Physiological indicators of Caribou calf health and compensatory responses to physiological perturbations Authors: Kristin Denryter, Jacob Pryor, Amanda Puype Published: 2026 ABSTRACT AbstractCaptive husbandry is a cornerstone of modern Caribou (Rangifer tarandus subspecies) conservation (e.g., captive rearing) and research in North America. Husbandry protocols have long emphasized feeding and nutrition programs, but information on health metrics—such as vital signs essential for evaluating animal health—is more limited, especially for young Caribou. We documented vital signs (i.e., body temperature, heart rate, and respiratory rate) of n = 8 hand-reared Caribou calves, each day, from 0 to 23 wk of age. We used multilevel, mixed-effects regression models to evaluate: (i) how vital signs changed as calves grew; and (ii) whether calves compensated (as indicated by changes in vital signs) for physiological perturbations (i.e., changes from homeostasis such as elevated body temperature). We excluded data from 1 calf that had a chronic illness; among healthy, juvenile Caribou in our study, mean (±SE) heart rate was 89 ± 0.7 beats/min, respiratory rate was 19 ± 0.2 breaths/min, and body temperature was 38.9 ± 0.01 °C. Heart rate, respiratory rate, and body temperature decreased as calf age and body mass increased; increasing calf age was a better predictor of heart rate, respiratory rate, and body temperature than body mass. Caribou calves exhibited physiological compensation for increasing body temperature by increasing heart rate and respiratory rate. Heart rate also increased as a function of increasing respiratory rate. Our findings for age-specific vital signs in Caribou calves can serve as baselines for future physiology studies and captive-rearing efforts that may enhance early detection and treatment of illness in captive Caribou. Captive husbandry is a cornerstone in the conservation and research of many ungulates in North America, especially Caribou (Rangifer tarandus subspecies), which are imperiled throughout much of their range (Vors and Boyce 2009; Johnson et al. 2015). Caribou have been maintained in captivity for conservation education and outreach and for captive breeding or maternal penning (Adams et al. 2019; Mueller et al. 2022; Lamb et al. 2024), as well as for studies of physiology and nutrition (Fancy and White 1987; Rominger et al. 1996; Parker and Barboza 2013). Tractable, tamed Caribou have proven to be an invaluable asset to research because they afford observers the opportunity to collect direct observations of nuanced behaviors in close proximity, thereby providing an unmatched level of detail into forage selection and nutrient intakes that could not be gathered by other means (Rominger and Oldemeyer 1990; Rominger et al. 1996; Denryter et al. 2017, 2020, 2022). While some studies of captive animals reveal differences from their wild counterparts (Turko et al. 2023), those findings are not universal. For example, tame ungulates retain innate behaviors and have been shown to be equivalent to their wild relatives for studies of foraging and nutrition (Buechner 1950; McMahan 1964; Bergerud and Nolan 1970; Wallmo and Neff 1970; Bergerud 1972; Olson-Rutz and Urness 1987; Spalinger et al. 1997). Technological advances provide new ways to collect data remotely (e.g., Trondrud et al. 2021; Høy-Petersen et al. 2023), but in some cases, there is no substitute for information gained by observing tame animals in natural environments (Parker et al. 1999; Cook et al. 2016; Denryter et al. 2022), making them a vital tool for some types of research. Success in rearing animals in captivity is contingent upon implementing appropriate protocols for the target species. Protocols for captive rearing of North American ungulates largely emphasize nutrition and growth rates (Schwartz et al. 1976; Hobbs and Baker 1979; Pekins and Mautz 1985; Parker and Wong 1987; Shochat and Robbins 1997; Parker and Barboza 2013). The nutrition-centric nature of existing protocols stemmed from an urgent need to understand nutritional management of captive ungulates because improper nutritional management contributed to high rates of mortality and morbidity in captive cervids (Shochat and Robbins 1997; Shochat et al. 1997). Emphasis on nutrition may have overshadowed understanding other elements that may be important in captive-rearing practices such as physiological parameters or vital signs that can be used to monitor and assess health. Vital signs including heart rate, respiratory rate, and body temperature are universally important indicators of homeostasis. Vital signs are essential in the management and husbandry of captive animals because they provide real-time information on health and homeostasis and can be collected with readily available equipment and by personnel with little training required for proficiency. Perturbations in vital signs often precede serious medical events (Cretikos et al. 2008) and hence close monitoring of vital signs can be an early clue of oncoming complications. Despite their importance, relatively little information has been published on heart rate, respiratory rate, and body temperature of juvenile cervids from birth through weaning. Recently, values of vital signs for hand-reared Moose (Alces alces) calves in Alaska were reported (Thompson et al. 2023), but to our knowledge, information on vital signs of Caribou calves is limited to animals <35 days old (Hissa et al. 1981; Soppela et al. 1986) or 12–18 mo old (Hammel et al. 1972; Yousef and Luick 1975; Cuyler and Øritsland 1986). Thus, there is a gap in information about vital signs in Caribou calves between those periods when they are rapidly growing (Parker and Barboza 2013). Limited age- and species-specific information on vital signs of young Caribou forces interpretation of vital signs relative to those of other species or older conspecifics. Often vital signs from livestock are used for comparison, but whether livestock values are appropriate for comparison with wild taxa is unclear as artificial selection may have altered livestock physiology (Robbins 1993). Regardless, age-specific vital signs in Caribou would be useful as baseline values in health assessments. Limited baseline information on vital signs in Caribou calves