Accepted Manuscript

Surface temperature of ewes during estrous cycle measured by infrared

thermography

Ana Carolina Barros de Freitas, Wilder Hernando Ortiz Vega, Celia Raquel Quirino,
Aylton Bartholazzi Junior, Caroline Marçal Gomes David, André Torres Geraldo,
Miguel Alejandro Silva Rua, Luisa Fernanda Cipagauta Rojas, Janeo Eustáquio de
Almeida Filho, Angelo José Burla Dias

PII: S0093-691X(18)30512-0
DOI: 10.1016/j.theriogenology.2018.07.015
Reference: THE 14629

To appear in: Theriogenology

Received Date: 1 March 2018

Revised Date: 8 July 2018
Accepted Date: 16 July 2018

Please cite this article as: Barros de Freitas AC, Ortiz Vega WH, Quirino CR, Junior AB, Gomes David
CarolineMarç, Geraldo AndréTorres, Silva Rua MA, Cipagauta Rojas LF, Eustáquio de Almeida Filho
J, Burla Dias AngeloJosé, Surface temperature of ewes during estrous cycle measured by infrared
thermography, Theriogenology (2018), doi: 10.1016/j.theriogenology.2018.07.015.

This is a PDF file of an unedited manuscript that has been accepted for publication. As a service to
our customers we are providing this early version of the manuscript. The manuscript will undergo
copyediting, typesetting, and review of the resulting proof before it is published in its final form. Please
note that during the production process errors may be discovered which could affect the content, and all
legal disclaimers that apply to the journal pertain.
 ACCEPTED MANUSCRIPT
1 “Revised”

2 Surface temperature of ewes during estrous cycle measured by infrared

3 thermography

5 Ana Carolina Barros de Freitasa,*, Wilder Hernando Ortiz Vegaa, Celia Raquel

PT
6 Quirinoa, Aylton Bartholazzi Juniora, Caroline Marçal Gomes Davida, André Torres

7 Geraldoa, Miguel Alejandro Silva Ruaa, Luisa Fernanda Cipagauta Rojasb, Janeo

RI
8 Eustáquio de Almeida Filhoc, Angelo José Burla Diasa

SC
9
a
10 Laboratory for Animal Breeding and Genetic Improvement, Norte Fluminense State

11
U
University, Av. Alberto Lamego, 2000, CEP: 28013-602, Campos dos Goytacazes,
AN
12 Rio de Janeiro, Brazil. E-mail: aninhabdf@hotmail.com, wilortvet@yahoo.es,

13 crqster@gmail.com, carolinedavid.mg@gmail.com, andre_atg@hotmail.com,

M

14 miguelvet-rua@gmail.com, junior_barth@hotmail.com, angeloburlauenf@gmail.com

D

b
15 Medicina Veterinaria y Zootecnia, Universidad de los Llanos, Km. 12 Vía Puerto
TE

16 López - PBX. 6616800, Villavicencio, Meta, Colombia. E-mail:

17 luisa.cipagauta@unillanos.edu.co
EP

c
18 Laboratory for Vegetal Genetic Improvement, Norte Fluminense State University,

19 Av. Alberto Lamego, 2000, CEP: 28013-602, Campos dos Goytacazes, Rio de
C

20 Janeiro, Brazil. E-mail: janeoeaf@gmail.com

AC

21

22 *Corresponding author: Ana Carolina Barros de Freitas. E-mail:

23 aninhabdf@hotmail.com

24

25

1
 ACCEPTED MANUSCRIPT
26 Abstract

27

28 The use of infrared thermography has not been used previously to assist sheep

29 reproduction. The aim of this study was evaluate the viability of infrared

30 thermography to identify body surface temperature patterns during ewes’ estrous

PT
31 cycle. The estrous cycles of 20 Santa Ines ewes were synchronized through

RI
32 hormone treatment (progesterone implant). The ewes were submitted to

33 measurements of body temperature (rectal and vaginal) with a digital thermometer

SC
34 and surface temperature of the anus, vulva, muzzle, left ear and left eye by infrared

35 thermography. Temperature and humidity of the environment and wet-bulb globe

36
U
temperature (WBGT) were measured. After the progesterone implant’s removal, the
AN
37 follicular dynamic was evaluated daily during five days with ultrasound. The estrous
M

38 cycle was separated into six phases to facilitate interpretation of the differences in

39 body temperatures in each period. Phase 1 was the period before estrous
D

40 synchronization, phase 2 was the beginning of estrous synchronization, phase 3 was

TE

41 the intermediate period and end of synchronization, phase 4 was the period of estrus

42 manifestations, phase 5 was the ovulation period, and phase 6 was the post-
EP

43 ovulatory period. There was a difference among phases for all body regions

44 evaluated (P<0.05). All temperatures were higher in the ovulation phase. The rectal
C

45 and vaginal temperatures measured by digital thermometer were lower during the
AC

46 diestrus period and increased during and after ovulation. The surface temperature of

47 the vulva, measured by thermography, was higher from the estrus phase (4) to the

48 ovulation and post-ovulation phases (5 and 6). The surface temperature of the

49 muzzle was higher from the ovulation phase (5) to the post-ovulation phase (6),

50 showing potential to detect ovulation. The thermography was efficient to detect small

2
 ACCEPTED MANUSCRIPT
51 temperature variations during different phases of the estrous cycle, allowing the

52 identification of different phases of the cycle in Santa Ines ewes.

