OUP user menu

Pregnancy during hibernation in Japanese black bears: effects on body temperature and blood biochemical profiles

Michito Shimozuru, Ruriko Iibuchi, Takuro Yoshimoto, Akiko Nagashima, Jun Tanaka, Toshio Tsubota
DOI: http://dx.doi.org/10.1644/12-MAMM-A-246.1 618-627 First published online: 11 June 2013


Bears from the family Ursidae are the only terrestrial mammals that go through gestation, parturition, and lactation during hibernation. This is the 1st study to examine the influence of reproductive status (i.e., nonpregnant, pseudopregnant, or pregnant) on body temperature and blood biochemical profiles in hibernating black bears. Pregnant bears appeared to have higher and more stable body temperatures (37-38°C) than nonpregnant ones (34-36°C) during pregnancy, which was followed by a rapid drop to levels comparable to those of nonpregnant individuals after parturition. In midpregnancy (i.e., January), pregnant bears had higher blood glucose and lower triglyceride concentrations than did nonpregnant ones, whereas blood concentrations of free fatty acids, glycerol, and ketone bodies did not differ significantly. Plasma urea, creatinine, and urea/creatinine levels were significantly lower in pregnant bears than in nonpregnant ones. Pseudopregnant bears showed similar changes in body temperature and blood profiles to pregnant ones, but blood glucose levels in January were significantly lower than those in pregnant bears. These results suggest that pregnant bears maintain homeothermic conditions and supply nutrients to the fetus by stimulating thermogenesis, gluconeogenesis, and urea recycling during hibernation. In addition, these physiological changes appear to be mainly caused by corpus luteum-derived factors (e.g., progesterone) but also were facilitated by placental factors.

Key words
  • bears
  • body temperature
  • gestation
  • hibernation
  • metabolism
  • pregnancy
  • Ursus

Mammalian hibernation is an adaptive strategy to survive during a food-scarce season by reducing energy consumption. Hibernation is characterized by inactivity, lower body temperature, reductions in breathing and heart rate, and metabolic suppression. Hibernation has been documented in 8 different groups of mammals (Carey et al. 2003), suggesting that hibernation may not have evolved within specific species but may be an ancestral trait that is shared by many species belonging to different orders. Among mammalian hibernators, bears, including the Japanese black bear (Ursus thibetanus japonicus), have a unique hibernation physiology, surviving for up to 6 months without eating, drinking, urinating, or defecating (Nelson 1973). Small hibernators, such as hamsters, woodchucks, and ground squirrels, enter torpor with a low body temperature (0–10°C) and exhibit periodic arousal by rewarming to euthermic levels (35-37°C [Wollnik and Schmidt 1995; Zervanos et al. 2010; Healy et al. 2012]). In contrast, bears do not exhibit such drastic changes in body temperature (i.e., torpor-arousal cycles), although their body temperature shows multiday cycles with a range from 30°C to 36°C, and they remain conscious during hibernation (Harlow et al. 2004; Tøien et al. 2011). During hibernation, heart rate decreases from 60 beats per minute (bpm) to around 10 bpm, and the metabolic rate is suppressed to 25–50% of basal levels during the active period (Laske et al. 2010; Tøien et al. 2011). Severe muscle atrophy and bone loss do not occur in hibernating bears (Harlow et al. 2001; McGee-Lawrence et al. 2009), although they are nearly immobile throughout hibernation.