also inhibits our understanding of the ability of Caribou calves to respond to and tolerate homeostatic perturbations. Homeostatic perturbations, sometimes called physiological stress, occur when the function of a body system deviates from normal (Mantovani 1995). In these cases, the body attempts to correct for the perturbation through a compensatory response such as releasing insulin in response to high levels of glucose in the bloodstream. Compensatory responses may not yet be fully developed in neonates that are transitioning from fetal to adult physiology (Palmer 1992; Doherty et al. 2023). Although calves have compensatory mechanisms to increase heat production, such as through shivering and nonshivering thermogenesis, they typically are more sensitive to thermal stress than adult animals (Broucek et al. 2009; Roland et al. 2016; Barnes et al. 2017). Hence, compensatory mechanisms for maintenance of body temperature in calves may be less well developed than in adults. As cervids, Caribou rely on respiration and panting, rather than sweating, as a primary physiological mechanism to dissipate heat (Parker and Robbins 1984). Heat dissipation also may be facilitated by vasodilation, which could result in a compensatory increase in heart rate to counter apparent low blood volume (hypovolemia) (Davies and Maconochie 2009). Thus, physiological compensation may be signaled by an increase in respiratory rate, heart rate, or both as a function of increasing body temperature (Sokolov and Grechkina 1998; Davies and Maconochie 2009). As part of a larger effort focused on rearing calves for studies of nutrition and physiology, we collected information on vital signs (i.e., heart rates, respiratory rates, and body temperature) of Caribou calves. Because information on vital signs is valuable to health assessment, our primary objective herein was to provide descriptive statistics for vital signs of Caribou calves from birth to weaning. We also aimed to evaluate whether 40.0 °C, which is a threshold used to indicate fever in domestic lambs (Ovis aries; Fielder 2022), was a reasonable threshold to indicate fever in Caribou calves. In addition to providing valuable descriptive statistics, we also tested 2 hypotheses. First, we hypothesized that Caribou calves mature physiologically from birth to weaning. We predicted that vital signs would be highest in neonatal calves and decrease over time, being lowest near weaning in autumn when growth and metabolism slow. Second, we hypothesized that Caribou calves compensate for homeostatic perturbations in body temperature and predicted that as body temperature increased, Caribou calves would increase heart and respiratory rates, and that heart rate would increase to support increased respiratory rates. The extent that Caribou calves compensate for perturbations would have implications to husbandry. Methods All research followed American Society of Mammalogist guidelines (Sikes et al. 2016); animal facilities were inspected and protocols were reviewed and approved by the Alaska Department of Fish and Game Animal Care and Use Committee (Protocol 0118-2023-013). In May 2023, we began hand-rearing female Alaskan barren-ground Caribou calves (R. t. granti) at the Foraging Ecology and Wildlife Nutritional Analysis (FaWNA) Lab of the Alaska Department of Fish and Game for use in foraging and nutrition studies to better understand factors contributing to poor calf recruitment and population declines in free-ranging Caribou herds in Alaska. Caribou calves were born in the wild and brought into captivity at ∼1–5 days old for hand rearing. Age was estimated following methods of Adams et al. (1995) based on conformation and posture (e.g., hunched or fully upright), whether the umbilicus was wet or dry, whether the hips were bony or had filled out at all, and hoof wear. We modified hand-rearing protocols previously developed for Caribou at the Robert G. White Large Animal Research Station at the University of Alaska, Fairbanks (Parker and Barboza 2013). The primary modification to hand-rearing protocols was the addition of MooseGro milk replacer (Grober Nutrition, Cambridge, Ontario), which was mixed at a ratio of 275 g powder to 800 g water, which provided protein and fat levels that were consistent with composition of Caribou milk (Parker and Barboza 2013). We maintained calves on this ration until they were 7 wk old, when we began mixing 3 parts MooseGro to 1-part Zoologic Milk Matrix 30/52 (PetAg, Hampshire, IL; mixed as 26.3 g powder in 73.7 g of water) by volume, to address a low calcium to phosphorus ratio (∼1:1) that we discovered in the MooseGro milk replacer. We offered calves milk replacer on a body-mass and age-based feeding schedule, offering calves 10% of their body mass per day in milk replacer for the first 8 wk and decreasing the daily amount of milk replacer offered by 100 g each week through weaning (e.g., if 2,000 g/d in week 9, then decrease to 1,900 g/d in week 10). Caribou calves had ad libitum access to native vegetation (e.g., Taraxacum spp.) and we supplemented calves daily with deciduous shrubs (e.g., Salix spp., Betula spp.), lichens (e.g., Cladina spp., Flavocetraria spp., Usnea spp.), Calf Manna (Manna-Pro, St. Louis, MO), and a pelleted ration (D-Ration, Alaska Mill and Feed, Anchorage, AK) developed specifically for Rangifer (Parker and Barboza 2013). Beginning in August, we provided Goat Minerals (Manna Pro, St. Louis, MO) to calves at a rate of 7–14 g/animal/d, and in September and October, we added 60 g thiamin (Su-Per B-1, Gateway, Holly, CO) and 120 g of probiotic powder (Gut Candy, Precision Feed Technologies, Lawton, OK) to pelleted feed each day to aid calves in transitioning to solid feed (D-Ration) during weaning. We also provided alfalfa hay at a rate of 0.25–0.50 kg/animal/d beginning in October when natural forages in the enclosures had senesced and were no longer available. Caribou calves were housed in 3 different enclosures during the study. First, we housed calves in an indoor quarantine facility, which was part of a former milking parlor with concrete floors that we covered with 2 cm thick rubber livestock mats and aspen shavings. Access to sunlight and fresh air was provided through screened windows and doors. The indoor facility did not