Available via license: CC BY-NC-ND 4.0
Content may be subject to copyright.
1
Information content of cheek and lip colour in relation to the timing of ovulation in women
1
Lucie Rigaill 1
2
1 Kyoto University Primate Research Institute KUPRI, Inuyama, Japan
3
lucie.rigaill@gmail.com
4
5
Abstract
6
Various animal species have evolved a sexual communication system with female displaying
7
and male discriminating information about the timing of ovulation through sexual signals. More
8
research is now investigating the potential ovulatory signalling function of female red skin
9
colour in human and non-human primates. However, to date it is still challenging to draft
10
satisfying hypotheses of the evolution and function of female red skin colour, due to
11
methodological discrepancies between human and non-human primate studies. The present
12
study used a within-individual design and objective methods to analyse the relationship
13
between fine-scale variation in cheek and lip colour (luminance and redness) and the estimated
14
day of ovulation in 15 cycling women. Lip, but not cheek, colour, appeared to contain
15
information about the timing of ovulation, with lips getting darker around the timing of
16
ovulation. This study adds to the growing evidence that female red skin colour may play a role
17
in sexual signalling in human and non-human primates but also underlines variation in trait
18
forms and functions at the species-level.
19
20
Keywords: Red skin colour, Information content, Ovulatory signalling, Lips, Humans, Primates
21
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
2
Introduction
22
Do females display visual information about their reproductive status that can modulate mate
23
attraction and mating strategies? This question is central to sexual selection theory. The number
24
of possible offspring by females is constrained by physiological factors such as menstrual
25
cyclicity, gestation and post-partum amenorrhea, and reproductive senescence. Reproduction is
26
usually more costly for females due to the production of larger gametes, extended maternal
27
care, and male monopolization [1,2]. Thus, in several species, females appear to have evolved
28
traits which are attractive to males and can act as probabilistic signals of ovulation to maximize
29
reproduction while balancing its associated costs. Across human and non-human primates
30
(hereafter, primates), there is evidence that male behaviour is modulated by female traits,
31
suggesting common evolutionary pathways and underlying mechanisms for sexual signalling
32
[3,4].
33
Among the different female traits, there is a growing interest in the potential role of red
34
skin colour in primate sexual communication. Intra-cycle variation in oestrogens induces
35
ovulation and also affects some chromatic (redness) and achromatic (luminance) parameters of
36
skin colour [5,6]. Circulating oestrogens bind to receptors in the skin, causing an increase in
37
blood flow and consequently a decrease in the perceived skin luminance (i.e., darkening of the
38
skin) [7,8]. Increase in blood flow can modulate the ratio of oxygenated/deoxygenated blood in
39
the skin vessels which may influence perceived redness. Primate studies provided evidence that
40
facial skin colour varies across the cycle in mandrills [9], informs about the probability of
41
ovulation in rhesus macaques [10,11], and simultaneously conceals ovulation and advertises
42
pregnancy in Japanese macaques [12,13]. Moreover, this colourful trait appears to be attractive
43
to males suggesting a role in mate attraction, at least in macaques [14].
44
Skin and lip colour also correlate with female attractiveness in human [15–18]. Skin
45
colour may thus be involved into human mate attraction, although the colour of the stimuli was
46
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
3
artificially manipulated in these studies which may impair the ecological validity of the results.
47
Studies of the potential signalling function of red skin colour in women yielded mixed results.
48
Oberzaucher et al. [19] described that cheeks were redder around ovulation compared to the
49
end of the cycle while Samson et al [20] did not replicate these findings. Recently, Burris et al.