53 Keywords: estrus, Ovis aries, ovulation, reproduction, ultrasound

54

55 1. Introduction

PT
56

RI
57 Infrared thermography (IRT) is a safe and noninvasive technique that can

58 measure the surface temperature of animals with precision [1]. Among other

SC
59 applications, it can be useful to detect estrus in animals [2], but the few studies that

60 have been conducted in cows and sows used different methodologies and showed

61
U
conflicting results. This estrus detection through temperature is possible due to the
AN
62 increase of blood circulation during estrus and consequent increase of vulva surface
M

63 temperature [3].

64 For successful mating, both males and females should express adequate
D

65 levels of sexual interaction (receptivity in female and libido in male) [4]. Traditionally,
TE

66 ram (or teaser ram) is used to detect ewes’ estrous, because it is subtle and hard to

67 detect otherwise [5], but the time necessary for observation can be labor-intensive
EP

68 for breeders. IRT can replace the ram or teaser ram for estrus detection, because it

69 is inexpensive, fast, efficient and provides important information without need for
C

70 physical contact with the animals [6]. Furthermore, the desire to mate is highly
AC

71 variable among individuals and the reproductive activity of rams is influenced by a

72 large number of external factors (e.g., photoperiod, weather conditions, social

73 interaction, nutritional cues, number of ewes, early experience, size and familiarity of

74 the paddock, stress and disease) [7].

3
 ACCEPTED MANUSCRIPT
75 All objects with temperature above -273.15 oC emit electromagnetic radiation,

76 related to surface temperature, located in the infrared part of the light spectrum [8],

77 which is invisible to the human eye [9]. Thermography allows thermal mapping of the

78 body from the infrared radiation emitted by the surface, forming an image that can be

79 seen by human eyes, providing information about temperature through visible colors

PT
80 [9]. The thermographic camera performs automatic calibration and the user selects

RI
81 the image, focuses and pulls the trigger to record it. The images are then evaluated

82 by software and the minimum, maximum and mean temperatures can be measured

SC
83 in every spot of the image.

84 Nowadays, thermographic cameras are reasonably priced, and besides

85
U
detecting temperature variations related to estrus, they can be used in sheep farming
AN
86 for early diagnoses of pathologies and lesions (e.g., mastitis, lameness, problems
M

87 with blood circulation or inflammatory responses), thermal physiology evaluation

88 (e.g., fever), counting of animal population and analysis of welfare (e.g., responses to
D

89 stress) [10]. Although this technology is very sensitive to environmental fluctuations

TE

90 [11], the environmental effects can be minimized by inserting the ambient

91 temperature and humidity in the camera before the images are obtained, in the
EP

92 software during image data analysis, or by considering these factors in the statistical

93 model.
C

94 The use of infrared thermography has not been used previously to assist
AC

95 sheep reproduction, so the aim of the present study was evaluate the viability of

96 infrared thermography to identify body surface temperature patterns during ewes’

97 estrous cycle.

98

99 2. Material and methods

4
 ACCEPTED MANUSCRIPT
100

101 The present study was approved by the university’s animal use ethics

102 committee in accordance with the Brazilian Society of Laboratory Science

103 Laboratory/Brazilian College of Animal Experimentation (SBCAL/COBEA), under

104 protocol number 454802, 2016. The experiment was conducted in February and

PT
105 March 2016 at Norte Fluminense State University (UENF), located in the city of

RI
106 Campos dos Goytacazes, Rio de Janeiro, Brazil (latitude 21º 45’ 15” S and longitude

107 41º 19' 28" W). In this region, February and March correspond to summer and the

SC
108 average of temperature, humidity and rainfall from 2011 to 2016 were 28.19 oC,

109 73.16% and 98.8 mm respectively.

110
U
Ewes of the Santa Ines hair breed (n = 20) were evaluated, with at least one
AN
111 parturition in their reproductive lives. The ewes were kept in natural pasture
M

112 (Cynodon spp.) and mineral salt and water were available ad libitum. During the

113 experiment, we measured and evaluated: body temperature; environmental

D

114 conditions (temperature and humidity, wet-bulb globe temperature – WBGT); ovarian
TE

115 follicular dynamic and estrous behavior (Figure 1).

116 The estrous cycle of the ewes was divided into 6 phases in order to facilitate
EP

117 the statistical analysis of temperatures in each period of the cycle, as described

118 below: Phase 1 (D-20 to D-16), data collection before estrus synchronization; Phase
C

119 2 (D-15 to D-12), uniformity of the estrous cycle; Phase 3 (D-14 to D-4), growth of
AC

120 pre-ovulatory follicles until sponge removal; Phase 4 (D-3 to D-1), ewes’ estrus

121 manifestation; Phase 5 (D0), ewes’ ovulation; Phase 6 (D5), ewes post-ovulatory

122 period.

123

124 2.1 Body and environmental temperatures

5
 ACCEPTED MANUSCRIPT
125 The body temperatures were measured twice daily (8:00 and 14:00 hours)

126 during 23 consecutive days (D -20 to D2) in a covered rectangular corral (6 m length,

127 2 m width and fence 3 m high) with the animals restrained (Figure 1). The ewes were

128 placed individually in the corral 30 minutes before the temperature measurements to

129 avoid exposure to the sun and consequently changes in temperature. The rectal and

PT
130 vaginal temperatures were measured with a Med® digital thermometer (Cotronic

RI
131 Technology Ltd., Shatin, Hong Kong).

