics. We found blue morphs specialize on prey at higher trophic levels
Article
Full-text available
Colour polymorphisms (CP's) continue to be of interest to evolutionary biologists because of their general tractability, importance in studies of selection and potential role in speciation. Since some of the earliest studies of CP, it has been evident that alternative colour morphs often differ in features other than colour. Here we review the rapidly accumulating evidence concerning the genetic mechanisms underlying correlations between CP and other traits in animals. We find that evidence for genetic correlations is now available for taxonomically diverse systems and that physical linkage and regulatory mechanisms including transcription factors, cis-regulatory elements, and hormone systems provide pathways for the ready accumulation or modification of these correlations. Moreover, physical linkage and regulatory mechanisms may both contribute to genetic correlation in some of the best-studied systems. These results raise the possibility that negative frequency-dependent selection and disruptive selection might often be acting on suites of traits and that the cumulative effects of such selection, as well as correlational selection, may be important to CP persistence and evolution. We consider additional evolutionary implications. We recommend continued efforts to elucidate the mechanisms underlying CP-correlated characters and the more frequent application of comparative approaches, looking at related species that vary in character correlations and patterns of selection. We also recommend efforts to elucidate how frequency-dependent selection may act on suites of characters.
Article
Full-text available
Based on the measurements of 1063 flower reflection spectra, we show that flower colours fall into distinct clusters in the colour space of a bee. It is demonstrated that this clustering is caused by a limited variability in the floral spectral reflectance curves. There are as few as 10 distinct types of such curves, five of which constitute 85% of all measurements. UV reflections are less frequent and always lower in intensity than reflections in other parts of the spectrum. A further cluster of colour loci is formed in the centre of the colour space. It contains the colour loci of green leaves, several other background materials and only very few flowers. We propose a system to classify the reflection functions of flowers, and a set of colour names for bee colours.
Article
Full-text available
Diurnal lizards are useful model organisms for the study of color polymorphisms. The cellular basis of nonavian reptilian coloration and dermal chromatophores, as well as the different pigments contained therein, is relatively well understood. Nonetheless, specific predictions cannot be made for the biochemical chromatophore constituents of individual species a priori. Despite well-documented associations between throat coloration and behavioral, physiological, and life history traits, the cellular basis of throat coloration in polymorphic Common Side-Blotched Lizards, Uta stansburiana, has never been studied. We used a combination of chromatographic techniques in conjunction with transmission electron microscopy and a thin-layer reflectance model to determine the cellular basis of polymorphic coloration in U. stansburiana. We found that different morphs express coloration through varying spatial arrangements of dietary xanthophylls (lutein and zeaxanthin), differential synthesis of drosopterins, differential melanin synthesis, and morph-specific iridophore reflecting platelets. Our results indicate that chromatic variations in this polymorphism cannot be attributed to alternative xanthophore pigmentation or structural reflectance alone. Instead, the chromatophores of U. stansburiana are multicomponent color signals.
Article
Evolutionary theory suggests that alternative colour morphs (i.e. genetically controlled phenotypes) may derive similar fitness under frequency-dependent selection. Here we experimentally demonstrate opposing effects of frequency-dependent social environments on plasma hormone levels (testosterone and corticosterone) and immune function between red- and black-headed male morphs of the Gouldian finch (Erythrura gouldiae). Red-headed males are highly sensitive to changes in the social environment, especially towards the relative density of their own aggressive morph, exhibiting high stress responses and immunosuppression in socially competitive environments. In contrast, the non-aggressive black-headed males follow a more passive strategy that appears to buffer them against social stresses. The differential effect of hormones on aggressive behaviour and immune performance reinforces the contrasting behavioural strategies employed by these colour morphs, and highlights the importance of the social environment in determining the individual basis of behavioural and physiological responses.
Article
All females from the Dona Ana Mountains, New Mexico, had orange throats which exhibited an insignificant tendency toward deeper orange in gravid and near gravid females. Females from Aguirre Spring, however, were polymorphic (yellow, gold, or pale orange) with respect to throat coloration early in the activity season. Later, females carrying enlarged follicles exhibited a deep orange throat color. This color change was especially significant during the first of 2 annual clutches. For the Aguirre Spring population both body size and throat color were good predictors of clutch size. -from Authors
Article
SYNOPSIS. An organism's phenotype results from an interaction of environment and genotype. Sex steroids play a role in translating sexual genotype into phenotype. The focus of this research has been to extend the model of sex steroid hormone action in sexual differentiation to individual variation in reproductive phenotype. The hypothesis generated, called the relative plasticity hypothesis, has been tested in a species with alternative phenotypes, tree lizards (Urosaurus ornatus). Such species are useful models for tests of these ideas because variation is extreme and easily studied. These tests have shown that permanent differentiation of the territorial and nonterritorial phenotypes is accomplished by hormonal mechanisms operating during early development. These are similar to organizational actions of ' hormones classically described for sexual differentiation. A unique finding of this work is that the adrenal hormone progesterone regulates differentiation of the two male types. Furthermore, the nonterritorial male tree lizard appears to conditionally switch tactics between sedentary satellite and nomadic behavior. Nomadic behavior is triggered under stressful environmental conditions. Nonterritorial tree lizards show greater inhibition of reproductive hormones following stress than do territorial tree lizards, suggesting that a differential sensitivity of reproductive hormones to stress is the endocrine basis of conditional tactic switching in this morph. This mechanism is similar to the classical activational effects of hormones. Together, these findings indicate that individual variation in sexual phenotype is produced by mechanisms similar to sexual differentiation. Further refinement of the models and integration with some evolutionary ideas is explored.