Another characteristic of bear hibernation is that female bears give birth to cubs and nurse them during hibernation (Nelson 1973). Bears from the family Ursidae are the only terrestrial mammals that go through all of their reproductive activities, including gestation, parturition, and the 1st several months of lactation, without consuming food and water (Ramsay and Dunbrack 1986). The mating season for black bears is from May to August (Yamamoto et al. 1998; Spady et al. 2007), but the fertilized egg undergoes quiescence at the blastocyst stage for 4–5 months, a phenomenon known as delayed implantation (Daniel 1974; Tsubota et al. 2001). Implantation occurs between late November and early December (Sato et al. 2000) when bears begin hibernation in the wild. True gestation lasts approximately 60 days, followed by parturition between late January and early February (Spady et al. 2007; Iibuchi et al. 2009). After parturition, mothers nourish cubs while remaining nearly immobile until their emergence from the den, usually in April (Koike and Hazumi 2008; Tøien et al. 2011). Black bears in wild habitats generally reproduce every other year (Kolenosky 1990), and hibernating females can be divided into reproductive and nonreproductive individuals. However, some females, especially under captive conditions, become pseudopregnant because of unsuccessful mating or chemosensory stimulation during the mating season (Okano et al. 2006). Pseudopregnant females show similar endocrinological changes during hibernation as pregnant females (e.g., elevated progesterone levels from November to early February [Sato et al. 2001; Schulz et al. 2003]).

The primary energy source for hibernating bears is stored body fat. In addition to their own needs, reproductive females also must provide energy and materials for fetal development and milk production. Consequently, their loss of mass is greater than that of nonreproductive females (Samson and Huot 1995; Harlow et al. 2002), and they lose a significant amount of muscle protein during hibernation (Tinker et al. 1998). These studies have suggested that catabolism of lipids and protein is enhanced in reproductive females relative to nonreproductive females. However, during the lactation period, no obvious differences in blood biochemical parameters, including glucose, lipid-related metabolites (e.g., free fatty acids [FFAs] and ketone bodies), and protein-related metabolites (e.g., total protein and urea), have been observed between lactating and nonlactating bears (Wright et al. 1999; LeBlanc et al. 2001). During lactation, bears nourish cubs with milk that is high in fat and low in sugar (Iibuchi et al. 2009), which allows them to suppress the consumption of glucose and glucogenic substances and to maintain similar blood profiles to those of nonlactating individuals. In contrast, during pregnancy, the primary energy source for fetal development is glucose, because the placental transfer of triglycerides (TGs), FFAs, and glycerol is very limited (Herrera 2002). In addition to producing glucose, bears must maintain their body temperature for fetal growth. Therefore, in order to develop a suitable body condition for fetal growth, pregnant bears are expected to show higher and more stable body temperatures, and higher blood glucose levels than nonpregnant females. However, to the best of our knowledge, no reports have addressed this issue. The purpose of the present study was to clarify how reproductive status affects thermoregulation and energy metabolism in hibernating bears by monitoring changes in body temperature and blood biochemical profiles during hibernation.

Materials and Methods

Animals.—A total of 17 adult female Japanese black bears between the ages of 5 and 15 were used in this study (n = 4, 3, 5, and 5 in 2008–2009, 2009–2010, 2010–2011, and 2011-2012, respectively). All females were kept at the Ani Mataginosato Bear Park in Akita Prefecture, northeastern Japan (39°55′N, 140°32′E) from 2008 to 2012. During the active period, from late April to late November, the bears were provided mainly with crushed corn (360 kcal/100 g, approximately 1.5 kg/individual; ration ingredients: 13.5% water, 8% crude protein, 3.8% crude fat, 71.7% nitrogen-free extract, 1.7% crude fiber, and 1.3% crude ash) once a day at 1600 h, and water was provided ad libitum. The bears also were provided with ration for bears and boars (KS; Oriental Yeast Co., Ltd., Tokyo, Japan; ration ingredients: 7% water, 21.4% crude protein, 9.6% crude fat, 51.6% nitrogen-free extract, 4.1% crude fiber, and 6.3% crude ash) and some fruit and vegetables as supplements intermittently. Two weeks prior to the onset of feed deprivation in late November or early December, the ration was reduced to one-third. At the end of November or the beginning of December, the bears were moved into indoor rooms for denning, where they were kept in isolation without feeding until the middle of the following April. The animals were provided with straw for nesting and with water ad libitum. The rooms were kept dark and not under temperature control. From the end of December to March, indoor temperatures were relatively constant (0-3°C), possibly because the facility was covered with snow. Feeding was resumed between 11 and 17 April. Because bears are anorectic and do not resume normal intake of food until 10–14 days after emergence from their dens (Nelson et al. 1983), the bears were fed one-third of the active period ration for 2 weeks after the onset of feeding. All experiments were conducted in accordance with guidelines of the American Society of Mammalogists (Sikes et al. 2011) and of the Animal Care and Use Committee of Hokkaido University (approval nos. JU8114, JU8134, and JU9124).

