The rhythmicity of body temperature is an important process to be studied, not only to advance knowledge of the temporal variability of thermal homeostasis, but also as a means to facilitate the study of biological rhythmicity in general. In view of the relative ease of monitoring body temperature and the robustness of its rhythm, the rhythmicity of body temperature has been widely used as an indicator of the rhythmicity of the biological clock [1, 2, 3]. The measurement of body temperature is an important procedure for assessment of physiological status of the animal  as well as an ideal indicator of its responses and adaptability to environmental stress factors .
Assessment of livestock production in the tropics has focused more on the adverse effect of heat stress during the hot-dry season, which is characterized by high ambient temperature and intense solar radiation, and the effect of cold-dry (harmattan) stress on livestock has been poorly investigated . The harmattan season, characterized by cold-dry and dusty wind, blowing north-east and west off the Sahara desert into the Gulf of Guinea, has been shown to be the most thermally stressful season in the Northern Guinea Savannah zone [7, 8, 9].
Studies on daily rhythmicity of physiological variables have been conducted mostly on humans and laboratory animals, while the subjects were maintained in a thermoneutral environment or controlled experimental photoperiod [3, 10]. While such studies remain relevant in understanding the endogenous and exogenous roles in producing circadian rhythms, there is mounting evidence that rhythm patterns often differ between laboratory-controlled settings and natural environment . Individuals under natural conditions are involved in complex interrelation with the microenvironment, which may produce changes in circadian activities that are not reflected in purely chronobiological studies carried out under laboratory conditions [10, 11]. Only of recent, few studies were conducted under natural conditions on livestock involving the influence of ambient temperature on the daily rythmicity of core temperature and blood biochemical variables [3, 12, 13, 14, 15]. The effects of natural environmental conditions on daily rhythm in livestock kept under extensive management system and pronounced natural seasonality requires further investigation.
Measurements of body surface temperature (BST) in animals with the aid of infrared thermography (IRT) and infrared thermometry (IRTM), which are non-contact methods of temperature measurements, offer several advantages over contact methods used in veterinary medicine [16, 17, 18]. A comprehensive collection of infrared imaging applications in livestock is described by Luzi et al. .
To date, there is a paucity of information in the available literature on the effect of the harmattan and hot-dry seasons on daily rhythmicity of core body temperature and body surface temperature (BST) of sheep and goats reared in the Northern Guinea Savannah zone. Studies with the aid of IRT for measurement of BST rhythmicity and thermographic standards in indigenous small ruminant species under natural tropical conditions and different photoperiods have not been reported. Such information, if obtained, may be of value to clinicians, researchers and agriculturists in the understanding of biological rhythm, and in enhancing the health status, welfare and productivity of sheep and goats under different natural conditions.
The aim of the present study was to investigate the circadian rhythmicity of rectal temperature (RT) and BST, and their characteristics in sheep and goats under different natural photoperiods and seasons.
Materials and Methods
Study area and management of animals
The experiment was conducted during the harmattan (November-December) and hot-dry (March-April) seasons of 2015–2016 at the Livestock Farm of the College of Agriculture and Animal Science, Mando-Kaduna (11° 10′ N, 07° 38′ E), located in the Northern Guinea Savannah zone of Nigeria. Twenty healthy, mature indigenous goats (Red Sokoto goats) and 20 indigenous sheep (Yankasa sheep), aged 1.5–2 years served as subjects. The animals were kept under a natural light-dark cycle during the hot-dry and harmattan seasons at the small ruminant research pen of the College of Agriculture and Animal Science, Mando-Kaduna, Nigeria. They were housed in individual stands in communal pens, made of concrete floor and cement block wall. The pens measured 2.42 m × 7.39 m wide and 1.12 m high of wire-mesh from the floor, which provided for adequate ventilation. The animals were fed beans husk, maize bran and groundnut hay, and given access to clean drinking water ad libitum. They were routinely dewormed, treated against ticks and vaccinated against most common diseases of small ruminants in the area of study. The health of the animals was assessed a week before the start of the study by screening of their blood samples for haemo-parasites, and by clinical evaluations of their rectal temperature, and respiration and heart rates for three consecutive days.