have mechanical climate controls and temperatures were equilibrated to ambient temperatures by opening large screened doors and windows and using fans to circulate air. Ambient temperatures and precipitation in the vicinity of the facility were recorded at the University of Alaska Fairbanks Matanuska Experiment Farm National Weather Service weather station during our study and ranged from −12.2 to 24.4 °C (10.6 ± 0.4 °C) and 0.26–4.06 cm (0.027 ± 0.004 cm). At ∼3 wk old, we moved calves to an outdoor facility comprising a barn with 9 m2 stalls, each with access to small grass pastures (∼30 m2 each). Calves were divided into groups of 2–4 based on temperament and bonding and separated into 3 stalls. Each stall had a rocky substrate to promote drainage, overlaid with aspen shavings for bedding. We changed bedding and cleaned stalls (including feed and water troughs) daily and removed fecal material after each bottle feeding. We installed fans with misters in the barn to help keep calves cool during periods of high ambient (>16°C) temperature (Denryter, 2026). At 14 wk, we moved calves to a 4,000 m2 enclosure on natural pasture with shelters. As in other housing, we provided wood shavings as bedding material in shelters and replaced bedding and cleaned shelters of excrement after feedings. Calves were comingled throughout the day in all facilities and separated into their groups only at feeding times or when an animal became ill, in which case the entire group of that animal was isolated. We weighed calves at the same time each day using a platform scale and load bars (S3 Weigh Scale Indicator, Tru-Test, Auckland, New Zealand; MP 800 load bars, Datamars Livestock, Temple, TX). We trained calves to load onto the scale using a variety of treats including lichens, cereal (Cheerios, General Mills, Minneapolis, MN), and birch and willow leaves as positive reinforcement. Some calves were amenable to halter and lead rope training and were led to the scale daily. Once atop the platform, we positioned each calf so that it stood squarely across the 2 load bars for even weight distribution. All calves were rewarded with additional treats for proper positioning on the scale. We recorded body mass of each calf to the nearest 0.5 kg. We collected rectal (hereafter, body) temperature, heart rate, and respiratory rate on calves 1–3 times daily, and more frequently if we suspected illness, such as when a calf refused a meal, appeared lethargic, or exhibited other behavioral changes. Generally, we collected vital signs more frequently in younger calves that were more cooperative than older calves. We recorded vital signs when the animals were bedded, typically 20–30 min after a bottle feeding, to allow the calves to settle and ensure cooperation, which also allowed us to collect standardized measurements of vital signs while calves were at rest. Trying to collect vital signs too soon after a feeding was unsuccessful as calves were generally uncooperative following the excitement of a meal. We obtained resting heart rates of calves by palpating the axillary artery, listening to the number of heart beats through a stethoscope (Littman Classic IIITM Monitoring Stethoscope, 3M, Maplewood, MN), or by counting the number of pulsations of the jugular vein in the neck, when visible, in 1 min. We determined respiratory rate by counting a complete cycle of the chest rising and falling as a single respiration and recording the number of respirations in 1 min. We counted heart and respiratory rates to the nearest whole number. We measured body temperature using individually, color-coded thermometers (Vicks SpeedRead V912US Digital Thermometer, Kaz USA Inc., Marlborough, MA) that were lubricated (MedPride Lubricating Jelly, Bayonne, NJ) and inserted into the rectum. Thermometers gave temperature readings to 0.1 °C accuracy in <3 s. If the thermometer did not provide a reading in <3 s, it generally would not provide an accurate reading, so we removed the thermometer, reset it, and tried again. We cleaned and disinfected thermometers using disinfectant wipes after each use. Calves were treated for illness and injury according to treatment plans prescribed by veterinarians. Calves with bacterial infections were treated with various antibiotics, depending on veterinary recommendations (including sensitivity testing), and were supplemented with oral probiotics (2.5 g, Goats Prefer Probiotic Plus Paste (given directly into the mouth), Vets Plus, Menomonie, WI), and thiamin injections (10–20 mg/kg, once daily, various manufacturers) to support energy metabolism and prevent issues with development of normal rumen flora and thiamin deficiency-induced polioencephalomalacia (Denryter and Beckmen 2024). When rectal temperature exceeded 40.0 °C, calves received 0.5 ml/45 kg of flunixin meglumine (Banamine, Merck, Rahway, NJ) up to every 6 h or 1 ml/45 kg every 12–24 h, usually intramuscularly unless the animal had a catheter, in which case we administered flunixin meglumine intravenously. Calves receiving flunixin meglumine also received 0.5–1 g sucralfate (Nostrum Labs, Kansas City, MO) mixed in milk replacer 3 times per day and 20 mg omeprazole (various manufacturers) mixed in milk replacer once per day. We excluded data from analyses from calves while they were being treated with flunixin meglumine or antibiotics, which included excluding the data from 1 animal entirely because she was treated consistently for several infections, including septic arthritis, for almost the entire duration of the study from June through September. Statistical analyses We conducted all statistical analyses in Stata 14 (StataCorp LP, College Station, TX). Given the repeated measures on the same individuals in our study, we used linear and generalized linear multilevel mixed-effects regression (hereafter, multilevel models) with an independent covariance matrix using the “mixed” and “meglm” commands, respectively, to evaluate relationships between predictor and response variables (Skrondal and Rabe-Hesketh 2004). We evaluated models with a random intercept and with and without random slope to track individual and used Akaike’s Information Criterion adjusted for small sample size (AICc) to rank competing models (Burnham and Anderson 2002). To test our hypothesis