This study was conducted in a buffalo production unit in south-eastern Mexico, at an elevation of 10 m above sea level. The zone is characterized by a subtropical wet climate or Cfa, according to the Koppen-Geiger classification, with an average temperature of 27°C and annual rainfall of 2,900 mm. The experimental periods were September to November 2020 and September to November 2021. The ambient temperature during the evaluation period had average maximum and minimum values of 36°C and 23°C. Maximum and minimum percentages of relative humidity fluctuated from 81 to 94%, respectively.

To evaluate calves from birth (day 0), 123 dam water buffaloes (Bubalus bubalis) of the Buffalypso breed, close to calving, were monitored. One week before calving, the animals were moved to a 5,000 m² maternity paddock (pasture) with a capacity of 20 buffaloes per group. The females were fed forage native to the region (Paspalum fasciculatum and Hymenachne amplexicaulis) with 50 g of mineral supplement. At birth, the calves were distributed into four groups according to quartiles (18–21). In total, 109 calves were included in this study. Fourteen female buffaloes and their calves were excluded because their agonistic temperament made it impossible to approach the calves to obtain the required measurements.

Immediately after the first intake of colostrum and formation of the mother-calf bond (observed as licking and ingestion of amniotic fluids and placental membranes by the dams) (7), the calves were weighed using a 100-kg Outmate Digital Crane Scale^® (Data Weighing Systems, USA) with a readout accuracy of 0.1 lb/0.05 kg-grams. Newborns were then divided into four experimental groups according to birth weight in each quartile. The first group (Q₁) included 25% of the calves with the lowest values recorded; quartile two (Q₂) included the lowest 25% of the mean; quartile three (Q₃) included 25% of values above the mean; and quartile four (Q₄) included 25% of the highest values. In this way, the distribution in quartiles was: Q₁ (37.8–41.2 kg, n = 25 calves); Q₂ (41.3–46.3 kg, n = 29 calves); Q₃ (46.4–56.3 kg, n = 29 calves); and Q₄ (56.4–60.3 kg, n = 26 calves).

Radiometric images were taken with a thermal camera model FLIR^® Thermal TM E80 (FLIR Systems, USA), with a resolution of 320 × 240 pixels, thermal sensitivity of <0.045°C, precision ± 2°C or 2% and an emissivity of 0.95. Each radiometric capture was performed at 1–2 m from the calf, focusing on three regions: lateral facial, frontal facial, and left or right latero- lateral. Six thermal windows were assessed: lacrimal gland, lacrimal caruncle, periocular region, nostrils, ear canal, and pelvic limbs (Figure 1). The measurement of surface temperatures in these windows was carried out for a period of 6 days, from the day of birth (day 0) to the 5th day postpartum (day 5).

Thermographic monitoring was performed twice: from 8:00 a.m. to 9:00 a.m. and from 4:00 p.m. to 5:00 p.m. in the central zone of Mexico. The ambient temperature during the September-November period ranged from 23 to 27°C in the morning (8:00-−9:00 a.m.) to 32–36°C (4:00 to 5:00 p.m.) in the afternoon. Relative humidity fluctuated from 81 to 93% in the morning to 82–94% in the afternoon. During the thermographic monitoring, neonates and their dams remained in the maternity paddocks all the time. The thermal imaging sequence was performed in the paddock. Buffalo calves were not moved or handled from their birthplace, so the thermal evaluation was carried out directly on the grass. Thermal images were taken in the following sequence: the camera placed in front of the animal to take the frontal facial region, then to the side to obtain the lateral facial view. Subsequently, the left latero-lateral and the right latero-lateral regions were captured on each side of the animal.

All radiometric images were stored in JPG format for later analysis using FLIR Tools Systems, USA^® software to set points in the proposed thermal windows to obtain maximum, minimum, and average temperatures from each calf on each sampling day during the mornings and afternoons.