132 The surface temperature of the anus, vulva, muzzle, left eye and left ear were

SC
133 measured with an infrared thermal camera (FLIR® series i50; FLIR Systems Co. Ltd.,

134 Shatin, Hong Kong). The images were captured at a distance of 1 m from the animal,

135
U
with emissivity of 0.98 [2] and rainbow palette. The camera was inclined so the image
AN
136 could be best visualized on the monitor. The perineal region was gently wiped with
M

137 dry paper towels to remove fecal matter and moisture. The images were transferred

138 to a computer and analyzed with the Flir Quickreport® version 1.2 SP2 (1.0.1.217)
D

139 2009 software. In the program, we selected the assessment regions (around the
TE

140 anus and in the central part of the vulva, muzzle, left eye and left ear) with the area

141 tool, represented by the smallest square (Figure 2). This tool identifies maximum,
EP

142 minimum and average temperature for the selected thermogram area. The mean

143 temperature was analyzed in each anatomical region, in each image captured.
C

144 During the evaluation of the images with the software, it was necessary to
AC

145 change the color palette (initially the rainbow palette), to sharpen the definition of

146 each structure. For evaluation of the anus and vulva, the “rain hi” palette was used;

147 for evaluation of the muzzle, “inverted gray” was used, for evaluation of the eye,

148 “gray” was used, and for the ear, “inverted gray” was used. The saturation of the

6
 ACCEPTED MANUSCRIPT
149 images was changed several times in order to make some areas more evident. The

150 changes in the color palettes did not alter the temperature readings.

151 The environmental variables (temperature, humidity and wet-bulb globe

152 temperature – WBGT) were recorded daily with the use of an Extech® HT30 meter

153 (Nashua, New Hampshire, USA). The instrument was always used inside the corral

PT
154 where the ewes were evaluated, without direct exposure to the sun. The

RI
155 measurements were performed from D -20 to D2, at 8:00 a.m., 9:00 a.m., 2:00 p.m.

156 and 3:00 p.m. The body temperatures were measured in the morning and afternoon,

SC
157 between the moments of measuring the environmental variables.

158

159
U
2.2 Hormonal treatment for estrous synchronization and ovulation
AN
160 To standardize the different phases of the estrous cycle, synchronization of
M

161 estrus and ovulation of all animals was performed. The hormonal treatment was

162 performed according to Aisen [12]. On day -15 (D -15), the sponge containing
D

163 medroxyprogesterone acetate (Progespon® for sheep, Zoetis, Cotia, São Paulo,
TE

164 Brazil) was introduced inside the vagina. On day -4 (D -4) the sponge was removed

165 and 300 UI of eCG (Sincro eCG®, Ourofino, Cravinhos, São Paulo, Brazil) and 1 ml
EP

166 de PGF2ɑ (Lutalyse®, Zoetis, Cotia, São Paulo, Brazil) were intramuscularly injected

167 in each ewe.

C

168
AC

169 2.3 Detection of estrus and ovarian evaluation by transrectal ultrasound

170 Estrus was investigated in the corral from D -3 to D0 (Phases 4 and 5, Figure

171 1), twice daily (9:00 a.m. and 6:00 p.m.), with the presence of a teaser ram. Ewes

172 that accepted being mounted were classified as in estrus [5]. Ovaries were evaluated

173 by transrectal ultrasound once daily (Phases 4, 5 and 6) using a Doppler ultrasound

7
 ACCEPTED MANUSCRIPT
174 device (Z6 Vet Diagnostic Ultrasound System - Henzhen Mindray Bio-Medical

175 Electronics Co., Ltd., Mindray Building, Nanshan, Shenzhen, China), at frequency 7

176 MHz.

177 To identify the phase corresponding to ovulation, follicular growth was

178 evaluated once a day (phases 4 and 5, Figure 1) until the dominant follicles identified

PT
179 reduce in size or no longer be detected in the ultrasound evaluation. We did not

RI
180 evaluate four ewes by transrectal ultrasound (from 20 used in the study), to avoid the

181 influence of manipulation on the rectal temperature. On D5, ultrasonography of all

SC
182 ewes (n = 20) was performed to confirm ovulation by the presence of a corpus

183 luteum (CL). The ewes were restrained according to [13] and the ovary observation

184
U
was performed according to Ginther and Kot [14] and Balaro et al. [15].
AN
185
M

186 2.4 Statistical analysis of results

187 The temperatures measured by the digital thermometer and the infrared
D

188 camera were fitted considering the following model:

TE

189
EP

190 Where: y is the vector with the phenotypes of a given temperature, µ is the intercept

191 of the model; βI is the coefficient of regression adjusted as covariable i (xi), centered
C

192 in the mean ( ), considering the covariables temperature, humidity and wet-bulb
AC

193 globe temperature; g is the random effect referring to the ewes, where

194 and the kinship matrix is an identity matrix (I), because the ewes were not related;

195 and e is the effect of the random error, assumed to be . The variance

196 components were estimated by restricted maximum likelihood (REML), and fixed

8
 ACCEPTED MANUSCRIPT
197 effect parameters by generalized least squares (GLS) using the ASReml package

198 [16] of R [17].

199 The variables were corrected (ycorr) according to the expression:

200

PT
201 The period of the experiment was separated into phases (Figure 1) to facilitate

202 interpretation of the differences in body temperatures (digital thermometer) and

RI
203 surface temperatures (infrared thermography) in each period of the estrous cycle:

SC
204 - Phase 1: days -20 to -16, before estrous synchronization.