that Caribou calves would undergo physiological maturation from birth to weaning, we evaluated heart rate, respiratory rate, and rectal temperature as functions of calf age in days or calf body mass (kg). We evaluated competing models for uninformed parameters (ΔAICc ≤2 per additional parameter over the top model; Arnold 2010; Leroux 2019; Sutherland et al. 2023). STATA automatically calculates the number of parameters, which we retrieved using the “ereturn list” postestimation command (Skrondal and Rabe-Hesketh 2004). Owing to different numbers of observations for body mass and age, we subset the data to ensure that we had the same number of samples for comparison within each model group. Because best-approximating (lowest AIC values) models can include variables with 85% confidence intervals that exclude zero (rather than just 95% confidence intervals), we followed recommendations of Arnold (2010) and Sutherland (2023) to report 85% confidence intervals in addition to 95% confidence intervals to avoid variable-selection ambivalence (i.e., rejecting a variable as important based on a 95% confidence interval when the 85% confidence interval is adequate for its inclusion in a best-approximating model). To test our hypothesis that Caribou calves compensated for homeostatic perturbations, we evaluated whether heart rates and respiratory rates increased as a function of body temperature, and whether heart rate increased as a function of respiratory rate. We assumed increases in heart rate would be needed to support increased rates of respiration, which allows for heat dissipation. As with our earlier analyses, we used mixed-effects regression with random intercepts for individual and with and without random slopes for individual and ranked models using AICc as described above. Results Between May and October 2023, we hand-reared n = 8 female Caribou calves. Initial body mass of calves ranged from 5.3 to 8.3 kg (6.5 ± 0.3 kg; mean ± SE) and final body masses at 23 wk ranged from 45.5 to 57.7 kg (52.7 ± 1.4 kg; Supplementary Data SD1). During the study, we collected n = 1,746 measurements of heart rate, n = 1,938 of respiratory rate, and n = 2,097 of body temperature on Caribou calves. Values for heart rate ranged from 40 to 300 heartbeats per min, for respiratory rate from 8 to 140 breaths per min, and for body temperature from 34.0 to 41.2 °C (Tables 1–3; Supplementary Data SD1). Minimum heart rate and respiratory rate occurred in older (>100 days) calves at rest (Fig. 1; Supplementary Data SD1). Minimum body temperature was recorded in a calf <48 h old that became hypoglycemic and hypothermic secondary to sepsis but was treated and recovered fully. Maximum heart rate, respiratory rate, and body temperature occurred in 2 calves that developed infections, were treated, and recovered fully. Data from sick calves were not used in estimation of mean values or in regression analyses. Mean (± SE) heart rate for healthy, juvenile Caribou in our study was 89 ± 0.7 beats/min, respiratory rate was 19 ± 0.2 breaths/min, and body temperature was 38.9 ± 0.01 °C. Relations between calf age and heart rate (A), respiratory rate (B), and body temperature (C) shown as mean (dashed blue line) and 95% confidence intervals (blue shading) from top linear and generalized linear multilevel regression models (Tables 4 and 5) for n = 7 Caribou calves hand-reared in captivity in Alaska from May to October 2023. Table 1Summary statistics of heart rate (beats/min) of n = 7 healthy female Caribou calves from the 1st week of life (0–7 days) through the 23rd week of life (154–161 days) at the Foraging Ecology and Wildlife Nutritional Analysis Lab in southcentral Alaska from May to October 2023.Age in weeks n MeanSDMinMaxSE95% CI 1 1311817.9881445.0108–128 2 3311116.7801402.9105–117 3 6811913.7821521.7116–123 4 6711813.2801441.6115–122 5 9311811.3901501.2115–120 6 4911310.6921401.5110–116 7 691047.6841200.9102–106 8 831058.8801241.0103–107 9 81939.4721161.091–95 10 61919.3641081.288–93 11 718510.1641081.283–88 12 778210.0641081.180–85 13 577911.0601041.577–82 14 498211.0641081.678–85 15 47738.060921.271–75 16 42698.044881.266–71 17 49669.6521021.464–69 18 45617.448781.159–63 19 42627.348761.159–64 20 42566.548721.054–58 21 48557.144721.053–57 22 50548.240761.252–56 23 10477.540602.443–52 Summary statistics of heart rate (beats/min) of n = 7 healthy female Caribou calves from the 1st week of life (0–7 days) through the 23rd week of life (154–161 days) at the Foraging Ecology and Wildlife Nutritional Analysis Lab in southcentral Alaska from May to October 2023. Table 2Summary statistics of respiratory rate (respirations/min) of n = 7 healthy female Caribou calves from the 1st week of life (0–7 days) through the 23rd week of life (154–161 days) at the Foraging Ecology and Wildlife Nutritional Analysis Lab in southcentral Alaska from May to October 2023.Age in weeks n MeanSDMinMaxSE95% CI 1 33266.316481.124–28 2 73288.216601.026–30 3 86254.118360.424–26 4 74266.014460.724–27 5 111245.016400.523–25 6 63225.914400.721–24 7 73204.312320.519–21 8 80193.710280.418–19 9 80184.510280.517–19 10 62183.110240.417–18 11 71152.910240.315–16 12 81163.69280.415–17 13 58142.610240.314–15 14 49164.310280.615–17 15 47132.110160.312–14 16 44133.510240.512–15 17 48165.410320.814–17 18 45154.710280.714–17 19 42142.210180.313–14 20 43132.58180.413–14 21 48143.310260.513–15 22 50132.68200.412–13 23 10101.88120.69–11 Summary statistics of respiratory rate (respirations/min) of n = 7 healthy female Caribou calves from the 1st week of life (0–7 days) through the 23rd week of life (154–161 days) at the Foraging Ecology and Wildlife Nutritional Analysis Lab in southcentral Alaska from May to October 2023. Table 3Summary statistics of body (rectal) temperature (°C) of n = 7 healthy female Caribou calves from the 1st week of life (0–7 days) through the 23rd week of life (154–161 days) at the Foraging Ecology and Wildlife Nutritional Analysis Lab in southcentral Alaska from May to October 2023.Age in weeks n MeanSDMinMaxSE95% CI 1 3138.70.4538.039.50.0838.6–38.9 2 3938.90.2438.439.40.0438.8–39.0 3 6839.00.1838.639.50.0239.0–39.1 4 6439.10.2138.239.90.0339.1–39.2 5 9539.10.1738.639.50.0239.0–39.1 6 7639.10.2438.339.90.0339.0–39.1 7 8039.00.1838.339.50.0239.0–39.1 8 