The GraphPad Prism statistical package (ver. 9.4.0) was used to obtain descriptive statistics for the weight of the animals (Q₁, Q₂, Q₃, Q₄), time of day (morning, afternoon), thermal window, and age of the animals (days 0, 1, 2, 3, 4, 5). The Shapiro-Wilk test was performed to analyze the normality of the data. The weight and age of the calves and the time of day were considered as independent variables, whereas the temperature of the thermal windows was considered a dependent variable. To evaluate the effects of these variables, analysis of variance (ANOVA) was used in a mixed linear model:

where:

Y = response variable (temperature of the thermal windows).

W_i = effect of the weight of the animals (Q₁, Q₂, Q₃, Q₄).

T_j = effect of time of day (morning, afternoon).

A_k= effect of the animals' age (days 0, 1, 2, 3, 4, 5).

β= random effect (per animal).

μ = population mean.

e = residue.

Correlations between the time of day and thermal window temperatures on day 0 were obtained using Spearman correlation. Tukey's post-hoc test was adopted to evaluate the differences between means. In all significance level of p < 0.05 was set.

The animals monitored in this study were not touched or stressed, as infrared thermography is a non-invasive technique; therefore, the protocol did not require approval from an Ethics Committee. During the study, handling of the animals was carried out under the guidelines of the Official Mexican Standard, NOM-062-ZOO-1999, which stipulates the technical specifications for the production, care, and ethical use of animals.

Table 1 lists the thermal responses of the pelvic limb region. On the morning of day 0, the surface temperature was 1.49°C higher in Q₃ and Q₄ than in Q₁ and Q₂ (p < 0.05). A similar pattern was observed on the following days, were the lowest temperatures occurred on day 0 but increased by 4.53°C in Q₁ and 6.43°C in Q₄ on day 1. On the afternoon of day 0, the lowest temperatures were observed on this day, but they showed significant increases (p < 0.05) of 4.35°C and 6.43 °C in Q₁ and Q₄, respectively, on day 1. The surface temperatures of the pelvic limb in Q₂ were 0.74, 0.94, and 1.19°C lower than those in Q₁, Q₃, and Q₄, respectively (p< 0.05). This response was observed in the following days since heavier calves (Q₃ and Q₄) maintained higher temperatures until the end of the evaluation period.

Regarding the surface temperature of the lacrimal caruncle (Table 2), in the morning, significant differences (p < 0.05) were observed among the quartiles and over time. On the morning of day 0, Q₄ was found to be 0.9°C higher than Q₂ (p < 0.05), and it can be observed that Q₄ did not show significant variations in the mornings of the days evaluated. On day 1, 2, and 3, buffalo calves from Q₃ and Q₄ showed the highest temperatures when compared to Q₁ and Q₂ (p < 0.05). On days 4 and 5, the temperatures of Q₁, Q₃, and Q₄ were significantly higher than those of Q₂ (p < 0.05). Readings from the afternoon of day 1 showed that Q₁, Q₃, and Q₄ presented a temperature 0.7°C higher than those of Q₂ (p < 0.05). Interestingly, at day 5, calves from Q₂ had lower lacrimal caruncle temperature than the rest of the groups, although a periodically increasing was reported during the 6 days (p < 0.05).

The surface temperature readings from the periocular region showed differences among quartiles and over time (Table 3). Regarding the difference among the four weight groups, in the morning of day 0, temperatures in Q₃ and Q₄ were 1.6°C higher than in Q₁ and Q₂ (p < 0.05). Although a progressive increase in temperature was recorded in all groups [e.g., the temperature of Q₁ on the morning of day 5 increased 0.6°C compared to day 3 (p < 0.05)], on day 5, Q₁, Q₃, and Q₄ were 0.7°C higher than Q₂ (p < 0.05). This response was also observed during the afternoon of day 5, when Q₁ and Q₄ were 0.6°C higher than Q₂ (p < 0.05) while Q₃ and Q₄ recorded the highest values during all the evaluation days (p < 0.05). It is important to mention here that in contrast to Q₁, Q₄ only showed a decrease of 0.91°C in the morning and afternoon of day 1 but then remained without any significant variation until the end of the evaluation period.