Blood collection and biochemistry analyses.—For bears that did not produce any cubs (i.e., pseudopregnant and nonpregnant females), blood was collected once every 4 periods: in late November or early December (November-December; just prior to the onset of feed deprivation; 27–28 November in 2008, 1–2 December in 2009, 22–24 November in 2010, and 23–24 November in 2011), early January (11 January in 2009, 8 January in 2010, 7–9 January in 2011, and 7–8 January in 2012), early March (3 March in 2009, 8 March in 2010, 2–3 March in 2011, and 1 March in 2012), and mid-April (1 week after the onset of feeding; 20 April in 2009, 23-24 April in 2010, 18–19 April in 2011, and 18–19 April in 2012). Pregnant females were sampled during the same period in November-December and January, but were not sampled after delivery (i.e., in March and April). Based on the determination of reproductive status, the numbers of pregnant, pseudopregnant, and nonpregnant females were 6, 6, and 5, respectively (detailed below). In November-December and April, each animal was isolated in an indoor room and provided with crushed corn (500 g/individual in November-December and April) at 1600 h. Water was provided ad libitum during isolation. After eating, a bear was fasted overnight (about 15–16 h) and an anesthetic was administered during the morning of the following day. In January and March, the bears were kept without feeding and were sampled in the morning. The bears were immobilized using blow darts with an intramuscular administration of 3.0 mg/kg zolazepam hydrochloride and tiletamine hydrochloride cocktail (Zoletil; Virbac, Carros, France) and 40 μg/kg medetomidine hydrochloride (Domitor; Zenoaq, Fukushima, Japan) based on their estimated body mass. After immobilization, the bears were weighed and blood samples were collected from the jugular vein into vacuum tubes. Heparin was used as the anticoagulant for the determination of TGs, glycerol, and total ketone bodies (total concentration of β-hydroxybutyric acid and acetoacetic acid), blood urea nitrogen, creatinine, and total proteins. A mixture of sodium fluoride, heparin sodium, and ethylenediaminetetraacetic acid disodium salt was used as the anticoagulant for glucose determination. Blood also was collected in vacuum tubes without an anticoagulant for the collection of serum to determine FFAs. The collected blood samples were centrifuged at 1,880 × g for 10 min, and the separated serum and plasma samples were rapidly frozen on dry ice. The blood samples were packed with dry ice, transported to the laboratory, and stored at −80°C until assay. After all of the experiments were finished, 200 βg/kg of atipamezole hydrochloride (Antisedan; Zenoaq) was administered intramuscularly as a medetomidine hydrochloride antagonist to aid recovery.

For blood biochemical analysis, we chose the following 9 representative blood parameters that reflect energy metabolic state. Each parameter was assessed in a manner similar to our previous studies (Kamine et al. 2012; Shimozuru et al. 2012). Plasma concentrations of glucose, TGs, blood urea nitrogen, creatinine, and total proteins were measured using an automatic blood analyzer (DRI-CHEM 7000; Fujifilm Medical Co., Ltd., Tokyo, Japan). Plasma urea and the urea/creatinine ratio were calculated as blood urea nitrogen/0.466 and urea/creatinine, respectively. Plasma glycerol, total ketone bodies, and serum FFAs were assayed using commercial kits (Glycerol Assay Kit; Cayman Chemical Company, Ann Arbor, Michigan; Ketorex Kit; Sanwa Chemical Company, Nagoya, Japan; and NEFA C Test; Wako, Osaka, Japan, respectively) according to the manufacturers’ protocols.