50
[21] found no intra-cycle variation in cheek colour. Interpretation of these findings is
51
constrained by some methodological limitations. While most studies used photography to
52
analyse skin colour, which provides accurate measurements, they did not always correct
53
lighting conditions across photography sessions using colour standards that may have altered
54
measurements [22,23]. Human studies usually rely on a restricted data set to assess intra-cycle
55
difference: e.g. 2 samples from mid and late cycle, which underestimate or overestimate cycle
56
effects, with the exception of [21]. Finally, human studies often failed to confirm ovulation with
57
sex hormones profiles (with the exception of [21]) especially when studying perception of intra-
58
cycle variation. Taken together, these methodological limitations constrain our understanding
59
of the potential role of female red skin colour in sexual signalling in humans.
60
To determine if women display a colourful trait that may play a role in ovulatory
61
signalling, there is a fundamental need for studies of the relationship between fine-scale
62
variation in trait expression and the probability of ovulation. Such studies would benefit from
63
objective and quantitative methods, inspired by primate studies, using standardized
64
photography, regular biological sampling, and hormonal estimation of the ovulation date [23–
65
26]. Following this premise, this study aims at investigating whether cheek and lip colour
66
contain information about intra-cycle variation in the probability of ovulation in women. If
67
cheek or lip colour is related to the timing of ovulation within a cycle, I predict that cheeks or
68
lips will be darker/redder around this timing.
69
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
4
Methods
70
Ethic statement
71
This study was approved by the Human Research Ethics Committee of Kyoto University
72
Primate Research Institute (KUPRI, project number 2017-13). All participants signed a written
73
agreement concerning their participation in the study, confidentiality, and the use of their
74
photographs for research purposes and illustrations. Participants received a financial
75
compensation in return for their participation (5,000 JPY ~ 45 USD).
76
77
Participants
78
In total, 18 women participated in the study (mean age = 28.3 ± SD 4.3 years, “Asian” = 11,
79
“Caucasian” = 7, additional information on the supplementary material). All participants were
80
naturally cycling and not taking hormonal contraceptives for at least 3 months. I collected
81
digital photographs and saliva samples every two days, excluding weekends and national
82
holidays, for the duration of one complete menstrual cycle (i.e., from the beginning of their
83
menstruations until the next ones) between September and June 2018-19, to limit the effect of
84
tanning. Sampling occurred at a fixed time of the day for each participant to control for possible
85
diurnal variation in sex hormones concentration [27]. I collected a total of 204 photographs and
86
saliva samples (mean par participant = 11.3 ± SD 1.5, range = 8-13).
87
88
Assessment of cheek and lip colour
89
Participants removed face and lip make-up a minimum of 30 min before sampling. A single
90
female experimenter (LR) photographed the participants, as the sex of the photographer can
91
influence participant facial temperature and potentially skin colour [28]. Sampling was
92
conducted in a unique location, i.e., an office room with closed curtains, ceiling lights turned
93
on, and a constant 24° C temperature. Participants sat in front of a beige background and
94
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
5
adopted a neutral expression. The camera was place 2m from the participant’s chair and held at
95
the same height of the participant’s face. The experimenter used a Nikon 5000D camera with a
96
12.3 megapixel CMOS censor and a Nikon AF-S NIKKOR 18-55mm f/3.5-5.6G lens.
97
Photographs were collected in NEF format, i.e., Nikon's RAW, with the flash disabled, the
98
shutter speed and aperture size determined automatically by the camera, and a 55mm focal to
99
limit distortion [29]. The experimenter manually set the white balance using a X-Rite White
100
Balance Card (GretagMacbeth ColorChecker) and a Gretag X-Rite Color Checker (colour
101
standards). This technique standardizes photographs such that colour measurements are
102
comparable across all photographs [23–25]. Photographs were converted to uncompressed 16-
103
bit TIFF files. Cheek colour was measured from a pair of points located on the area described
104
in [21]. Lip colour was measured from the whole lip skin (supplementary material, Figure S1).
105
CIELAB values were extracted using Colourworker software (Chrometics Ltd. available at:
106
http://www.colourworker.com/) which provided luminance (L*) and red-green ratio (a*,
107
hereafter redness) values calculated from the estimated reflectance spectrum according to the
108
standard CIE (Commission Internationale de l’Éclairage) equations [30].