Regarding the thermal response of the auditory canal in the water buffalo calves (Table 4), during the morning of the six evaluation days Q₃ and Q₄ had higher temperature than Q₁ and Q₂, (p < 0.05). The same response was observed in the afternoon, and only on day 5, the temperatures of Q₁ and Q₄ were 0.3°C higher than those obtained in Q₂ and Q₃ (p < 0.05). The series of readings taken over time indicated that Q₁ did not present significant differences until day 2, when an increase of 1.91°C was found. In contrast, Q₄ showed a significant decrease of 1.65°C. on day 1, followed by a significant increase of 0.96°C on day 2. Q₁ and Q₄ only presented significant variations toward the end of the afternoon evaluations, but not during the morning study period, when Q₁ tended to show a decrease on day three, with a difference of 1.06°C with respect to day 2. Later, the temperature began to increase, by 0.94 and 0.99°C on days 4 and 5, respectively.

The surface temperature in the nostrils, as shown in Table 5, had the lowest values during the morning of day 0 in Q₁ and Q₂ (p < 0.05). Consecutively, on day 5, Q₃ and Q₄ were 1.4°C higher than Q₁ and Q₂. When considering afternoon temperatures, they had a similar pattern between groups. On day 0, the temperatures of Q₃ and Q₄ were 0.8°C higher than Q₁ and 1.4°C higher than Q₂. This response was maintained during the rest of the days, and when comparing events, observations showed that Q₁, Q₃, and Q₄ had significantly higher morning and afternoon temperatures on day 0 than on day 1 (p < 0.05). The temperature on day 1 was between 0.9 and 1°C higher than on days 2, 3, 4, and 5 (p < 0.05). Regarding Q₂, readings showed that in both the morning and afternoon of day 0 the temperatures were 1.1°C higher than on day 1 (p < 0.05), and that on the same day there was a difference of 1°C compared to day 3 (p < 0.05).

The lacrimal gland showed significant differences on the morning of day 0, where Q₃ and Q₄ were 0.7°C higher than Q₁ and 0.2°C higher than Q₂, respectively (p < 0.05; Table 6). During the rest of the days, temperatures of calves belonging to Q₃ and Q₄ were higher than Q₁ and Q₂ (p < 0.05). On the afternoon of days 0 and 5, the temperatures of Q₁, Q₃, and Q₄ were significantly higher than those in Q₂ (p < 0.05), while on day 2, Q₃ and Q₄ had temperatures 0.7°C higher than those of Q₁ and Q₂ (p < 0.05).

In all groups, the lacrimal gland and lacrimal caruncle had the highest temperatures. In Q₂, the lacrimal caruncle had significantly higher temperatures, with a difference of 0.16°C compared to the lacrimal gland. Similarly, the Q₃ and Q₄ newborns had temperature differences of 0.2–0.11°C between these two areas (Table 7). With respect to the periocular region, all the study animals had high temperatures in the thermal window, except those in Q₄, where no significant differences were found with the auditory canal (36.601 ± 0.02 vs. 36.491 ± 0.03). The windows with the lowest temperatures were the nostrils and the pelvic limbs. In these zones, the buffaloes in Q₁ and Q₂ showed differences of up to 1.47°C.

Regarding the comparison among weights, the lacrimal gland in Q₁ had the lowest temperature, with a difference of up to 0.62°C compared to Q₄. The lacrimal caruncle in Q₁ and Q₂ did not show significant differences within the groups, while there were differences in Q₃ and Q₄, who had the highest temperatures (37.132 ± 0.02 and 37.331 ± 0.02, respectively). The periocular region, auditory canal, nostrils, and pelvic limbs in Q₂ had the lowest temperatures, with significant differences of 1.07, 1.95, 1.74, and 2.67°C, respectively, compared to Q₄.

The temperatures of the lacrimal gland, lacrimal caruncle, periocular region, auditory canal, nostrils, and pelvic limb were 0.3°C higher in the afternoon than the morning (p < 0.05; Figure 2). The lacrimal gland and lacrimal caruncle presented significantly higher values than the periocular region, auditory canal, nostrils, and pelvic limb, in both the morning and afternoon (p < 0.05). Similarly, the temperatures of the periocular region, auditory canal, nostrils, and pelvic limb differed (p < 0.05). Notably, the lowest temperatures were recorded in the pelvic limb (Figure 2).