Measurements of rectal temperature and thermal environental conditions
The ambient temperature (AT) and relative humidity (RH) inside the pen, and the RT and BST of the sheep and goats were measured on two separate days at three-hour intervals for two days. Each measurement began at 08:30 h on the first day and ended at 08:30 h the second day. The animals were easily restrained for measurements of RT, recorded from each goat as an indicator of the body temperature by a digital clinical thermometer (Electron Thermometer, COCET, Kangfu Medical Equipment, Zhejiang Yueqinq, China), inserted at about 2 cm into the rectum via the anus. The readings were taken after an alarm sound was heard, indicating the end of the measurement. Within the same period of RT measurements, BST was also recorded at five different body locations using a digital infrared thermometer (Model AR330, China) with a resolution of 0.1°C. The locations comprised the inter-digital space, coronary band, mid-head, nose and ear-base regions. The IRT was focused at an approximate distance of not more than 1.2 meters from the site of recording .
The dry- and wet-bulb temperature values were obtained concurrently with the RT readings at the experimental site, using a dry- and wet-bulb thermometer (Hartmann Company, England). The animals were humanely handled and treated according to Code of Recommendations for the Welfare of Livestock  and as approved by the Research, Seminar and Welfare Committee of the Animal Health Department of the College (No. 44–14/16).
Each time series was evaluated for rhythmicity by repeated-measures analysis of variance (ANOVA model-3) and by the cosinor procedure [3, 22]. Cosinor analysis was used to determine the RT and BST daily rhythms of individual animals. The mean mesor (rhythm-adjusted mean), amplitude (half the range of excursion or a measure of the extent of predictable change within a cycle), acrophase (time of peak) and robustness (strength of rhythmicity computed as the fraction of the variance explained by the cosine model) values of the variables of daily rhythm were calculated for each animal and for each time series of the study period. Rhythm robustness was computed as a percentage of the maximal score, attained by the chi-square periodogram statistic for ideal data sets of comparable size and 24-h periodicity. The effects of animal species, season and interaction between seasons and time of day within subjects were compared. The difference was tested using two-way repeated-measures ANOVA. Values of p < 0.05 were considered significant.
Ambient temperature and relative humidity
The AT and RH during the study periods are depicted in Figures 1a and b. The mean, maximum and minimum AT values of 28.4 ± 0.2, 32.5 and 23.0°C recorded during the hot-dry season were higher (P < 0.05) than the corresponding values of 20.1 ± 0.1, 28.2 and 16°C, obtained during the harmattan season (Figure 1a and b). A similar trend was observed in the RH values (Figure 1b). In general, the AT and RH values were lower during the scotophase than the photophase.
The mean, maximum and minimum RT values obtained during the study period in the goats were 38.65°C ± 0.05, 39.5°C and 37.8°C, respectively; and in sheep the values were 38.14 ± 0.05°C, 39.20°C and 37.0°C, respectively.
In both seasons the RT of the sheep and goats exhibited a strong and clear daily rhythm (Figures 2a and b). Higher peak values of the daily rhythm of RT were recorded in both sheep and goats during the photophase (07:00–19:00 h) as compared to scotophase.
Characteristics of rectal temperature
Table 1 shows the RT rhythmic parameters of mesor, amplitude, acrophase and robustness. During the harmattan in sheep, the mesor (38.2) was lower, and the amplitude was greater (1.20); while the acrophase was delayed to 20:00 h as compared with the mesor value of 38.8, amplitude of 0.85 and acrophase of 17:30 h observed in goats. During the hot-dry season, the rhythmic parameter of mesor in sheep (38.8 ± 0.5°C) was not different from the value obtained in goats (38.7 ± 0.3°C). The amplitude values were slightly different (0.55 in sheep versus 0.45 in goats). Sheep and goats attained the acrophase (highest peak) at the same time (18:00 h). The goats had a more robust RT rhythm (70%) as compared to the sheep (56%) (Table 1).