109
110
Determination of the estimated ovulation
111
Participants followed the sampling guideline recommended by Salimetrics (Salimetrics, LLC,
112
USA). Samples were collected during the photo session, labelled and stored within 15 min at -
113
80°C until analyses at KUPRI. Saliva samples were analysed for oestradiol (17ß-estradiol) and
114
progesterone (4-pregenene-3,20-dione) using Salimetrics enzyme immunoassays kits. The
115
oestradiol assay had a sensitivity of 0.1 pg/ml with an intra-assay coefficient of variation (CV)
116
of 6.2 % and inter-assay CV of 10.8 %. The progesterone assay had a sensitivity of 5 pg/ml
117
with an intra-assay CV of 6.6 % and inter-assay CV of 9.7 %.
118
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
6
The onset of the luteal phase was defined as the sample with a progesterone
119
concentration at least 2 standard deviations greater than the mean of the 2-3 preceding baseline
120
values [26]. Ovulation was considered to have occurred when a mid-cycle peak of oestradiol
121
was detected around the onset of the luteal phase. Because sampling occurred every 2-3 days,
122
I considered the day of peak oestradiol as the most likely day of ovulation and labelled it as day
123
0. The day directly preceding the estimated ovulation day was labelled as day -1, the day
124
directly following it as day +1, and so on. Ovulation could not be determined (i.e., abnormal
125
hormone variations) for 3 out of the 18 participants.
126
127
Statistical analyses
128
I used photographs of 15 participants showing ovulatory cycles to analyse intra-cycle variation
129
in cheek and lip luminance and redness (N = 165 photographs, mean per participant = 11.0 ±
130
SD 1.4, range = 8-13).
131
This study tests for a possible quadratic effect of the days relative to ovulation on cheek
132
and lip colour: higher effect toward the estimated ovulation and lower effect toward the
133
beginning and end of the cycle. I thus constructed general linear mixed-effects models (LMMs)
134
for cheek and lip colour (luminance and redness respectively) which included the linear (DAY)
135
and quadratic (DAY2) effects of the days relative to the estimated ovulation as fixed effects,
136
and fitted by maximum likelihood using lme4 [31] and lmerTest [32] packages in R version
137
3.6.0 [33]. Inspection of the cumulative distribution functions revealed good fits to the normal
138
distribution for luminance and to the lognormal distribution for redness. Prior to modelling, the
139
days relative to the estimated ovulation were standardized to mean = 0 and SD = 1 to improve
140
model performance and interpretability [34]. On one hand, we can expect that the relationship
141
between the timing of ovulation and colour would be similar across participants; I thus
142
constructed random intercept models (RIM) with random intercepts for participant identity. On
143
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
7
the other hand, this relationship may vary between participants; I thus constructed uncorrelated
144
random slope models (RSM) with random slopes for the linear and quadratic terms. I also
145
constructed the respective RIM and RSM null models in which the predictor variables were
146
removed but the random effect structure was maintained. This resulted in 4 candidate models
147
for each colour/skin combination: RIM-full, RIM-null, RSM-full, and RSM-null. I ensured that
148
all relevant model assumptions were met by visually inspecting histograms of the residuals and
149
plots of residuals against fitted values. I used an information-theory approach to objectively
150
compare and rank the 4 candidate models in terms of how well they fitted the existing data thus
151
assessing the likelihood that one or more models among the candidates is/are best supported by
152
the data [35,36]. I used the function model.sel of the MuMIn package [37] to rank models based
153
on the Akaike's information criterion corrected for small sample size (AICc values). I reported
154
the weight of the models which indicates to what extent one candidate model is more likely
155
than another to provide a reasonable explanation of the variance in the data. I extracted weighted
156
parameter estimates (β), standard errors (s.e.), and 95% confidence intervals (95% CIs) of
157
model intercepts and predictors from conditional averaging of the 4 candidate models (function
158
model.avg).