Monitoring of body temperature.—Because it is less invasive and involves an easier methodology, subcutaneous temperature was monitored as an alternative to core body temperature. Ventral and neck subcutaneous temperature (Tvs and Tns, respectively) were monitored during hibernation for 8 of 17 animals (n = 2 in 2009–2010, 5 in 2010–2011, and 1 in 2011–2012) using data loggers (3 g, 17-mm diameter, 6-mm thick; iButton DS1922L; Dallas Maxim Integrated Products, San Jose, California). The loggers were synchronized and programmed (resolution ± 0.0625°C; temperature was measured every 60 min; maximum recording duration was 171 days) and coated with less-invasive paraffin wax (Sigma-Aldrich Co., St. Louis, Missouri) before implantation. After blood collection in November-December, we shaved a section of hair (3×3 cm) on the central abdomen and on the dorsal part of the neck and washed the skin thoroughly with antimicrobial cleaner (Isodine Wash; Meiji Seika Pharma Co., Ltd., Tokyo, Japan). The loggers were implanted subcutaneously into these regions. On the same day as the blood sampling in April, the loggers were removed via skin incisions. After implantation or removal was finished, meloxicam (Metacam; Boehringer Ingelheim, Ingelheim, Germany) was administered subcutaneously at 0.2 mg/kg as an analgesia. The monitored bears included 2 pregnant, 3 pseudopregnant, and 3 nonpregnant females (details below). Body temperature recordings started 3 days after implantation, and any data within 3 days of blood sampling were excluded from our analyses. In each year, the same model of data logger was placed in front of the denning room to record coincident ambient temperature (Ta).

Determination of reproductive status.—In January, the pregnancy status of each bear was determined by ultrasonography (SonoSite 180; SonoSite, Inc., Bothell, Washington). When a fetus was detected, the bear was categorized as pregnant. Following delivery, lactating bears were not sampled in March or April to avoid disturbing nurturing.

To discriminate between pseudo- and nonpregnant females, we assessed serum progesterone (P4) levels using competitive double-antibody enzyme immunoassays. A total of 200 μl of serum was mixed vigorously with 2 ml of diethyl ether, and the ether layer was recovered by decanting to another tube after snap-freezing at −80°C. After evaporation, the residue was redissolved in 200 μl of phosphate-buffered saline with bovine serum albumin, and stored at −20°C until assayed. The primary and secondary antiserums that were used for enzyme immunoassays were rabbit anti-progesterone-3(E)CMO-BSA (FKA 302-E; COSMO Bio, Tokyo, Japan) and goat anti-rabbit γ-globulin serum (Seikagaku Co., Tokyo, Japan), respectively. Progesterone-3-CMO-HRP (FKA301; COSMO Bio) was used for the competitive reaction. Twenty-five microliters per well of standard progesterone solution (Nacalai Tesque, Inc., Kyoto, Japan) and extracted sample were assayed in triplicate. The minimum detectable level of P4 was 4.3 pg per well (i.e., 172 pg/ml), and the intra- and interassay coefficients of variation were less than 10%. Several studies have shown that pregnant and pseudopregnancy-induced black bears exhibit high P4 levels (>1.0 ng/ml) during the period from November to the end of January, whereas P4 concentrations of nonpregnant, that is, not pseudopregnancy-induced, bears remain lower (<1.0 ng/ml) during this period (Hellgren et al. 1991; Sato et al. 2001; Schulz et al. 2003). Therefore, in this study, bears that did not have fetuses or cubs and showed high P4 levels (>1.0 ng/ml) in January were categorized as pseudo-pregnant; the remaining bears (i.e., no fetus or cub and low P4 level) were categorized as nonpregnant.