205 - Phase 2: days -15 to -12, beginning of the estrous synchronization.

206 -
U
Phase 3: days -11 to -4, intermediate period and end of synchronization,
AN
207 during the growth of pre-ovulatory follicles until sponge removal.

208 - Phase 4: days -3 to -1, estrus manifestation.

M

209 - Phase 5: day 0, ovulation.

D

210 - Phase 6: day 5, post-ovulatory period.

TE

211 Analysis of variance of temperatures was performed (PROC GLM, SAS). The

212 model was:

EP

213 Yijl = µ + Idi + ECCj + (ECC Id)ji + eijl

214 Where: Yijl= dependent variable; µ = general mean, associated with the
C

215 dependent variable; Idi = effect of age i; ECCj= effect of the body score j; (ECC Id)ji =
AC

216 effect of the interaction of body score and age ji; eijl = random residual associated

217 with each observation . Means were compared by the SNK test with 5%

218 probability. Correlation was performed by CORR of the SAS software [18].

219 A sensitivity test was performed to identify estrus through the vulvar IRT,

220 considering two IRT temperature increases: 0.6 °C and 1.0 °C. If the ewe presented

9
 ACCEPTED MANUSCRIPT
221 vulvar temperature rise between one day previous to estrus (D-4) and during the

222 period of estrus (D-3 and D-1) and the remained high for at least two subsequent

223 evaluations (each 12 h), the prediction result was considered true positive. When

224 temperature was no longer observed or the temperature did not remain high for the

225 next two evaluations, the prediction result was considered false negative, according

PT
226 to Talukder et al. [19]. Sensitivity was defined as the proportion of estrus

RI
227 manifestation period in which the IRT estrus alert occurred. It was calculated by the

228 formula: true positive/(true positive + false negative).

SC
229

230 3. Results

231
U
Each anatomical region was measured 2 times/day/animal, during 23
AN
232 consecutive days. Considering that we used 20 animals and measured 7 anatomical
M

233 regions, a total of 6440 measurements were performed in the study, with 920 in each

234 anatomical region (rectum, vagina, anus, vulva, muzzle, left eye and left ear). Each
D

235 environmental variable (ambient temperature, relative humidity and WBGT) was
TE

236 measured 82 times during the study. The mean and standard deviation of

237 environmental variables were 28.89±2.12 °C for ambient temperature, 74.18±9.44%

EP

238 for relative humidity and 26.27±1.3 °C for WBGT. All the ewes responded to the

239 synchronization protocol used, so they remained in all phases of the study.
C

240 The mean body condition score was 3.5. The estrus manifestations were
AC

241 observed from 24 hours after the sponge removal and lasted about 12 to 36 hours

242 (mean 25.9 h and SD 8 h). This result was confirmed by mating behavior

243 (acceptance of mating attempt by the teaser ram) and by the ultrasound evaluation of

244 follicular development, ovulation and presence of CL. A total of 28 pre-ovulatory

245 follicles (mean size 0.70 ± 0.12 cm) were identified by ultrasonography in the 16

10
 ACCEPTED MANUSCRIPT
246 ewes evaluated during the phases 4 and 5. However, during the phase 6 (5 days

247 after ovulation), a total of 33 corpus luteum (mean size 1.05 ± 0.22 cm) were

248 confirmed in all ewes of the study, because the four ewes used as temperature

249 control during follicular growth and ovulation also presented CL on that ultrasound

250 evaluation.

PT
251 The biological relevance of IRT is mainly due to the differences in

RI
252 temperatures observed in the diestrus period (phases 1, 2 and 3) related to phase 4

253 (estrus) or 5 (ovulation) in vulva and muzzle anatomical regions, respectively (Table

SC
254 1). Figure 3 presents a graphical representation of the body and surface

255 temperatures in relation to the days of the experiment. The body temperature (digital

256
U
thermometer) was higher than the surface temperature (infrared thermography) in all
AN
257 measurements. The muzzle and ear temperatures were the ones with the greatest
M

258 amplitude, with averages above 1 °C between the phases. In all cases there was an

259 increase in temperature from phase 4 (D -3), equivalent to one day after sponge
D

260 removal and the estrus period. The highest estrous cycle temperatures occurred
TE

261 during estrus or during ovulation, depending on the instrument used and the

262 anatomical region measured (Table 1).

EP

263

264 Table 1: Mean, standard deviation (SD), amplitudes of temperatures in parentheses

C

265 (max-min), and P-value of body and surface temperatures in sheep in the different
AC

266 phases.

Temp. Phase 1 Phase 2 Phase 3 Phase 4 Phase 5 Phase 6

P
°C Mean±SD Mean±SD Mean±SD Mean±SD Mean±SD Mean±SD
(Max-Min) (Max-Min) (Max-Min) (Max-Min) (Max-Min) (Max-Min)
Rectum 38.6±0.3b 38.4±0.3b 38.5±0.3b b
38.6±0.3
a
38.7±0.2
a
38.8±0.2 <.001
(DT) (37.9-39.1) (37.4-39.6) (37.2-39.3) (38.1-39.5) (38.2-39.2) (38.6-39.2)
Vagina 38.6±0.3b 38.5±0.3b 38.5±0.3b b
38.6±0.4
a
38.7±0.2
a
38.9±0.2 0.002
(DT) (38.0-39.4) (37.6-39.8) (37.6-39.4) (37.9-39.2) (38.2-39.3) (38.6-39.3)