159
160
Results
161
RIMs generally performed better than their respective RSMs (Table 1). Concerning cheek
162
colour, the RIM-null for luminance had a weight of 0.75, while the RIM-full for redness had a
163
weight of 0.51 (vs. 0.37 for its RIM-null), suggesting some weak evidence for variation in cheek
164
redness only (Table 1). Concerning lip colour, the RIM-full for luminance had a weight of 0.53
165
(vs. 0.31 for RIM-null) and the RIM-null for redness had a weight of 0.50 (vs. 0.37 for its RIM-
166
null), suggesting some weak evidence for variation in lip luminance only (Table 1). The timing
167
of ovulation explained little to none of the variance in cheek redness (Table 2, Figure 1). Lip
168
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
8
luminance varied according to the timing of ovulation with lips being darker around ovulation
169
(ß = 0.21 ± 0.09 s.e., 95% CI = 0.03; 0.39, Table 2, Figure 2).
170
171
Table 1. Characteristics of candidate models for cheek and lip redness and luminance. AICc:
172
corrected Akaike’s information criterion, weight: model probabilities
173
logLik
AICc
weight
logLik
AICc
weight
Cheek luminance
Cheek redness
RIM - null
-247.751
501.7
0.75
RIM - full
279.799
-549.2
0.51
RIM - full
-247.292
505.0
0.14
RIM - null
277.351
-548.6
0.37
RSM - null
-247.751
505.9
0.09
RSM - full
279.799
-544.9
0.06
RSM - full
-247.292
509.3
0.02
RSM - null
277.621
-544.9
0.06
Lip luminance
Lip redness
RIM - full
-272.607
555.6
0.53
RIM-null
221.984
-437.8
0.60
RIM - null
-275.332
556.8
0.30
RIM-full
223.354
-436.3
0.29
RSM - null
-274.316
559.0
0.10
RSM-null
222.055
-433.7
0.08
RSM - full
-272.345
559.4
0.08
RSM-full
223.356
-432.0
0.03
174
Table 2. Model averaged parameters estimates (ß) ± standard errors (s.e.) and confidence
175
intervals (95% CIs) from conditional averaging of all candidate models. Result for which CI
176
does not include zero are presented in bold.
177
Intercept
DAY
DAY2
Cheek luminance
26.99 ± 0.50 (26.01; 27.97)
-0.06 ± 0.07 (-0.21; 0.08)
0.03 ± 0.07 (-0.11; 0.17)
Cheek redness
2.05 ± 0.02 (2.01; 2.10)
0.00 ± 0.00 (-0.00; 0.01)
0.01 ± 0.00 (-0.00; 0.01)
Lip luminance
20.77 ± 0.32 (20.13; 21.40)
0.01 ± 0.09 (-0.18; 0.19)
0.21 ± 0.09 (0.03; 0.39)
Lip redness
2.20 ± 0.03 (2.14; 2.27)
0.01 ± 0.00 (-0.00; 0.01)
0.00 ± 0.00 (-0.00; 0.01)
178
Figure 1. Relationship between cheek (a) luminance and (b) redness and the days relative to
179
the estimated ovulation. Raw data are presented as grey circle. The red line presents the global
180
model prediction; the black lines show prediction for each participant.
181
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
9
182
183
Figure 2. Relationship between lip (a) luminance and (b) redness and the days relative to the
184
estimated ovulation. Raw data are presented as grey circle. The red line presents the global
185
model prediction; the black lines show prediction for each participant.
186
187
188
Discussion
189
Using objective methods to analyse fine-scale and intra-individual changes in cheek and lips
190
colour according to the timing of ovulation, I found that lip luminance contains information
191
about the timing of ovulation in women. Cheek colour may not be related to the timing of
192
ovulation in agreement with a previous study [21].