Data analysis and statistical procedures.—The statistical analyses were performed using SPSS version 20 (SPSS Inc. 2011). The statistical significance of time-course changes in body mass and blood biochemical values was assessed separately by reproductive status. In non- and pseudopregnant bears, values in January, March, and April were compared to values in November-December using a paired Mest with Bonferroni's correction (i.e., comparisons were considered significant at P < 0.05/3 = 0.0167). Similarly, in pregnant bears, values in January were compared to values in November-December through paired t-tests, and the same significance level was used. The statistical significance of any differences in body mass or blood biochemical values among the reproductive status groups was assessed using Student's t-tests with Bonferroni's correction in November-December and January (i.e., pregnant versus pseudopregnant versus nonpregnant females) and in March and April (i.e., pseudopregnant versus nonpregnant females). Comparisons were considered significant at P < 0.05/3 = 0.0167. All values are presented as mean ± SD.

For body temperature analysis, Tvs and Tns data for each bear were averaged per day. Because of the limited number of monitored bears, statistical comparisons among reproductive status groups were not performed.


Reproductive status.—Changes in serum P4 levels and body mass are presented in Fig. 1. Six female bears were categorized as pregnant following the detection of a fetus in January. All of these bears gave birth to 1 or 2 cubs during the period between 1 and 6 February, whereas none of the remaining bears, including non- or pseudopregnant individuals, delivered cubs. Among the remaining bears, 6 and 5 bears were categorized as pseudo- and nonpregnant, respectively, on the basis of serum P4 levels. No significant differences in P4 levels were detected between pregnant and pseudopregnant females in November-December or January (P > 0.05). In pregnant and pseudopregnant females, serum P4 concentrations ranged from 1.2 to 4.5 ng/ml in November-December and from 2.7 to 5.6 ng/ml in January, whereas P4 concentrations ranged from 0.2 to 0.6 ng/ml in November-December and from 0.2 to 0.3 ng/ml in January in nonpregnant females. In March and April, P4 concentrations were lower than 1 ng/ml in all pseudo- and nonpregnant females. Taken together, the numbers of pregnant, pseudopregnant, and nonpregnant females were 6 (n = 2 in 2008–2009, 1 in 2009–2010, and 3 in 2011–2012), 6 (n = 1 in 2008–2009, 1 in 2009–2010, 2 in 2010–2011, and 2 in 2011-2012), and 5 (n = 1 in 2008–2009, 1 in 2009–2010, and 3 in 2010–2011), respectively. No significant differences in body mass were found among the 3 groups in November-December or in January or between pseudo- and nonpregnant females in March or April (P > 0.05). Also, daily loss of mass from November-December to January, which was calculated as BM[November_December] − BM[January]/interval date, where BM is body mass, did not differ significantly among pregnant (255 ± 38 g/day), pseudopregnant (230 ± 90 g/day), and nonpregnant females (193 ± 47 g/day).

Fig. 1

Changes in A) serum progesterone concentrations and B) body mass in nonpregnant (NP, n = 5), pseudopregnant (PP, n = 6), and pregnant (P, n = 6) bears. Error bars are SDs.