There was a positive correlation of 0.99 between the surface temperatures of these thermal windows and the time of day (p < 0.0001; Table 8).

In this study, the heavier animals in Q₃ and Q₄–weight >46 kg—maintained higher surface temperatures than the lighter ones in Q₁ and Q₂ (p < 0.05). These results show a relationship between birth weight and the thermoregulation of these calves (1), a finding that has been reported for other species, such as pigs and sheep (22, 23). This could be due to a possible difference in the disposition of energy resources, such as brown adipose tissue and glucose, between a low birth weight or a high birth weight animal. In other words, when a newborn has a low birth weight, its response to compensate for hypothermia will be more limited compared to an animal with a higher weight.

Due to the exposure to cold in mammals, a coordinated response is exerted with the Central Nervous System that produces the activation of the sympathetic nervous system, which promotes the neurosecretion of catecholamines (24). Consequently, a response at the vasomotor level is observed with peripheral vasoconstriction to reduce heat loss. In addition to this, a metabolic response is generated due to the interaction with norepinephrine that produces the breakdown of triglycerides in adipocytes that leads to intracellular activation of cAMP and protein kinase A, for the mitochondrial combustion of substrates and the production of heat (25). The concept of thermoneutrality considers the range where the basal metabolism maintains body temperature (26). In this way, the previous explanation would confirm the fact that an animal with a low birth weight would present low levels of substrates to compensate for heat loss and reach the thermal comfort zone.

The association between basal metabolism and the ability to thermoregulate could explain why a low-birth-weight animal has a lower body temperature. In this sense, Diesch et al. (27) showed that calves born in windy and humid regions with an ambient temperature of 10°C delayed the time for first standing. This could be due to the organism using the energetic reserves to activate thermoregulatory mechanisms, such as vasomotor changes, and shivering and non-shivering thermogenesis (1). These mechanisms depend on the animal's metabolic rate, which is lower per unit area in low-weight neonates and reduces their tolerance to cold temperatures. The basal metabolic rate (rate of energy expenditure per unit time) is higher in smaller animals because they have a higher surface/mass ratio which results in increased corporal heat loss (28–30) in other words, large surface area relative to mass shows a disadvantage in retaining heat (25, 31). Consequently, calves with low body weight may be more susceptible to hypothermia when heat loss exceeds heat production (32).

On the other hand, it is necessary to mention that the surface temperature was significantly lower during days 0, 1, and 2 compared to the rest of the days (p < 0.05). This event is probably due to age since the low critical temperature threshold is higher due to the development of thermoregulatory capacity (2). In other words, the growth of the animal leads to an increase in subcutaneous fat reserves, but also an increase in the thickness of the skin and the length of the hair that allow a greater tolerance to cold (27). Johanson and Berger (33) determined that birth weight can be considered a predictor of perinatal mortality, since animals with higher weights have a reduced incidence of mortality due to hypothermia, as has been reported for cattle in eastern Montana. In this study, animals that weighed over 17.42 kg experienced lower temperature decreases when exposed to cold stress (34). Likewise, for Nili-Ravi buffaloes, Hassan et al. (35) indicated that birth weight had a significant effect on mortality rates, as they found a relationship between low birth weight, poor thermoregulation and deficient dam-calf bonding. It has been suggested that newborns with low birth weight and those with insufficient milk supply from the mother are prone to perform compensatory behaviors, such as allosuckling, to prevent nutritional deficiency (36, 37). Similarly, Chnitier et al. (38) found that newborns lambs with low weights had a higher mortality rate (52.2%) than medium- and high-weight animals (24.4 and 1.8%, respectively). Those authors further determined that during winter, when lambs were exposed to a decrease in environmental temperature, the mortality rate was 23.5%, 14% higher than that during the summer.