193
Different facial features may respond differently to variation in the probability of
194
ovulation. Intra-cycle variation in colour may be less cryptic for the lip skin as a result of its
195
higher vascularization and the finer cellular layers [38] or due to tissue-specific differences in
196
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
10
oestrogen receptor expression or sensitivity. The latter has been already suggested to explain
197
the differential effect of cycle phase on facial and hindquarters colour in Japanese macaques
198
[13]. Further studies on tissue-specific expression and sensitivity of oestrogen receptors would
199
help to clarify this question. It is also possible that hormonal contraceptives have on-going and
200
long-term effect on red skin colour (e.g., female odours [39,40]), a question that I could not
201
assess in the present study and requires further investigation. Alternatively, colour changes in
202
the facial skin may be more condition-dependent and related to individual health thus masking
203
a potential cycle effect [15,16,41].
204
So, can lip colour act as an ovulatory signal? The present study suggest that lip colour
205
contains information about the timing of ovulation. To determine whether this information is
206
conveyed, further studies should assess men and women responses toward intra-cycle variation
207
in lip colour to determine whether these changes are perceptible. However, caution should be
208
used when designing such experiments. Studies should use an intra-individual design, i.e., using
209
individual pictures rather than pictures from different participants or composite stimuli, as the
210
differences in skin colour appeared to be relatively greater between than within participant in
211
the present study. Studies should also collect and test multiple samples per individual in order
212
to detect fine-scale variation, and confirm ovulation using hormonal data or ovulation detection
213
kit. Such experimental design should reduce the risk of over- or under-estimating a potential
214
cycle effect. However, whether variation in this trait is detectable under “optimal” conditions,
215
i.e., artificial environments and controlled settings, does not necessarily entail consequences on
216
human sexual communication and mate preferences since laboratory studies can have low
217
ecological validity. At best, studies on perception could provide further evidence that lip and/or
218
cheek colour has the potential to play a role in ovulatory signalling.
219
Humans may have inherited the biological bases for female red skin colour and sexual
220
signalling from a primate ancestor. Female red skin colour has been suggested to play a role
221
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
11
into the sexual communication of some primate species, i.e., signalling [9–11] or concealing
222
ovulation [13], advertising pregnancy [9,12]. The differences in red skin traits (face, lips, and
223
sexual skin) and signalling functions probably result from the different socio-environmental
224
constraints on mating across species. Concealing the reproductive status may have evolved in
225
species facing higher male monopolization or pair-bonding, lower intra-sexual competition, and
226
higher costs on signalling; while higher infanticide risks and intra-sexual competition, limited
227
mating opportunities, and lower costs on signalling may have favoured exaggerated or
228
multimodal signalling [42,43]. However, these proposed effects do not appear to fully explain
229
the inter-species differences observed in female red skin colour function, as species with similar
230
socio-ecology express traits that may be involved in either ovulatory signalling or concealing
231
(e.g. [11,13]). More studies using longitudinal and ecologically valid designs are needed to
232
better understand the possible evolutionary pathways and underlying mechanisms leading to
233
the evolution of female red skin colour in primate species including humans.
234
235
Acknowledgments
236
I acknowledge the Human Research Ethics Committee of Kyoto University Primate Research
237
Institute for permission to conduct my research. I would like to sincerely thank C. Neumann
238
for his insightful comments on the study and his valuable advice with statistical analyses, J.
239
Duboscq and C. Garcia for comments on an earlier version of the manuscript, and K. Mouri for
240
her general support in the lab.
241
242
Funding
243
LR is funded by the Japanese Society for the Promotion of Science (project 18H05810 JSPS).