9739.00.2038.339.50.0239.0–39.0 9 7939.00.2038.439.40.0238.9–39.0 10 6239.00.2138.439.40.0338.9–39.0 11 6339.00.2138.639.40.0338.9–39.0 12 7439.00.2438.339.50.0338.9–39.0 13 4838.90.2138.439.40.0338.8–38.9 14 4938.90.1538.639.20.0238.8–38.9 15 5038.90.2338.539.60.0338.8–38.9 16 4438.70.1838.339.10.0338.7–38.8 17 4938.70.1738.239.20.0238.7–38.8 18 4538.80.1938.439.30.0338.8–38.9 19 4238.70.2238.239.10.0338.6–38.7 20 4438.70.1938.239.00.0338.6–38.8 21 4938.60.2738.239.40.0438.6–38.7 22 5038.50.2437.838.90.0338.5–38.6 23 1038.30.2637.838.80.0838.1–38.4 Summary statistics of body (rectal) temperature (°C) of n = 7 healthy female Caribou calves from the 1st week of life (0–7 days) through the 23rd week of life (154–161 days) at the Foraging Ecology and Wildlife Nutritional Analysis Lab in southcentral Alaska from May to October 2023. Age and body mass of Caribou calves were highly correlated (Spearman’s correlation coefficient = 0.985, P < 0.001) and were not used together in any models to explain changes in vital signs. Calf age (in days) best explained heart rate, respiratory rate, and body temperature (Fig. 1A–C). The top models for heart rate (Wald Χ2 = 1,249.55, P < 0.001), respiratory rate (Wald Χ2 = 659.59, P < 0.001), and body temperature (Wald Χ2 = 57.03, P < 0.001) included both a random intercept and random slope for individual calf (Tables 4 and 5). For every 1 d increase in age, heart rate decreased by 0.5 beats/min, and there was a 0.1% decrease in respiratory rate, and a 0.0001% decrease in body temperature (Tables 4 and 5). Table 4Ranks of candidate multilevel regression models for heart rate (heartbeats/min), respiratory rate (respirations/min), and body (rectal) temperature (°C) of n = 7 healthy female Caribou calves from May to October 2023 at the Foraging Ecology and Wildlife Nutritional Analysis Lab in southcentral Alaska.Model k n LLAICcΔAICcWi Heart rate ∼ Age + RSa 51240–4601.3219212.6910.0001.000 Heart rate ∼ Age 41240–4609.9309227.89215.2010.000 Heart rate ∼ Body mass + RSa 51240–4631.5499273.14660.4550.000 Heart rate ∼ Body mass 41240–4652.8669313.765101.0740.000 Heart rate (null model) 31240–5704.09811414.2162201.5250.000 Respiratory rate ∼ Age + RSa 51360–3864.9907740.0250.0000.991 Respiratory rate ∼ Body mass + RSa 51360–3870.1917750.42710.4020.005 Respiratory rate ∼ Age 41360–3871.5297751.08711.0620.004 Respiratory rate ∼ Body mass 41360–3878.4457764.92024.8950.000 Respiratory rate (null model) 31360–4372.3848750.7861010.7610.000 Body temperature ∼ Age + RSa 51288187.360–364.6720.0001.000 Body temperature ∼ Body mass + RSa 51288164.321–318.59546.0770.000 Body temperature ∼ Age 41288154.365–300.69963.9730.000 Body temperature ∼ Body mass 41288132.691–257.351107.3220.000 Body temperature (null model) 31288–92.713191.446556.1180.000 Heart rate ∼ Body temperature + Age 51168–4225.4748460.9990.0001.000 Heart rate ∼ Body temperature 41168–5144.44010296.9131835.9140.000 Heart rate (null model) 31168–5377.67310761.3672300.3680.000 Heart rate ∼ Body temperature + Age + RSa , b –––––– Heart rate ∼ Body temperature + RSa , b –––––– Respiratory rate ∼ Body temperature + Age 51164–3217.5656445.1810.0000.737 Respiratory rate ∼ Body temperature + Age + RSa 61164–3217.5866447.2442.0630.263 Respiratory rate ∼ Body temperature 41164–3487.2966982.627537.4460.000 Respiratory rate ∼ Body temperature + RSa 51164–3487.2916984.633539.4520.000 Respiratory rate (null model) 31164–3603.9867213.993768.8120.000 Heart rate ∼ Respiratory rate + Age + RSa 61210–4379.2688770.6070.0001.000 Heart rate ∼ Respiratory rate + Age 51210–4392.5018795.05224.4450.000 Heart rate ∼ Respiratory rate 41210–5214.93510437.9031667.2960.000 Heart rate ∼ Respiratory rate + RSa 51210–5214.93510439.9201669.3130.000 Heart rate (null model) 31210–5555.35211116.7242346.1170.000All variables are continuous including age (days) and body mass (kg). Columns are the number of parameters in the model (k), number of samples (n), log likelihood of the model (LL), Akaike Information Criterion corrected for small sample size (ΔAICc), and weight of inference (ωi).aModel includes a random slope (RS), in addition to a random intercept, for animal identifier.bModel did not converge. Ranks of candidate multilevel regression models for heart rate (heartbeats/min), respiratory rate (respirations/min), and body (rectal) temperature (°C) of n = 7 healthy female Caribou calves from May to October 2023 at the Foraging Ecology and Wildlife Nutritional Analysis Lab in southcentral Alaska. All variables are continuous including age (days) and body mass (kg). Columns are the number of parameters in the model (k), number of samples (n), log likelihood of the model (LL), Akaike Information Criterion corrected for small sample size (ΔAICc), and weight of inference (ωi). Model includes a random slope (RS), in addition to a random intercept, for animal identifier. Model did not converge. Table 5Results from top multilevel regression models, as indicated by Akaike Information Criteria corrected for small sample size (Table 4), characterizing relations between heart rate (heartbeats/min), respiratory rate (respirations/min), age (in days), body (rectal) temperature, and body mass (in kg) of n = 7 healthy female Caribou calves from May to October 2023 at the Foraging Ecology and Wildlife Nutritional Analysis Lab in southcentral Alaska; all variables were continuous and all models included a random intercept for individual.ModelParameterCoef.SE z P 95% CI85% CI Heart rate ∼ Age + RSa Age−0.5360.02−35.4<0.001−0.566–(−0.507)−0.558–(−0.515)Intercept1291.8968.1<0.001125–132126–131 Respiratory rate ∼ Age + RSa , b Age−0.0062.22 × 10−4−25.7<0.001−0.006–(−0.005)−0.006–(−0.005)Intercept3.300.01277<0.0013.28–3.333.29–3.32 Body temperature ∼ Age + RSa , b Age−9.25 × 10−51.23 × 10−5−7.55<0.001−1.17 × 10−4–(−6.85 × 10−5)−1.10 × 10−4–(−7.49 × 10−5)Intercept3.676.92 × 10−45302<0.0013.67–3.673.67–3.67 Heart rate ∼ Body temperatureb Body temperature0.090.025.94<0.0010.061–0.1210.069–0.114Age−0.0069.32 × 10−5−66.4<0.001−0.006–(−0.006)−0.006–(−0.006)Intercept1.360.602.250.0240.176–2.540.490–2.23 Respiratory rate ∼ Body temperatureb Body temperature0.070.032.220.0270.008–0.1340.025–0.117Age−0.0051.92 × 10−3−26.0<0.001−0.005–(−0.005)−0.005–(−0.005)Intercept0.471.250.380.708−1.99–2.93−1.34–2.28 