Body temperature.—Patterns in daily mean Tvs, Tns, and Ta are presented in Fig. 2 and Supporting Information S1 (DOI: 10.1644/12-MAMM-A-246.1.S1). In each year, Ta dropped at the beginning of December and was maintained within a stable range from 0°C to 3°C until the end of March, when it gradually increased. In nonpregnant females (bears 163, 160, and 59), Tvs was maintained within the range of 34–36°C throughout the hibernation period. Tns decreased at the beginning of December, in parallel with the decrease in Ta, and was maintained at a temperature 4—6°C below Tvs until March, when it gradually increased. In pregnant (bears 174 and 181) and pseudopregnant (bears 102, 98, and 14) females, Tvs was maintained within the range of 37–38°C during the period from December to January. In pregnant bears, the existence of cubs was initially confirmed by daily checkups on 1 February in 2010 (bear 174) and on 1 February in 2012 (bear 181). Until 3 days before delivery, daily mean Tvs remained with the range of 36.7-37.1°C in both bears; thereafter, Tvs began to decrease to near 36.0°C until the day of delivery. Until 3–6 days after delivery, Tvs of both bears further decreased to below 35°C. Similarly, in pseudopregnant females, a rapid decline in Tvs was observed from late January to mid-February. Tns patterns in pregnant and pseudopregnant females varied among individuals. In some cases (i.e., pregnant bear 181 and pseudopregnant bear 102), Tns was maintained at a higher temperature (32-35°C) from December to January and then decreased rapidly, similar to the pattern in Tvs. However, in other cases (i.e., pregnant bear 174 and pseudopregnant bears 98 and 14), Tns decreased at the beginning of December, similar to the pattern in nonpregnant females, and notable changes in Tns were not observed.

Fig. 2

Representative patterns in body temperature during hibernation in nonpregnant, pseudopregnant, and pregnant bears. Tvs: ventral subcutaneous temperature; Tns: neck subcutaneous temperature; Ta: ambient temperature. The pregnant bear (bear 181) gave birth to 2 cubs on 1 February (dotted line). Data for all monitored bears, including 3 nonpregnant, 3 pseudopregnant, and 2 pregnant females, are presented in Supporting Information S1.

Fig. 3

Changes in blood concentrations of A) glucose, B) triglycerides, C) free fatty acids, D) glycerol, E) total ketone bodies, F) urea, G) creatinine, H) the urea/creatinine ratio, and I) total protein during hibernation in nonpregnant (NP, n = 5), pseudopregnant (PP, n = 6), and pregnant (P, n = 6) bears. Different letters indicate significant differences among the groups for each month (P < 0.0167). Error bars are SDs.

Time-course changes in blood biochemical profiles during the hibernation period.—Changes in blood biochemical values and statistical data for the time-course analysis are presented in Fig. 3 and Supporting Information S2 (DOI: 10.1644/12-MAMM-A-246.1.S2), respectively. Glucose levels did not show significant changes in any of the groups (i.e., nonpregnant, pseudopregnant, or pregnant bears). Plasma concentrations of TGs and total ketone bodies increased significantly in January and March, compared to November-December, in all groups. Similarly, compared to November-December, FFA and glycerol levels increased in January in nonpregnant and pregnant bears, and in March in non- and pseudopregnant females, respectively. Plasma urea decreased significantly in January in pseudopregnant and pregnant bears and in March in pseudopregnant animals, whereas plasma creatinine increased in January in nonpregnant and pseudopregnant bears and in March in pseudopregnant females. Urea/creatinine ratios decreased significantly in January in all groups and in March in non- and pseudopregnant females. Plasma total protein concentrations increased significantly in January in pregnant bears.

The effect of reproductive status on blood biochemical values.—In November-December, no significant differences were found among the 3 groups in any of the parameters. In contrast, in January, significant differences were found among the groups in plasma glucose, TGs, urea and creatinine concentrations, and urea/creatinine ratios, although serum concentrations of FFAs and plasma concentrations of glycerol, total ketone bodies, and total proteins did not differ significantly among the groups. Plasma glucose concentrations were significantly higher in pregnant bears than in pseudopregnant (t10 = 3.459, P < 0.01) and nonpregnant (t9 = 6.071, P < 0.001) animals, and were higher in pseudopregnant bears than in nonpregnant ones (t9 = 3.140, P < 0.0167). Plasma TG concentrations were significantly lower in pregnant and pseudopregnant bears than in nonpregnant females (t9 = −3.710, P < 0.01, for pregnant versus nonpregnant; t9 = −3.806, P < 0.01, for pseudopregnant versus nonpregnant). Plasma urea concentrations were significantly lower in pregnant and pseudopregnant bears than in nonpregnant females (t9 = −4.483, P < 0.01, for pregnant versus nonpregnant; t9 = −3.513, P < 0.01, for pseudopregnant versus nonpregnant). Plasma creatinine concentrations and urea/creatinine ratios were significantly lower in pregnant bears than in nonpregnant females (t9 = −4.473, P < 0.01, for creatinine; t9 = −2.993, P < 0.0167, for urea/creatinine ratio). In March and April, no significant differences were found between pseudo- and nonpregnant bears in any of the blood parameters (P > 0.05).