As our findings show, energy resources differ depending on the weight of newborns, so those with higher weight have a greater chance of using the energy that is available to them to compensate for temperature changes at birth (39). For example, newborn piglets weighing <1 kg have lower energy resources and greater sensitivity to cold stress during the 1st h of life (40). Similarly, data from newborn Holstein-Friesian calves immersed in water at a temperature of 15–17°C to induce cold stress showed 3-fold increases in catecholamine and blood glucose concentrations, suggesting that hypothermia activates the autonomic nervous system to increase blood glucose levels that contribute to shivering and non-shivering thermogenesis (41, 42).

Another factor linked to birth weight that can be altered is the newborn's vitality and ability to reach the mother's udder (43). Ingesting colostrum helps maintain metabolic resources that aid thermoregulation (7). When this behavior is limited, the ability of young animals to acquire more nutrients is reduced (1). A study by Stanko et al. (44) showed that adding exogenous sources of energy in the form of glucose or colostrum maintained thermostability during the 1st day after birth in newborn Brahman calves exposed to a cold environment (5–25°C). A study by Schmidek et al. (45) in cattle identified factors predisposing to the delayed first intake of colostrum and their implications for the calf mortality rate. Bueno et al. (46) also emphasize the importance of first-colostrum feeding latency, and its effect on calf cortisol levels, with adverse effects on immunoglobulin absorption.

In the present study, regardless of birth weight, the calves presented higher than surface temperatures on days 3, 4, and 5, possibly influenced by the quality of nutrition provided during the 1st days of life. Conte et al. (47) reported that the ingestion of good-quality colostrum can mitigate the negative impact of thermal stress because its lipidic and niacin contents participate in the metabolic and vasodilator responses that aid thermoregulation. The energy provided by colostrum is another element that contributes to obtaining resources and maintaining the thermal balance. According to Silva and Bittar (2), a domestic bovine calf requires 4.6–4.7 Mcal/kg to maintain thermoneutrality. In this regard, studies that have focused on colostrum composition have reported fat, protein, and total solids percentages of 11.31, 8.73, and 25.31%, respectively. These three elements also provide neonates to energy (48). These factors could explain the significantly lower temperatures on days 0, 1, and 2, suggesting a compensatory effect of colostrum ingestion that, in a study of newborn Holstein calves, brought about an increase of ~1°C in prescapular temperature (49). The differences found in the days after birth support the hypothesis that colostrum consumption is essential for newborn ruminants to achieve thermoregulatory success (1). Other experimental data have shown that feeding neonates high-fat diets resulted in an increase in body temperature of 0.5°C under exposure to cold climates, an environmental factor also associated with enhanced muscle development in newborns (50, 51).

Therefore, the results presented affirm that the birth weight of buffalo calves is related to the amount of energy available at birth and can affect their thermoregulatory capacity. In addition, said thermoregulatory capacity is developed according to the age of the animals, which can lead to morphophysiological changes.

According to Mota-Rojas et al. (52), ruminants have some anatomical regions with a high disposition of blood capillaries, arteriovenous anastomoses, and glabrous skin. These areas are known as thermal windows. These regions facilitate heat exchange between the animal and its environment; therefore, exposure to cold temperatures generates a peripheral vasoconstriction response that decreases blood flow (41).

In the present study, temperatures of the lacrimal gland and lacrimal caruncle did not differ between groups Q₁ and Q₂. Moreover, the temperatures recorded in these two thermal windows were higher than those in the other evaluated regions. This suggests that the ocular region could be a useful window for non-invasive assessment of core temperature (23, 53). The difference between these two regions may be attributable to the activation of the sympathetic nervous system (SNS) and the consequent secretion of catecholamines that participate in peripheral vasoconstriction (54–56). This vasomotor response preserves the core temperature in metabolically active organs, such as the brain, heart, and abdomen (28, 57). The effects of this response were described by Shu et al. (58). This body of evidence suggests that the high temperatures exhibited in these two windows are due to their proximity to the brain (59).