Heart rate ∼ Respiratory rateb Respiratory rate0.0037.29 × 10−43.74<0.0010.001–0.0040.002–0.004Age−0.0062.01 × 10−4−30.9<0.001−0.007–(−0.006)−0.007–(−0.006)Intercept4.870.02207<0.0014.82–4.914.83–4.90aModel includes a random slope for individual in addition to a random intercept.bResults in the table are coefficients for generalized linear regression models with log link (gamma family) but are reported in text and figures using exponentiated coefficients. Results from top multilevel regression models, as indicated by Akaike Information Criteria corrected for small sample size (Table 4), characterizing relations between heart rate (heartbeats/min), respiratory rate (respirations/min), age (in days), body (rectal) temperature, and body mass (in kg) of n = 7 healthy female Caribou calves from May to October 2023 at the Foraging Ecology and Wildlife Nutritional Analysis Lab in southcentral Alaska; all variables were continuous and all models included a random intercept for individual. Model includes a random slope for individual in addition to a random intercept. Results in the table are coefficients for generalized linear regression models with log link (gamma family) but are reported in text and figures using exponentiated coefficients. Heart rate and respiratory rate both increased as a function of increasing body temperature and the best models for these included calf age as a covariate. When the effect of age was controlled for, heart rate increased by ∼10% for every 1 °C increase in body temperature (Wald Χ2 = 7499.07, P < 0.001; Fig. 2A and B) and respiratory rate increased by ∼7% (Wald Χ2 = 1118.48, P < 0.001; Tables 4 and 5; Fig. 3A and B). Heart rate also was a function of respiratory rate and when age was controlled for, every 1-unit increase in respiratory rate resulted in an ∼0.3% increase heart rate (Wald Χ2 = 1193.33, P < 0.001; Tables 4 and 5; Fig. 4A and B). Relations between actual heart rate and (exponentiated) heart rate predicted from generalized linear multilevel regression models with body temperature and age as covariates (A; Tables 4 and 5) and heart rate as a function of body temperature (B) for n = 7 Caribou calves hand-reared in captivity in Alaska from May to October 2023. The dashed blue line in (A) indicates where a 1:1 relationship (perfect agreement) between predicted and actual values would occur. Relations between actual respiratory rate and (exponentiated) respiratory rate predicted from generalized linear multilevel regression models with body temperature and age as covariates (A; Tables 4 and 5) and respiratory rate as a function of body temperature (B) for n = 7 Caribou calves hand-reared in captivity in Alaska from May to October 2023. The dashed blue line in (A) indicates where a 1:1 relationship (perfect agreement) between predicted and actual values would occur. Relations between actual heart rate and (exponentiated) heart rate predicted from generalized linear multilevel regression models with respiratory rate and age as covariates (A; Tables 4 and 5) and heart rate as a function of respiratory rate (B) for n = 7 Caribou calves hand-reared in captivity in Alaska from May to October 2023. The dashed blue line in (A) indicates where a 1:1 relationship (perfect agreement) between predicted and actual values would occur. Discussion The neonatal period in animals is the time of greatest vulnerability (Roffe 1993; Adams et al. 1995; Jenkins and Barten 2005; Gustine et al. 2006), even in captive settings (Gayle et al. 1998; Figueroa et al. 2024). Being able to recognize signs of illness early, based on changes in vital signs or otherwise, is critical for initiating treatment and improving calf outcomes. We documented 3 vital signs—body temperature, heart rate, and respiratory rate—in captive female Caribou calves from 0 to 23 wk old to determine normal ranges of values for these vital signs and characterize their changes as calves grew. All vital signs that we measured were highest, on average, in neonatal Caribou and declined over time as Caribou grew, consistent with our hypothesis that Caribou mature physiologically from birth to weaning. Consistent with our second hypothesis, Caribou calves compensated for homeostatic perturbations in body temperature by altering heart rate and respiratory rate (Sokolov and Grechkina 1998; Davies and Maconochie 2009). To our knowledge, detailed information on heart rate, respiratory rate, and body temperature of Caribou calves covering the entire period from neonatality through weaning was not available prior to our study. Our findings can serve as baseline metrics for vital signs that are species and age-specific, which can aid health assessments of Caribou calves in captivity. To evaluate changes in vital signs relative to calf growth, we conducted intensive daily monitoring of captive Caribou calves and determined ranges of body temperature, heart rate, and respiratory rate for Caribou calves of different ages (Tables 1–3). All vital signs were greatest and most variable in calves during early life when they had bacterial infections, before their immune systems had matured (Cortese 2009; Cardoso et al. 2021), but these were excluded from our analyses. As calves increased in age and body mass, heart rate, respiratory rate, and body temperature all decreased, consistent with predictions of allometric scaling (Kleiber 1932; da Silva et al. 2006) and reports for other taxa (McTaggart-Cowan and Wood 1955; Hammel et al. 1972; Yousef and Luick 1975; Cuyler and Øritsland 1986; Soppela et al. 1986; Thompson et al. 2023). We assumed that age and body mass would be equally good predictors of changes in vital signs, but this was not supported by our models. Age was a better predictor than body mass of changes in heart rate, respiratory rate, and body temperature. Although generally body mass increased linearly with age, some animals experienced slower mass gain than others, which may be why age better explained changes in vital rates than body mass. Age and body mass, however, were highly correlated and thus we caution against putting too much confidence in 1 explanatory variable over the other. Estimates of age-specific vital signs are not well established for Caribou, so we made comparisons based on limited data on vital signs reported