Body temperature.—In nonpregnant females, Tvs during hibernation ranged from 34°C to 36°C, which was comparable to core body temperatures in other bear species during hibernation (Harlow et al. 2004; Tøien et al. 2011). Subcutaneous temperature is lower than intraperitoneal core temperature in other mammals, such as rhesus macaques (Macaca mulattaTaffe 2011), and this difference was expected to be larger in a cold environment, but the effect of Ta on Tvs would be minimized in immobile animals that remain in a curled posture during hibernation. Pregnant bears appeared to have higher and more stable Tvs than did nonpregnant females during pregnancy, which was followed by a rapid drop to a level that was comparable to that of nonpregnant bears during lactation. Similar shifts during reproductive events have been reported in European brown bears (Ursus arctos arctosHissa 1997) and American black bears (Ursus americanusTøien et al. 2011). This suggests that more energy is required for hibernating pregnant bears to maintain a high and stable body temperature for fetus development. Comparable body temperature patterns in pseudopregnant bears indicated that high and stable body temperatures were attributable to endocrinological changes during hibernation that were shared between pregnant and pseudopregnant bears. If an endocrinological factor exists to control thermogenesis based on bear reproductive status during hibernation, likely candidates would be pituitary or ovarian hormone. In addition, the concentration of such a hormone is expected to be higher in pregnant and pseudopregnant bears from November to January, as compared to nonpregnant bears, whereas the level should be low regardless of reproductive status in February-March. Among candidate hormones that have been investigated in this and previous studies (i.e., luteinizing hormone, follicle-stimulation hormone, prolactin, P4, estradiol-17 β, and inhibin—Hellgren et al. 1991; Sato et al. 2000, 2001), only P4 meets the above conditions. P4 has thermogenic effects in other species (Nakayama et al. 1975; Marrone et al. 1976), which supports its involvement in the maintenance of high body temperatures in pregnant bears. Previous studies (Hellgren et al. 1991; Sato et al. 2001) have demonstrated that in pregnant and pseudopregnant bears, serum P4 levels exhibit a sharp and transient rise (5-20 ng/ml) during the preimplantation period in December, begin to decrease in January, and then suddenly drop prior to parturition. These changes would induce an elevation in Tvs from the end of November to the beginning of December in pregnant and pseudopregnant bears, and the sharp drop in Tvs that occurred just before parturition.

Compared to Tvs, Tns was approximately 4–6°C lower and was more variable during hibernation, which is due to the thermal gradient in the body and the effect of Ta. In parallel with the decrease in Ta in early December, Tns decreased without relation to Tvs in most bears, although patterns in Tns seemed to correspond to patterns in Tvs after February. Furthermore, unlike Tvs, Tns did not always reflect individual reproductive status in November-December to January; only 2 of 5 pregnant and pseudopregnant bears (i.e., bear 102 in 2009–2010 and bear 181 in 2011–2012) maintained higher Tns levels than nonpregnant bears. This would be partially due to the location of the temperature logger on the back of the bear's neck. The dorsal neck area, where the loggers were implanted, would be exposed to ambient air but in some bears it also could have been in contact with the wall or the straw nest material. This could have affected the Tns measurements, although bear postures were not monitored during hibernation. It can be concluded that, compared with Tvs, Tns is less reliable for monitoring the reproductive status of bears and for estimating the delivery date.