11
 ACCEPTED MANUSCRIPT
bc c b ab a a
Anus 37.5±0.7 37.5±0.7 37.6±0.7 37.9±0.5 38.1±0.6 38.1±0.4 <.001
(IRT) (35.2-38.9) (34.7-38.9) (34.7-39.1) (36.0-39.1) (36.5-39.0) (37.1-39.2)
c c b a a a
Vulva 36.2±0.9 36.2±0.7 36.7±0.7 37.1±0.7 37.1±0.8 37.0±0.5 <.001
(IRT) (33.0-38.6) (34.5-37.8) (34.1-38.2) (34.5-38.5) (35.2-38.7) (36.2-37.8)
Muzzle 35.0±1.1c 35.1±1.0c 35.5±1.0b 35.4±1.0
b
36.0±0.9
a a
36.1±0.7 <.001
(IRT) (31.0-37.8) (32.3-37.3) (31.4-37.7) (31.8-37.6) (33.4-37.7) (34.2-37.6)
ab b ab ab a b
Eye 36.7±0.7 36.5±0.7 36.6±0.6 36.7±0.5 36.8±0.5 36.5±0.6 0.009
(IRT) (34.1-38.0) (33.4-38.2) (33.8-37.8) (35.6-37.5) (35.9-37.9) (35.2-37.5)
c d b ab a a

PT
Ear 33.3±1.8 32.7±2.0 34.2±2.1 34.7±1.6 35.3±1.7 35.4±1.7 <.001
(IRT) (28.5-39.2) (27.6-37.1) (28.6-41.2) (29.5-37.9) (30.0-38.1) (31.1-38.8)
267 Temp = temperature; °C = degreee Celsius; DT = digital thermometer; IRT = infrared thermography.
268 Phase 1 (D-20 to D-16) = data collection before the synchronization of estrus. Phase 2 (D-15 to D-12)

RI
269 = start of estrus synchronization. Phase 3 (D-11 to D-4) = period of estrus synchronization, until
270 sponge removal. Phase 4 (D-3 to D-1) = estrus manifestation period. Phase 5 (D0) = ovulation. Phase
271 6 (D5) = period after ovulation.

SC
272 Different letters in the same row indicate statistical difference between the means (P<0.05).
273

274 The differences (p<0.05) between body and surface temperatures in the

275
U
different phases of the experimental period are presented in Table 1. There was a
AN
276 difference between the phases for all analyzed variables. Rectal and vaginal
M

277 temperatures, measured with the digital thermometer, were lower during the period of

278 exposure to progesterone and during estrus (phases 1, 2, 3 and 4), and increased
D

279 during and after ovulation (phases 5 and 6, respectively).

TE

280 The measurements made by infrared thermography demonstrated that the

281 surface temperatures of the anus, muzzle and ear increased gradually from the
EP

282 insertion of the sponge (phase 2), reaching the highest temperatures during ovulation
C

283 (phase 5) and after it (phase 6). The highest vulva surface temperature occurred at
AC

284 estrus (phase 4) and remained high during and after ovulation. The eye temperature

285 gradually increased from the insertion of the sponge (phase 2), achieving its

286 maximum value at ovulation (phase 5), and decreased soon thereafter (phase 6).

287 The rectal and vaginal temperatures were highest during and after ovulation

288 (Phases 5 and 6). The anus and vulva surface temperatures were higher from estrus

289 to post-ovulation (Phases 4, 5 and 6). All variables had significant pairwise

12
 ACCEPTED MANUSCRIPT
290 correlations (p <0.05), the great majority being of moderate magnitude. The

291 correlation between the rectal and vaginal body temperature (r=0.80) and between

292 the surface of anus and vulva (r=0.62) were high.

293 When the increase of vulvar IRT considered was 0.6 °C, the sensitivity was

294 70% (14 ewes with true positive results and 6 with false positive results). However,

PT
295 when the increase considered was 1.0 °C, the sensitivity was 45% (9 ewes with true

RI
296 positive results and 11 with false positive results).

297

SC
298 4. Discussion

299

300
U
The detection of estrus in ewes is labor intensive and time consuming, but
AN
301 failures in estrus detection can cause economic losses [3]. Ewes do not show evident
M

302 signs of estrus, such as swollen vulva or mucosal discharge through the vagina, so

303 the presence of a male is necessary. Since females show active search behavior in
D

304 relation to the male and immobility during mating, this is considered the most
TE

305 important and reliable sign of the occurrence of estrus in this species [20].

306 In the present study, we observed pairwise correlations between all the
EP

307 environmental variables (air temperature, relative humidity and WBGT) and between

308 the environmental temperature and the body and surface temperatures. Most of the
C

309 correlations were moderate and positive, although all correlations between humidity
AC

310 and the other environmental variables or the body and surface temperatures were

311 always negative and almost always moderate.

312 The variations of temperature and air humidity contribute to the surface

313 temperature variation, thus affecting the thermographic accuracy, since the highest

314 temperature of animals generally occurs in the hottest period of the year [8][11][19].

13
 ACCEPTED MANUSCRIPT
315 According to Landgraf et al. [21], the longer the period of data collection (with

316 consequent change in environmental conditions), the greater the effect of

317 temperature will be. Therefore, it is evident that thermography is sensitive to

318 environmental changes, so that air temperature and humidity and WBGT were

319 evaluated in this experiment and considered in the statistical analysis as correction

PT
320 factors to decrease the degree of exogenous interference.