|Harmattan Season||Hot-dry Season|
|Goat (n=20)||Sheep (n=20)||Goat (n=20)||Sheep (n=20)|
|Rectal temperature||38.8 ± 0.2a||38.2 ± 0.3b||38.8 ± 0.5a||38.7 ± 0.4a|
|Inter-digital space||22.7 ± 0.4a||22.0 ± 0.2a||27.4 ± 0.1b||26.5 ± 0.5b|
|Coronary band||22.5 ± 0.4a||21.4 ± 0.5a||27.6 ± 0.2b||26.1 ± 0.4b|
|Nose||24.1 ± 0.2a||23.4 ± 0.4a||29.0 ± 0.4b||28.7 ± 0.5b|
|Head||27.7 ± 0.3a||26.1 ± 0.2b||31.2 ± 0.5c||30.8 ± 0.2c|
|Ear||27.2 ± 0.2a||25.0 ± 0.5b||30.7 ± 0.4c||30.7 ± 0.4c|
|Rectal temperature||0.85 ± 0.02a||1.20 ± 0.05b||0.55 ± 0.01c||0.45 ± 0.04c|
|Inter-digital space||5.60 ± 0.10a||6.35 ± 0.20b||4.51 ± 0.10c||3.01 ± 0.02d|
|Coronary band||4.60 ± 0.01a||3.50 ± 0.04b||4.50 ± 0.07a||3.50 ± 0.04b|
|Nose||7.15 ± 0.20a||6.65 ± 0.08b||5.30 ± 0.01c||4.10 ± 0.02d|
|Head||5.65 ± 0.08a||5.25 ±0.04a||4.90 ± 0.02b||4.80 ± 0.06b|
|Ear||7.00 ± 0.21a||6.00 ± 0.11b||5.50 ± 0.05c||5.00 ± 0.05c|
|Rectal temperature||17:30 ± 0.14a||20:30 ± 0.10b||18:00 ± 0.12c||18:00 ± 0.14c|
|Inter-digital space||17:30 ± 0.20a||17:30 ± 0.11a||11:30 ± 0.10b||11:30 ± 0.10b|
|Coronary band||17:30 ± 0.25a||17:30 ± 0.10a||11:30 ± 0.11b||11:30 ± 0.10b|
|Nose||17:30 ± 0.15a||17:30 ± 0.14a||11:30 ± 0.10b||11:30 ± 0.10b|
|Head||17:30 ± 0.15a||14:30 ± 0.15b||11:30 ± 0.10c||11:30 ± 0.11c|
|Ear||17:30 ± 0.14a||14:30 ± 0.12b||11:30 ± 0.11c||11:30 ± 0.12c|
|Rectal temperature||69 ± 10.5a||54 ± 12.7b||71 ± 15.2a||58 ± 10.5b|
|Inter-digital space||60 ± 9.5a||52 ± 11.5a||61 ± 8.7a||58 ± 11.2a|
|Coronary band||59 ± 10.0a||55 ± 9.5a||60 ± 10.2a||60 ± 7.5a|
|Nose||54 ± 7.8a||44 ± 8.6a||50 ± 9.5a||48 ± 10.0a|
|Head||52 ± 8.5a||50 ± 9.2a||60 ± 10.5a||55 ± 11.4a|
|Ear||60 ± 12.4a||58 ± 10.2a||62 ± 8.9a||58 ± 7.3a|
Body surface temperatures
The BST at various regions of the body in both sheep and goats exhibited a daily, but weak rhythmicity (Figures 3, 4, 5, 6, 7). The feet (inter-digital and coronary band) temperatures (Figure 3 and 4) were lower (p < 0.05) than those recorded in other regions, while the highest values were obtained in the head, ear and nose regions (Figure 5 and 7). The BST value differed based on the season and animal’s species. It was generally lower during the harmattan season, especially in the night, than in the hot-dry season; and in the afternoon hours of the day. Overall, the BST of goats was significantly (p < 0.05) higher than that of the sheep in all the seasons.