Difference between the responses of the lacrimal caruncle and lacrimal gland may also reflect the fact that the latter is irrigated by the infraorbital artery, a structure innervated by the sympathetic branch of the facial nerve (52, 58, 60). Given these anatomical traits, these regions could be helpful for indirectly assessing autonomic nervous system activity as a stress response (61, 62).

The ocular or periocular window can provide a better approximation of core temperature (52, 59). Hoffman et al. (63), for example, established that the IRT readings (mean = 37°C) taken from the eyes of 22 cows and nine calves demonstrated that this is a suitable area for monitoring body temperature, unlike the ear and shoulder. Recently, studies have been done in buffaloes, in which it has been determined that the orbital thermal window could be a validated zone in this specie to record temperature (14, 52, 59).

In contrast to these regions, the auditory canal and pelvic limb registered low temperatures. The recorded values indicate heat loss through the skin due to dermal vasoconstriction and closing of arteriovenous anastomoses (28). IRT readings from the ears and limbs of newborn piglets should be useful to recognize the early stages of hypothermia, although it is important to mention that the concave shape of the auditory canal could alter radiation in that zone (23). For this reason, evaluating the thermal response of the limbs, where blood flow may be reduced due to the vasomotor changes that occur under exposure to cold, could be an alternative (1), even though this possibility requires further study.

The nostrils showed significantly lower temperatures than those recorded in the ocular, lacrimal caruncle, and auricular regions. These results are attributed to the tachypnea response as a means of achieving thermoneutrality (64). In this regard, the increase in oxygen intake compensates for one of the main consequences of hypothermia; namely, hypoxemia (57, 65). Even under conditions of regular respiratory rate, the convective processes associated with the inhaled and exhaled airflows, provide evaporation of fluid present in the nostrils that cause the corresponding cooling. The response in this window reflects increased airflow, which could be helpful in monitoring the breathing patterns of animals (66). On this topic, Kim and Hidaka (67) used IRT and computer vision systems to state breathing patterns in cattle. The authors obtained an accuracy of 76%. However, when evaluating temperatures in this region, elements such as the moist surface must be considered (68), as they can alter measurements due to heat dissipation by evaporation (69).

The IRT technique is useful because it responds to the vascular structure of each thermal window, thus allowing for more reliable approximations to detect changes in the body temperature of newborn buffaloes associated with hypothermic states.

The thermal responses of the six windows evaluated were affected by the time of the day (morning vs. afternoon). In addition to this, it was possible to observe that the temperatures of all the thermal windows presented positive correlation with the time of day (p < 0.05). The temperature readings of the buffalo calves were significantly higher in the afternoon. A possible explanation for this is the effect of environmental factors. Villanueva-García et al. (23), Mota-Rojas et al. (69), and Kozat (15) mentioned that elements such as wind and rain can cause hypothermia in newborns, whereas solar radiation has the opposite effect (70).

In the six thermal windows, during the afternoon, heat loss by evaporation is attenuated, so the amount of blood in the dermal tissue increases due to the activation of the SNA (71, 72). A study of Murrah water buffaloes reported that exposure to 45°C during the summer increased blood flow by 5–17 units (in the dorsal, abdominal, and auricular areas), as well as the animals' respiratory frequency (73). Marai and Heeeb (74) mention that the characteristics for the water buffalo to remain in an optimal productive state are an air temperature of 13–18°C, relative humidity of 55–60%, and an air speed of 5–8 km/h. An environmental temperature lower than 15°C is a crucial factor that could challenge thermoregulatory mechanisms in newborn, while wind speed and solar radiation could affect the IRT reading. This could be a limitation within the present study and an element to consider for future studies.

The ability of water buffaloes to maintain thermoneutrality during periods of intense heat also depends on characteristics such as the presence of a thick epidermis, low hair density, and abundant melanin (14, 52). Interestingly, these features may help in a certain way to increase resilience to cold because the difficulty that these animals experience in dissipating heat could be beneficial by preventing greater heat loss through evaporation (73).

According to our findings, the time of the day influences the thermostability of newborn water buffaloes (75); therefore, this must be considered an important factor for the perinatal management of the offspring of this species.