RI
321 Although we used environmental variables as correction factors in the

322 statistical analysis, it is possible to record this information directly in the

SC
323 thermographic camera before the images are obtained, since the camera’s software

324 calculates the correction [21]. Another option is to insert the information in the image

325
U
evaluation software during this evaluation. These possibilities enhance the practical
AN
326 applicability of the technique, by minimizing the environmental influence on the
M

327 surface temperatures of the animals without the need for statistical tools.

328 The sensitivity for the temperature of 0.6 °C was superior to that for 1.0 °C.
D

329 Talukder et al. [19] observed in cows sensitivity value of 83% and 58% for estrus
TE

330 detection by vulval IRT using temperature from 1.0 °C to 1.4 °C respectively, greater

331 than observed in this study.

EP

332 We utilized the average temperature, since the standard deviation of the data

333 are significantly reduced for such temperatures when compared with the maximum
C

334 and minimum values. The body temperature (digital thermometer) was always higher
AC

335 than the surface (infrared thermography), which is in agreement with the results

336 reported by [2]. In this study, only the vulva, measured by thermography, showed the

337 highest temperature during estrus. In relation to this result, different temperature

338 patterns have been identified, for example, our result was opposite to the found by [2]

339 in cattle, where thermography showed little efficacy to detect estrus (phase 4). Using

14
 ACCEPTED MANUSCRIPT
340 infrared thermography in gilts, [22] found increased in the vulvar temperature during

341 of estrus in relation to the diestrus, since [23] demonstrated a decrease in

342 temperature measured by IRT vulvar 12 hours before ovulation in sows.

343 The highest temperatures were observed during the period of ovulation (phase

344 5) in all anatomical regions, and remained high or decreased thereafter (which

PT
345 occurred in the eye and muzzle). This result was different from that showed by [11],

RI
346 which used infrared thermography to monitor the estrous cycle in mares and reported

347 an increase in the vulvar temperature during the follicular growth and a decrease of

SC
348 the same temperature during the establishment of CL.

349 The body temperature of the rectum and vagina serve as reference for the

350
U
body core temperature, with the rectal temperature being most commonly used [24]
AN
351 [25]. In the present study, the rectal and vaginal temperatures, measured with the
M

352 digital thermometer, reached the highest levels during and after ovulation (phase 5

353 and 6), and presented high correlation with each other (r=0.80). George et al. [24]
D

354 also found a high correlation (r=0.96) between these regions in sheep, which means
TE

355 that vaginal temperature can also be used as way to measure core body

356 temperature, similar to rectal temperature.

EP

357 The anus surface temperature, measured with infrared thermography, was

358 stable across measurements (SD=0.71 °C), presenting the highest means during and
C
AC

359 after ovulation (phase 5 and 6). As happened with the rectal and vaginal

360 temperatures, the correlation between the anus and vulva temperatures were also

361 high. (r=0.62). These two temperatures are associated with the decline of serum

362 progesterone and peak of luteinizing hormone (LH) during estrus, but their

363 measurement has the drawbacks of consuming time and being invasive [2], due to

364 the need to keep the tail raised [19].

15
 ACCEPTED MANUSCRIPT
365 The surface temperature of the vulva, measured by thermography, was the

366 variable which had the closest relationship to the estrus period due to its great

367 stability (SD = 0.84) and showed higher mean temperature from estrus to the post-

368 ovulation period (phases 4, 5 and 6). Stelletta et al. [26] also detected a higher mean

369 temperature in ewes in estrus when compared to the control group in anestrus and

PT
370 affirmed that females in estrus showed an increase in blood flow from the internal

RI
371 iliac artery to the genital tract and also hormonal changes. Signs of estrus, including

372 effects on the vulva (edema and redness due to increased blood flow in vulvar

SC
373 tissues), are induced by an increase of estradiol level in the bloodstream [2]. Ewes

374 present spontaneous ovulation near the end of estrus, about 24 to 27 hours after

375 initiation [5].

U
AN
376 The temperature of the eye measured by thermography was the most constant
M

377 (SD = 0.61 ° C) during the estral cycle. The highest temperature was observed at

378 phase 5 (ovulation), but the difference with phase 4 (estrus) was only 0.1 °C, which
D

379 may not be relevant in practice. Due to the high precision, practicality and elimination
TE

380 of the discomfort and stress that can be caused by containment, which can increase

381 the body temperature, thermography of the eye can be more advisable than of the
EP

382 anus and vulva, since these measurments need containment to raise the tail, in

383 addition to the possibility of fecal contamination [19]. So, the eye temperature can be
C

384 used as a measure of core body temperature [21],[24][25][8] because it undergoes

AC

385 less variation in regular measurements, does not involve contact with the skin or dirt

386 and its temperature was the closest to the values of the rectal, vaginal and vulvar

387 temperature.