for Reindeer (R t. tarandus, R. t. platyrhynchus, R. t. fennicus), Columbian Black-tailed Deer (Odocoileus hemionus columbianus), and Moose. Comparing vital signs in Caribou with other taxa was not always straightforward because the level of detail for other taxa often did not match the level of detail for our data. During the first month of life, heart rates of Caribou in our study were variable and on the lower end of the range reported for ≤35-d-old Finnish Reindeer (Soppela et al. 1986). Finnish Reindeer calves were exposed to much colder and warmer temperatures than our Caribou calves experienced, reflecting different environments in Finland and southcentral Alaska. Environmental differences may explain some differences between these studies, especially given that at the coldest temperatures, Reindeer calves increased metabolic rate (and hence heart and respiratory rates) at least 5-fold to maintain body temperatures (Soppela et al. 1986). As in Moose, heart rates of Caribou calves decreased with increasing age (Thompson et al. 2023). Compared to domestic sheep and goats (Capra hircus), heart rates of Caribou calves generally were lower (California Veterinary Emergency Team 2024). Information on respiratory rates of young cervids was even more limited than information on heart rates. Respiratory rates of Moose increased with age (Thompson et al. 2023), contrasting with our findings for Caribou. Elevated respiratory rates of Moose likely reflect increased activity levels of Moose calves that were taken on walks prior to collection of respiratory rates (Thompson et al. 2023) as compared to respiratory rates of Caribou, which we collected when animals were inactive. As with heart rates, respiratory rates of Caribou calves were lower than in domestic sheep lambs and goat kids (California Veterinary Emergency Team 2024). Differences in vital rates between Caribou and domestic livestock may reflect true physiological differences or may be related to differences in activity level or tolerance of handling for collection of vital rate data. Body temperatures of Caribou calves were similar to those reported for Reindeer calves, Moose calves, Black-tailed Deer fawns, and domestic lambs and kids (McTaggart-Cowan and Wood 1955; Soppela et al. 1986; Thompson et al. 2023; California Veterinary Emergency Team 2024). Mean (rectal) body temperature of ≤1-mo-old Caribou calves in our study was slightly lower than in similarly aged Reindeer calves (Soppela et al. 1986). Differences in body temperature of Reindeer and Caribou calves could reflect environmental differences (described above) or differences in thermogenesis due to variation in the amount of brown adipose tissue or food intake (Blix et al. 1984; White et al. 1984; Beale et al. 2018). As in our Caribou calves, body temperature of Moose calves and Deer fawns decreased as animals matured (McTaggart-Cowan and Wood 1955; Thompson et al. 2023). By 23 wk of age (representing the end of our study period), body temperatures of Caribou calves were similar to those reported for 12- to 18-mo Caribou (Hammel et al. 1972; Yousef and Luick 1975; Cuyler and Øritsland 1986), which may reflect a level of physiological maturation or seasonally reduced metabolic rates (McEwan and Whitehead 1970). Although our results suggest similarities in vital signs across taxa, they also highlight potential interspecific differences that are relevant to health assessments. Caribou calves exhibited compensatory physiological responses in heart and respiratory rates to increasing body temperatures. Increased hearts rates likely aided in heat dissipation in at least 1 of 2 ways. First, increased heart rates are needed to support increased respiration for heat dissipation via panting, and second, increased heart rates also likely reflect vasodilation of the circulatory system that aids in heat dissipation (Sokolov and Grechkina 1998; Davies and Maconochie 2009). Alternatively, as blood vessels dilate, but blood volume does not change, the body may sense hypovolemia (low blood volume) and respond by increasing heart rate (Davies and Maconochie 2009). Additional work would be required to distinguish whether heart rate increased in Caribou because of apparent hypovolemia, to support increased respiratory rates, or both. Increasing respiratory rates likely allowed Caribou to dissipate heat via evaporative cooling as they exhaled (Parker and Robbins 1984), but the relationships between respiratory rates and body temperature were more variable than other relationships in this study, suggesting other factors we did not model influenced respiratory rates. Environmental variables such air temperature, humidity, wind, and solar radiation can influence physiological responses of animals, especially thermoregulatory responses (Thompson et al. 2020, 2023). Variation in environmental conditions likely underpins some of the unexplained variation in vital signs, especially respiratory rates (Fig. 3) among Caribou in our study. Our goal was not to predict how environmental variables influenced body temperature and other physiological responses, so we did not collect information on environmental conditions concurrently with vital rates. Although environmental variables may have contributed to variation in vital rates, body temperatures, heart rates, and respiratory rate continued to decline as calves aged, even as seasonal ambient temperatures peaked in July–August and then rapidly declined in September and October. Decreases in vital signs despite increasing ambient temperatures may reflect increased capacity for physiological compensation to changing environmental conditions as calves mature, which could have contributed additional variation to vital sign data. The ability to compensate for homeostatic perturbations is critical, but individuals may not be able to compensate fully for these perturbations. We did not have time-series data needed to evaluate whether compensation was partial or complete. Nonetheless, younger animals, with their large surface area relative to body mass, face a higher risk of partial compensation because they have less capacity to dissipate heat than larger animals (Broucek et al. 2009; Roland et al. 2016; Barnes et al. 2017). The ability of animals to dissipate heat (or not) may