244
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
12
References
245
1. Galef BG. 1983 Costs and Benefits of Mammalian Reproduction. In Symbiosis in
246
Parent-Offspring Interactions (eds LA Rosenblum, H Moltz), pp. 249–278. Boston, MA:
247
Springer US. (doi:10.1007/978-1-4684-4565-7_11)
248
2. Gittleman JL, Thompson SD. 1988 Energy Allocation in Mammalian Reproduction.
249
Integr Comp Biol 28, 863–875. (doi:10.1093/icb/28.3.863)
250
3. Campbell C. 2007 Primate sexuality and reproduction. In Primates in perspective, pp.
251
423–437. New York: Campbell CJ, Fuentes A, MacKinnon KC, Panger M, Bearder SK.
252
4. Dixson AF. 2012 Primate sexuality: comparative studies of the prosimians, monkeys,
253
apes and human beings. 2nd edition. Oxford University Press. New York.
254
5. Rhodes L, Argersinger ME, Gantert LT, Friscino BH, Hom G, Pikounis B, Hess DL,
255
Rhodes WL. 1997 Effects of administration of testosterone, dihydrotestosterone, oestrogen
256
and fadrozole, an aromatase inhibitor, on sex skin colour in intact male rhesus macaques.
257
Journal of Reproduction and Fertility 111, 51–57. (doi:10.1530/jrf.0.1110051)
258
6. Thornton MJ. 2002 The biological actions of estrogens on skin. Experimental
259
Dermatology 11, 487–502. (doi:10.1034/j.1600-0625.2002.110601.x)
260
7. Pelletier G, Ren L. 2004 Localization of sex steroid receptors in human skin. Histol.
261
Histopathol. 19, 629–636. (doi:10.14670/HH-19.629)
262
8. Thornton Mj, Taylor AH, Mulligan K, Al-Azzawi F, Lyon CC, O’Driscoll J,
263
Messenger AG. 2003 The Distribution of Estrogen Receptor β Is Distinct to That of Estrogen
264
Receptor α and the Androgen Receptor in Human Skin and the Pilosebaceous Unit. Journal of
265
Investigative Dermatology Symposium Proceedings 8, 100–103. (doi:10.1046/j.1523-
266
1747.2003.12181.x)
267
9. Setchell JM, Wickings EJ, Knapp LA. 2006 Signal content of red facial coloration in
268
female mandrills (Mandrillus sphinx). Proceedings of the Royal Society B: Biological
269
Sciences 273, 2395–2400. (doi:10.1098/rspb.2006.3573)
270
10. Dubuc C, Brent LJN, Accamando AK, Gerald MS, MacLarnon A, Semple S,
271
Heistermann M, Engelhardt A. 2009 Sexual skin color contains information about the timing
272
of the fertile phase in free-ranging Macaca mulatta. International Journal of Primatology 30,
273
777–789. (doi:10.1007/s10764-009-9369-7)
274
11. Higham JP, Brent LJN, Dubuc C, Accamando AK, Engelhardt A, Gerald MS,
275
Heistermann M, Stevens M. 2010 Color signal information content and the eye of the
276
beholder: a case study in the rhesus macaque. Behavioral Ecology 21, 739–746.
277
(doi:10.1093/beheco/arq047)
278
12. Rigaill L, MacIntosh AJJ, Higham JP, Winters S, Shimizu K, Mouri K, Furuichi T,
279
Garcia C. 2015 Multimodal advertisement of pregnancy in free-ranging female Japanese
280
macaques (Macaca fuscata). PLOS ONE 10, e0135127. (doi:10.1371/journal.pone.0135127)
281
13. Rigaill L, Higham JP, Winters S, Garcia C. 2019 The redder the better? Information
282
content of red skin coloration in female Japanese macaques. Behavioral Ecology and
283
Sociobiology 73, 103. (doi:10.1007/s00265-019-2712-x)
284
14. Higham JP, Winters S. 2015 Color and mate choice in non-human animals. In
285
Handbook of Color Psychology, pp. 502–530. Cambridge: A. Elliot, M. Fairchild, & A.
286
Franklin.
287
15. Re DE, Whitehead RD, Xiao D, Perrett DI. 2011 Oxygenated-Blood Colour Change
288
Thresholds for Perceived Facial Redness, Health, and Attractiveness. PLOS ONE 6, e17859.