388 The temperature of the muzzle measured by thermography presented higher

389 variation than observed in the anus, vulva and eye during the experiment (SD = 1.07

16
 ACCEPTED MANUSCRIPT
390 °C), increasing gradually from the insertion of the sponge (phase 2) and reaching the

391 highest temperatures in ovulation (phase 5) and afterward (phase 6). The

392 temperature in phase 5 was 0.6 °C higher than in phase 4 and this was the only

393 anatomical region with the practical potential to detect ovulation. When assessing a

394 thermal image, asymmetry of 1 °C or more is often significant and indicates possible

PT
395 pathology [10]. In cows, there is greater variability in the muzzle temperature

RI
396 compared to the vulva, which can occur due to water intake, as a consequence of the

397 thermoregulation of the muzzle as a physiological response, and by the habit of

SC
398 licking the muzzle frequently [24]. In sheep, the high correlation found in the work of

399 George et al. [24] (r = 0.65) between muzzle and rectum (r = 0.63) and muzzle and

400
U
eye (r = 0.73) may be a consequence of the low number of samples collected and the
AN
401 behavior of sheep, which do not have the habit of licking their muzzle.
M

402 The ear was the anatomical region that showed the highest temperature

403 variation during the experiment (SD = 2.08 °C), obtaining the highest averages during
D

404 and after ovulation (phases 5 and 6). This result may have occurred due to the
TE

405 anatomical position and hair cover of the ear. Not all edges of the ear are in direct

406 contact with the body, thus being more influenced by environmental conditions.
EP

407 Lodetti [27] affirmed that some parts of the body not covered by hair are the only

408 areas on which it appears possible to read the correct surface temperature of
C

409 animals, while Church et al. [10] noted that hair coats should be free of dirt, moisture
AC

410 or foreign material.

411

412 5. Conclusion

413

17
 ACCEPTED MANUSCRIPT
414 The IRT was efficient to identify superficial temperature variations during the

415 different phases of estrus cycle in Santa Ines ewes and the better anatomical regions

416 to carry out these measurements are vulva and muzzle.

417 IRT is a promising tool for reproductive management in ewes. The differences

418 in temperature detected between phases of the estrous cycle can be used in the

PT
419 development of new technologies associated with precision in livestock farming.

RI
420

421 Conflict of interests

SC
422 The authors declare they have no competing interests.

423

424 Acknowledgements
U
AN
425 We thank professors Eliemar Campostrini and Luis Fonseca Matos for loaning
M

426 the equipment and veterinary doctor Natália Torres for the technical instructions and

427 Alexandra Elbakyan. Funding was provided by Fundação de Amparo a Pesquisa do

D

428 Estado do Rio de Janeiro (FAPERJ), Conselho Nacional de Desenvolvimento

TE

429 Científico e Tecnológico (CNPq) and Coordenação de Aperfeiçoamento de Pessoal

430 de Ensino Superior (CAPES).

EP

431

432 References
C

433 [1] KNÍŽKOVÁ I, KUNC P. Applications of infrared thermography in animal

AC

434 production. J Fac Agric 2007;22:329–36.

435 [2] Talukder S, Thomson PC, Kerrisk KL, Clark CEF, Celi P. Evaluation of infrared

436 thermography body temperature and collar-mounted accelerometer and

437 acoustic technology for predicting time of ovulation of cows in a pasture-based

438 system. Theriogenology 2015;83:739–48.

18
 ACCEPTED MANUSCRIPT
439 doi:10.1016/j.theriogenology.2014.11.005.

440 [3] Scolari SC, Clark SG, Knox R V, Tamassia MA. Vulvar skin temperature

441 changes significantly during estrus in swine as determined by digital infrared

442 thermography. J Swine Heal Prod 2011;19:151–5.

443 [4] Ekiz EE, Ekiz B, Koçak Ö. Effects of ram presence during synchronization

PT
444 period and previous experience on certain estrus parameters and sexual

RI
445 behaviors in Kivircik ewes. Turkish J Vet Anim Sci 2013;37:189–93.

446 doi:10.3906/vet-1202-30.

SC
447 [5] Hafez B, Hafez ESE. Reprodução animal. 7th ed. Barueri, SP: Manole; 2004.

448 [6] McManus C, Tanure CB, Peripolli V, Seixas L, Fischer V, Gabbi AM, et al.

449
U
Infrared thermography in animal production: An overview. Comput Electron
AN
450 Agric 2016;123:10–6. doi:10.1016/j.compag.2016.01.027.
M

451 [7] Dwyer CM. The welfare of sheep. 6th ed. Edinburgh: Springer; 2008.

452 doi:10.1007/978-1-4020-8553-6_1.
D

453 [8] Švejdová K, Šoch M, Šimková A, Zábranský L, Novák P, Brouček J, et al.

TE

454 Measuring the Body Surface Temperature of Animals Using a Thermographic

455 Camera. Acta Univ Cibiniensis Ser E Food Technol 2013;17:99–106.

EP

456 doi:10.2478/aucft-2013-0017.

457 [9] Roberto JVB, Souza BB. Utilização da termografia de infravermelho na

C

458 medicina veterinária e na produção animal. J Anim Behav Biometeorol

AC

459 2014;2:73–84. doi:10.14269/2318-1265/jabb.v2n3p73-84.

460 [10] Church JS, Cook NJ, Schaefer AL. Recent applications of infrared

461 thermography for animal welfare and veterinary research: everything from

462 chicks to elephants. InfraMation 2009:10.

463 [11] Simões VG, Lyazrhi F, Picard-hagen N, Gayrard V, Martineau G, Waret-szkuta

19
 ACCEPTED MANUSCRIPT
464 A. Variations in the vulvar temperature of sows during proestrus and estrus as

465 determined by infrared thermography and its relation to ovulation.

466 Theriogenology 2014;82:1080–5.

467 [12] Aisen EG. Reprodução ovina e caprina. São Paulo, SP: MedVet, 1 ed., . 1st

468 ed. São Paulo: Medvet; 2008.