have biologically significant consequences, particularly in a warming climate. In endotherms, body temperature is one of the most highly defended traits (Beale et al. 2018), meaning heat dissipation is necessary, but it also is energetically expensive (Fancy and White 1985). Additionally, animals consume less energy when heat dissipation is limited (West 2003; Broucek et al. 2009; Roland et al. 2016; Barnes et al. 2017; Beale et al. 2018; Denryter 2026). Increased metabolic costs and heat-associated reductions in intake can result in poorer growth and smaller body size (Broucek et al. 2009; Roland et al. 2016; Barnes et al. 2017). Body size may be less critical for captive animals, but for free-ranging animals, smaller-bodied juveniles experience greater risk of mortality over winter (Cook et al. 2004). Thus, the ability of Caribou calves to dissipate heat may be important as Caribou contend with a warming climate, particularly as young animals are the most sensitive to thermal stress (Broucek et al. 2009; Roland et al. 2016; Barnes et al. 2017). For captive Caribou, heat dissipation can be enhanced through artificial improvements to habitats (e.g., additional shade, fans with and without misters, and cool substrates) that support evaporative cooling. In contrast, free-ranging animals may need to alter habitat selection or shift distributions to ameliorate thermal stress (Verzuh et al. 2021; Denryter and Fischer 2022) and if that is not adequate, they could be forced to sacrifice food intake. Our data indicate that the threshold for fever in Caribou calves is lower than the 40.0 °C threshold used for domestic lambs as the upper 95% confidence interval for body temperature of Caribou calves never exceeded 39.2 °C (Table 3). More reasonable thresholds for fever (and hypothermia) would comprise values substantially deviating from 95% confidence intervals, which are generally used as benchmarks for “normal” body temperature (Geneva et al. 2019; Diamond et al. 2021). Body temperatures constituting a fever decrease over time, and hence body temperature (and other vital signs) should be evaluated relative to age-specific ranges. Aberrations in body temperature outside age-specific ranges require close monitoring as fever and hypothermia can develop quickly, especially in younger animals, and often require rapid intervention. Often, febrile calves refused to eat until the fever was reduced, which meant delays in identifying and treating fever carried some nutritional cost. Because flunixin meglumine was very effective at reducing body temperature and improving appetite (i.e., when given intramuscularly reduced body temperature ≥1°C within 15–20 min), earlier identification of fever may allow for earlier intervention with medication and a lower overall nutritional cost to febrile Caribou calves. Although more conservative, passive cooling techniques (applying cool damp cloths, providing cooling mats) may be advised to treat fever in some situations, but we observed that they were largely ineffective at reducing fever. Our dataset of Caribou vital signs provides information that can improve health assessments but has some important limitations. Initially, we expected age and body mass to be equally valuable at explaining changes in vital signs, but our results suggest otherwise and there are several possible explanations for this discrepancy. First, body mass data likely were less accurate than calf age data. For example, the scale we used to weigh calves was sensitive to ambient environmental conditions and did not perform as well when temperatures dropped below freezing. When this happened, body mass sometimes fluctuated by ≥2.5 kg from day to day, which was not biologically feasible based on intake and activity levels, so we noted when the body mass measurements may not be reliable. Issues with accuracy of body mass, however, only arose in the last few weeks of the study as temperatures dropped. Although body mass was less accurate on some days, the weekly patterns of body mass change were generally accurate, following expected patterns for rates of gain. Small errors in collection of body mass data may explain why age, which did not have the same noise associated with its measurements, was a better predictor of vital signs than body mass. Additionally, many individual staff collected vital signs of Caribou calves, which could have affected consistency in measurements; however, staff were trained using the same protocols and senior staff conducted quality control by checking new staff and providing additional training and by checking logs each day for potential errors and resolving them. Our estimate of standard error for all vital signs was less than the measurement accuracy (i.e., <0.1 °C, <1 heartbeat/min, <1 breath/min), which likely occurred because of our large number of observations contributing to a large denominator in the standard error calculation. Finally, although our sample size was small in terms of unique Caribou, it encompasses thousands of observations of heart rate, respiratory rate, and body temperature of young Caribou. Vital signs of Caribou gathered in our study can be used as age-specific baselines (95% confidence intervals, Tables 1–3) for monitoring health of Rangifer calves. In the absence of information on normal vital signs, animal caregivers must rely on changes in behavior that can be subtle and easily missed, particularly by inexperienced personnel. Thus, information gathered through our work provides objective means by which caregivers can assess changes in animal health before behavioral changes are recognized. Early recognition of perturbations in heart rate, respiratory rate, or body temperature can allow for earlier intervention and improved husbandry. In addition to their importance in monitoring health of Reindeer and Caribou calves, information on normal vital signs also provides a baseline for future studies of physiology, metabolism, and homeostasis including responses to perturbations (McEwan et al. 1965). Physiological ecology may be an important area of future Caribou research, particularly given that northern environments occupied by Caribou are experiencing warming at a faster rate than other parts of the planet (Rantanen et al. 2022). Supplementary Material