289
(doi:10.1371/journal.pone.0017859)
290
16. Pazda AD, Thorstenson CA, Elliot AJ, Perrett DI. 2016 Women’s Facial Redness
291
Increases Their Perceived Attractiveness: Mediation Through Perceived Healthiness.
292
Perception 45, 739–754. (doi:10.1177/0301006616633386)
293
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
13
17. Young SG. 2015 The effect of red on male perceptions of female attractiveness:
294
Moderation by baseline attractiveness of female faces. European Journal of Social
295
Psychology 45, 146–151. (doi:10.1002/ejsp.2098)
296
18. Stephen ID, McKeegan AM. 2010 Lip colour affects perceived sex typicality and
297
attractiveness of human faces. Perception 39, 1104–1110. (doi:10.1068/p6730)
298
19. Oberzaucher E, Katina S, Schmehl SF, Holzleitner IJ, Mehu-Blantar I, Grammer K.
299
2012 The myth of hidden ovulation: Shape and texture changes in the face during the
300
menstrual cycle. Journal of Evolutionary Psychology 10, 163–175.
301
(doi:10.1556/JEP.10.2012.4.1)
302
20. Samson N, Fink B, Matts P. 2011 Does a woman’s skin color indicate her fertility
303
level? Preliminary findings. Swiss Journal of Psychology 70, 199–202. (doi:10.1024/1421-
304
0185/a000057)
305
21. Burriss RP, Troscianko J, Lovell PG, Fulford AJC, Stevens M, Quigley R, Payne J,
306
Saxton TK, Rowland HM. 2015 Changes in Women’s Facial Skin Color over the Ovulatory
307
Cycle are Not Detectable by the Human Visual System. PLOS ONE 10, e0130093.
308
(doi:10.1371/journal.pone.0130093)
309
22. Rowland HM, Burriss RP. 2017 Human colour in mate choice and competition.
310
Philosophical Transactions of the Royal Society B: Biological Sciences 372.
311
(doi:10.1098/rstb.2016.0350)
312
23. White T, Dalrymple R, Noble D, O’Hanlon J, Zurek D, Umbers K. 2015 Reproducible
313
research in the study of biological coloration. Faculty of Science, Medicine and Health -
314
Papers , 51–57. (doi:10.1016/j.anbehav.2015.05.007)
315
24. Bergman TJ, Beehner JC. 2008 A simple method for measuring colour in wild
316
animals: validation and use on chest patch colour in geladas (Theropithecus gelada).
317
Biological Journal of the Linnean Society 94, 231–240. (doi:10.1111/j.1095-
318
8312.2008.00981.x)
319
25. Stevens M, Stoddard MC, Higham JP. 2009 Studying primate color: towards visual
320
system-dependent methods. International Journal of Primatology 30, 893–917.
321
(doi:10.1007/s10764-009-9356-z)
322
26. Hodges JK, Heistermann M. 2011 Field endocrinology: monitoring hormonal changes
323
in free-ranging primates. In Field and laboratory methods in primatology: a practical guide,
324
Joanna M. Setchell and Deborah J. Curtis.
325
27. Bao AM, Liu RY, Someren E van, Hofman MA, Cao YX, Zhou JN. 2003 Diurnal
326
rhythm of free estradiol during the menstrual cycle. European Journal of Endocrinology 148,
327
227–232. (doi:10.1530/eje.0.1480227)
328
28. Hahn AC, Whitehead RD, Albrecht M, Lefevre CE, Perrett DI. 2012 Hot or not?
329
Thermal reactions to social contact. Biology Letters , rsbl20120338.
330
(doi:10.1098/rsbl.2012.0338)
331
29. Třebický V, Fialová J, Kleisner K, Havlíček J. 2016 Focal Length Affects Depicted
332
Shape and Perception of Facial Images. PLOS ONE 11, e0149313.