PT
469 [13] Paes de Barros MB. Estudo da dinâmica folicular no protocolo de nove dias de

RI
470 sincronização de cios com o uso de um novo dispositivo de liberação

471 sustentada de progesterona, associado ou não ao eCG, em ovelhas Santa

SC
472 Inês 2010.

473 [14] Ginther OJ, Kot K. Follicular dynamics during the ovulatory season in goats.

474 Theriogenology 1994;42:987–1001.

U
AN
475 [15] Balaro MFA, Brandão FZ, Peneiras AB V, Oba E, Fonseca JF, Almosny NRP,
M

476 et al. Reproductive performance, metabolic and hormonal profiles of Santa

477 Inês ewes in winter and summer under tropical conditions. Trop Anim Health
D

478 Prod 2015;47:627–31. doi:10.1007/s11250-015-0757-z.

TE

479 [16] Butler D. asreml: asreml() fits the linear mixed model 2009.

480 [17] R Core Team. R: A Language and Environment for Statistical Computing.
EP

481 2014.

482 [18] SAS Institute. SAS/STAT® 9.3 User’s Guide. 2012.

C

483 [19] Talukder S, Kerrisk KL, Ingenhoff L, Thomson PC, Garcia SC, Celi P.
AC

484 Theriogenology Infrared technology for estrus detection and as a predictor of

485 time of ovulation in dairy cows in a pasture-based system. Theriogenology

486 2014;81:925–35. doi:10.1016/j.theriogenology.2014.01.009.

487 [20] Gonçalves PBD, Figueiredo JR, Freitas VJF. Biotécnicas aplicadas à

488 reprodução animal. São Paulo: VARELA EDITORA E LIVRARIA LTOA; 2002.

20
 ACCEPTED MANUSCRIPT
489 [21] Landgraf T, Zipser S, Stewart M, Dowling SK, Schaefer AL. Modelling and

490 Correction of Influences on Surface Temperature Measurements using infrared

491 thermography for animal health and welfare assessments 2014:1–6.

492 [22] Sykes DJ, Couvillion JS, Cromiak A, Bowers S, Schenck E, Crenshaw M, et al.

493 The use of digital infrared thermal imaging to detect estrus in gilts.

PT
494 Theriogenology 2012;78:147–52. doi:10.1016/j.theriogenology.2012.01.030.

RI
495 [23] Luño V, Gil L, Jereza RA, Malo C, González N, Grandía J, et al. Determination

496 of ovulation time in sows based on skin temperature and genital electrical

SC
497 resistance changes. Vet Rec 2013;172:579. doi:10.1136/vr.101221.

498 [24] George WD, Godfrey RW, Ketring RC, Vinson MC, Willard ST. Relationship

499
U
among eye and muzzle temperatures measured using digital infrared thermal
AN
500 imaging and vaginal and rectal temperatures in hair sheep and cattle. J Anim
M

501 Sci 2014;92:4949–55. doi:10.2527/jas2014-8087.

502 [25] Hoffmann G, Schmidt M, Ammon C, Rose-Meierhöfer S, Burfeind O,

D

503 Heuwieser W, et al. Monitoring the body temperature of cows and calves using
TE

504 video recordings from an infrared thermography camera. Vet Res Commun

505 2012;37:91–9. doi:10.1007/s11259-012-9549-3.

EP

506 [26] Stelletta C, Gianesella M, Vencato J, Fiore E, Morgante M. Thermographic

507 Applications in Veterinary Medicine. In: Prakash R V, editor. Infrared

C

508 Thermogr., Rijeka, Croatia: InTech; 2012, p. 117–40.

AC

509 [27] Lodetti E. Thermography: current status and advances in livestock animals and

510 in veterinary medicine. Brescia: FONDAZIONE INIZIATIVE

511 ZOOPROFILATTICHE E ZOOTECNICHE - BRESCIA; 2013.

512

513

21
 ACCEPTED MANUSCRIPT
514 Figures

515

516 Figure 1 Experimental design for days (D) and phases of study.

517

518 Figure 2 Anatomical regions delimited in the Flir Quickreport® version 1.2 SP2

PT
519 (1.0.1.217) 2009 software in sheep.

RI
520

521 Area of measurement around the anus (a1), in the center of the vulva (a2), ear (b),

SC
522 muzzle (c) and eye (d). Warm areas appear white and cooler areas appear black.

523

524
U
Figure 3 Variation of temperature in sheep for each anatomical region according to
AN
525 day of the experiment.
M

526

527 DT = digital thermometer; IRT = infrared thermography.

D

528
TE
C EP
AC

22
 ACCEPTED MANUSCRIPT
Figure 1

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
Figure 2

PT
RI
U SC
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
Figure 3

40

39

38

37 Rectum (DT)
Temperature 0C

Vagina (DT)

PT
36
Anus (IRT)
35
Vulva (IRT)

RI
34 Muzzle (IRT)
Eye (IRT)
33

SC
Ear (IRT)
32

31

U
-15 -14 -13 -12 -11 -10 -9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2
Day
AN
M
D
TE
C EP
AC
 ACCEPTED MANUSCRIPT
Highlights

- The surface temperature of ewes varies during the estrous cycle.

- Thermography can identify temperature variation during estrus.

PT
- IRT is a promising tool for reproductive management in ewes.

RI
- Vulva and muzzle are good anatomical regions for temperature evaluation of

estrous cycle.

U SC
AN
M
D
TE
C EP
AC