Characteristics of body surface temperature
The mesors of BST recorded in both sheep and goats were significantly higher (p < 0.05) during the hot-dry than harmattan season (Table 1). In general, the extent of predictable changes (amplitudes) of the daily rhythm of BST in the sheep and goats was greater (p < 0.001) during the harmattan season as compared to the hot-dry season. The goats had greater (p < 0.05) amplitude in all the seasons. The acrophases of daily rhythm of BST were all restricted to the photophase (Table 1) with peak at 11:30 h in both sheep and goats during the hot-dry season; while during the harmattan season, the acrophases of BST coincided in time near those of the RT (17:30 h). Exceptions were the acrophases of the head and ear in sheep, which were recorded at 14:30 h. The robustness of the RT and BST rhythms of the goats and sheep is shown in Table 1. The robustness of RT was significantly (p < 0.05) stronger than the BST in all seasons. The goats had a more robust RT rhythm (70%) as compared to the sheep (56%). The seasons had no significant effect on robustness.
The mean AT and RH recorded in the present study in both seasons were within the thermoneutral zone of 22 – 35°C and 28 – 60%, respectively, established for sheep and goats [5, 23, 24]. However, the minimum AT (16°C) recorded during the scotophase of the harmattan season was outside the thermoneutral zone established for goats and sheep in the tropical regions.
The mean RT values recorded were within the normal reference intervals values of 38 – 40°C for sheep and goats in the tropics [5, 8]. However, the RT values recorded in both species during the early morning hours of the harmattan season were considerably lower, which suggests that the season induced hypothermia in the sheep and goats. The result supports several reports that different seasons of the year exert different effects on body core temperature [4, 25]. The RT and BST in both ruminant species exhibited individual variation typical of mammals . Overall, at each hour of measurements, the RT and BST of sheep were lower than those of goats in both seasons, apparently due to species variation.
The application of the periodic model and the fitting of cosinor procedure showed that the RT exhibited a strong daily rhythmicity, while the BST exhibited a weak daily rhythmicity in both sheep and goats in all the seasons. The result agrees with the previous finding of diurnal rhythm in the RT of Comisana sheep  and Red Sokoto goats recorded only during the day time by Ayo et al. . The oscillatory pattern in RT and BST differed with seasons, with stronger daily oscillation being observed during the harmattan season than during the hot-dry season.
The mesors of RT recorded in sheep and goats were consistently higher than those of BST recorded at various sites of the body, which agreed with the findings of Piccione et al. , who showed that surface body temperature was lower than core temperature in sheep, and this result is true of all endothermic animals [4, 26]. The lower mesor of RT recorded in sheep when the AT was 16 – 28°C during the harmattan season was in contrast to that recorded by Piccione et al. , who found no effect of cold AT (9°C) or thermoneutral AT on mesor in sheep. The present result demonstrated the adverse effect of the harmattan season on the thermoregulation of sheep, and confirmed the earlier report that the harmattan season is the most thermally stressful of all the seasons prevailing in the Northern Guinea Savannah zone [5, 7]. Although the AT (16 – 28.2°C) during the harmattan season was not as low as 9°C, its hypothermic effect on daily rhythmicity in RT of sheep was more severe than that recorded in Comisana sheep, exposed to cold AT of 9°C. The present finding suggests that, aside from breed differences, additional factors, including cold-dry air, heavy dust-laden particles and wide fluctuations in AT, which are characteristics of the harmattan season, may be responsible for the hypothermic effect.
The mesors of daily rhythm of BST in both sheep and goats demonstrated that they were modulated by AT; with higher values occurring during the hot-dry season than harmattan season, and the goats had higher mesors in all the body sites as compared to those of the sheep which had cooler BST. In both sheep and goats, the feet (inter-digital and coronary) temperatures were cooler than the RT and temperatures in other parts of the body measured, while the highest temperatures recorded in the head, ear and nose showed that these parts of the body were the hottest. The feet temperature in the present study varied compared with the finding of Piccione et al. , who showed that feet temperature was 9°C cooler than the rectum; however, the head temperature was similar to the value of 9°C obtained by Piccione et al. . The lower feet temperature, especially during the harmattan (cold-dry) season, as compared to other sites confirmed that the feet play a more crucial role in thermoregulation than other parts of the body, probably because vasoconstriction in the feet is a good mechanism of heat conservation [3, 27].