333
(doi:10.1371/journal.pone.0149313)
334
30. Vorobyev M, Gumbert A, Kunze J, Giurfa M, Menzel R. 1997 Flowers Through
335
Insect Eyes. Israel Journal of Plant Sciences 45, 93–101.
336
(doi:10.1080/07929978.1997.10676676)
337
31. Bates D, Mächler M, Bolker B, Walker S. 2015 Fitting linear mixed-effects models
338
using lme4. Journal of Statistical Software 67. (doi:10.18637/jss.v067.i01)
339
32. Kuznetsova A, Brockhoff PB, Christensen RHB. 2017 lmerTest Package: Tests in
340
Linear Mixed Effects Models. Journal of Statistical Software 82, 1–26.
341
(doi:10.18637/jss.v082.i13)
342
33. R Core Team. 2019 R: A language and environment for statistical computing. See
343
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
14
https://www.R-project.org/.
344
34. Harrison XA, Donaldson L, Correa-Cano ME, Evans J, Fisher DN, Goodwin CED,
345
Robinson BS, Hodgson DJ, Inger R. 2018 A brief introduction to mixed effects modelling and
346
multi-model inference in ecology. PeerJ 6, e4794. (doi:10.7717/peerj.4794)
347
35. Burnham KP, Anderson DR, Huyvaert KP. 2011 AIC model selection and multimodel
348
inference in behavioral ecology: some background, observations, and comparisons. Behav
349
Ecol Sociobiol 65, 23–35. (doi:10.1007/s00265-010-1029-6)
350
36. Symonds MRE, Moussalli A. 2011 A brief guide to model selection, multimodel
351
inference and model averaging in behavioural ecology using Akaike’s information criterion.
352
Behav Ecol Sociobiol 65, 13–21. (doi:10.1007/s00265-010-1037-6)
353
37. Bartoń K. 2017 MuMIn: Multi-Model Inference. R package version 1.40.0.
354
38. Crouzet C, Fournier H, Papon X, Hentati N, Cronier P, Mercier Ph. 1998 Anatomy of
355
the arterial vascularization of the lips. Surgical and Radiologic Anatomy 20, 273–278.
356
(doi:10.1007/BF01628490)
357
39. Crawford JC, Boulet M, Drea CM. 2010 Smelling wrong: hormonal contraception in
358
lemurs alters critical female odour cues. Proceedings of the Royal Society of London B:
359
Biological Sciences , rspb20101203. (doi:10.1098/rspb.2010.1203)
360
40. Kuukasjärvi S, Eriksson CJP, Koskela E, Mappes T, Nissinen K, Rantala MJ. 2004
361
Attractiveness of women’s body odors over the menstrual cycle: the role of oral
362
contraceptives and receiver sex. Behavioral Ecology 15, 579–584.
363
(doi:10.1093/beheco/arh050)
364
41. Thorstenson CA, Pazda AD, Elliot AJ, Perrett DI. 2017 Facial Redness Increases
365
Men’s Perceived Healthiness and Attractiveness. Perception 46, 650–664.
366
(doi:10.1177/0301006616680124)
367
42. Rooker K, Gavrilets S. 2018 On the evolution of visual female sexual signalling.
368
Proceedings of the Royal Society B: Biological Sciences 285, 20172875.
369
(doi:10.1098/rspb.2017.2875)
370
43. Fitzpatrick CL, Servedio MR. 2018 The evolution of male mate choice and female
371
ornamentation: a review of mathematical models. Curr Zool 64, 323–333.
372
(doi:10.1093/cz/zoy029)
373
374
.CC-BY-NC-ND 4.0 International licenseIt is made available under a
(which was not peer-reviewed) is the author/funder, who has granted bioRxiv a license to display the preprint in perpetuity.
The copyright holder for this preprint. http://dx.doi.org/10.1101/859801doi: bioRxiv preprint first posted online Nov. 29, 2019;