In general, the lower mesor in BST in sheep than in goats may be due to species difference, integumentary variations and body fat. In addition, the white hair-skin colour characteristic of the Yankasa sheep has been shown to influence evaporative cooling and repel heat; unlike the dark-brown colour of the goats’ skin, which may enhance solar absorption and, thus, increase heat load on the skin surface [28, 29].
The greater amplitude of RT and BST during the harmattan season may be a result of the adverse effect of cold AT and greater variation of AT on the animals. Greater amplitude of core temperature has been descrbed in mammals reared under cold conditions in temperate regions [3, 6, 30, 31].
In homeothermy, the limit of variability for amplitude of daily rhythmicity of core temperature is considered to be ± 2°C , which is wider than the amplitude of daily rhythmicity of RT values of 0.6 – 1.2°C recorded in sheep and goats in the present study. However, the amplitude obtained in the current study is similar to those of 0.3 – 1.2°C and 0.4 – 1.9°C, reported in sheep  and goats , respectively, but greater than the amplitudes of 0.2 – 0.4°C recorded in sheep [3, 34] and 0.26 – 0.33°C in humans . In both studies in sheep and goats mentioned above and in the current study, the highest amplitude of daily rhythmicity of RT values was recorded when AT oscillated about 20°C, which is also similar to the finding of Lovegrave and Heldmaier . The amplitude of daily rhythmicity of BST obtained at various sites in the present study was between 3 – 7°C, which was greater than the amplitude values of 0.5 – 1.3°C in sheep . The differences in the amplitudes may be due to variations in animal species and AT.
The RT during both harmattan and hot-dry seasons in the sheep and goats peaked earlier (17:30–20:00 h) than the acrophase value of 21:38 h reported in sheep . During the hot-dry season, the acrophase of BST at various sites in both sheep and goats were recorded in the morning (11:30 h), but during the harmattan season the acrophase of BST was delayed till late afternoon, at about 17:30 h (8 hours after sunrise).
The relationship between RT and BST of the sheep and goats recorded during the hot-dry season showed that the BST, especially the feet temperature, reached acrophase about six hours before RT, whereas during the cold-dry (harmattan) BST peaked at the same (17:30 h) time as RT (17:30 h). The present result is in contrast with the finding of Picconne et al. , who obtained a peak in foot temperature six and 13 hours before RT in sheep kept under cold condition (9°C) and themoneutral zone (25°C), respectively. Our result also disagrees with the finding in humans that foot and core temperatures attained their peaks at 12 h apart [37, 38]. The finding that BST, especially the feet temperature, peaked several hours before RT suggests that the mechanism of thermoregulation in the feet may be involved in production of the oscillation in RT as earlier suggested [35, 38]. In the present study, however, ambient temperature fluctuated daily, thus introducing a confounding factor not present in laboratory studies.
The maximal robustness of the RT and BST rhythms recorded in the sheep and goats in the present study did not differ between seasons, although the oscillation of BST was less robust than in RT and the goats had a relatively stronger robustness than the sheep. Similar robustness values of 44–70% were reported for goats, sheep, horses [34, 39, 40] and dogs  in temperate regions. However, the robustness obtained in the present result was less than the maximal robustness of 90% and 81% recorded in cattle  and dromedary camels . The robustness of BST obtained in the present study was similar to those reported in the limbs of sheep , but less than the robustness of 32% reported in camel’s skin .
The present result on BST indicates that IRT may be a useful non-invasive and accurate tool to detect the variations in BST in small ruminants, and the differences in BST distribution may be used for diagnosis of diseases; especially diseases accompanied with inflammatory conditions, leg disorders and detection of skin injuries, and topical applications of therapeutic agents. Overall, the present results showed that exogenous environmental cues, including seasonal variations and light-dark cycles, provide important information for adjusting body temperature in sheep and goats towards the maintenance of a stable thermoregulation.
In conclusion, the results of the present study demonstrated the effects of seasonal changes on the daily rhythmicity in RT and BST and their special features in indigenous sheep and goats under natural light-dark cycle. The data may serve as additional normal reference values for sheep and goats under different natural climatic conditions, and may be useful for the diagnosis, treatment and prevention of diseases. They may also be beneficial in the evaluation of the welfare and productivity of small ruminants under different environmental conditions.