Changes in blood profiles: time-course and pregnancy effects.—On the whole, time-course changes in blood biochemical parameters during hibernation were comparable to results reported in earlier studies (Ahlquist et al. 1984; Lohuis et al. 2005; Shimozuru et al. 2012), although the patterns seemed to differ slightly depending on the reproductive group. The increases in TGs, FFAs, glycerol, and ketone bodies clearly indicate an energy shift from glucose to lipids. Furthermore, the absence of hypoglycemia during hibernation suggests an enhancement of gluconeogenesis, which is mainly achieved through the use of glycerol and not of amino acids (Fedorov et al. 2009; Shimozuru et al. 2012). The increases in creatinine and total proteins and the decreases in urea and the urea/creatinine ratio, which are characteristic of the physiological state in hibernating bears (Nelson et al. 1984; Hellgren 1995), reflect muscle protein conservation, reduced gluconeogenesis from proteins and amino acids, and the lack of urination. In addition, in April, most of the blood biochemical parameters had returned to levels that were comparable to those observed in November-December, suggesting that normalization of the energy metabolism occurs relatively rapidly with the commencement of feeding.

Although pregnancy-related hormonal changes (e.g., elevated progesterone levels) were already observed in late November (Sato et al. 2001), differences in blood biochemical profiles among the groups were not observed until January. This suggests that differences in the blood biochemical profiles that were observed in January were not simply due to pregnancy-related hormonal changes but due to the interaction between those changes and fasting. Plasma glucose levels are highly variable among individuals and among studies of hibernating bears (Erickson and Youatt 1961; Palumbo et al. 1983; Lohuis et al. 2005). Hissa et al. (1994) hypothesized that this is partially due to differences in reproductive status among individuals, which is supported for the 1st time in the current study. The maintenance of high blood glucose levels in pregnant bears would facilitate fetal development and would be achieved by enhancing gluconeogenesis from glycerol or reducing glucose use in peripheral tissues (e.g., muscles), or both. Similar glucoregulatory mechanisms during pregnancy have been reported in ruminants (Bell and Bauman 1997). In the present study, pregnant bears had higher glucose levels than did pseudopregnant animals, which suggests that more than just pregnancy-related hormones that are secreted by the ovary (e.g., progesterone), but also placenta-derived hormones or cytokines, or both, contribute to the maintenance of high glucose levels. In support of this conclusion, it was reported that progesterone and placental extracts stimulate gluconeogenesis in rats (Dahm et al. 1977) and sheep (Thordarson et al. 1987), respectively. Thordarson et al. (1987) suggested that, among placental factors, placental lactogens are likely to play a role in the maintenance of high glucose levels during gestation in sheep; however, the existence of placental lactogens has been documented only in primates, rodents, and ruminants (Gootwine 2004), but not in carnivores. Therefore, this issue requires further study to determine which factors and associated mechanisms are involved in glucoregulation during pregnancy in bears.

During a fasting state, the majority of TGs in the circulatory system are transported in very low-density lipoprotein derived from the liver. Therefore, the lower levels of TGs in pregnant and pseudopregnant bears, as compared to nonpregnant bears, in January may be attributable to reduced activity of the resynthesis of TGs in the liver. In the liver, exogenous glycerol is converted into glycerol-3-phosphate, which will be used for gluconeogenesis (Ahlquist et al. 1984), or combined with fatty acids to synthesize TGs (i.e., re-esterification of fatty acids). Because of the preferential use of glycerol-3-phosphate for gluconeogenesis in pregnant and pseudopregnant bears, reesterification would be reduced, consequently decreasing circulating TG levels relative to those in nonpregnant bears. However, this process does not provide a sufficient explanation for the fact that pregnancy and pseudopregnancy did not affect levels of circulating glycerol, FFAs, or ketone bodies. If glycerol utilization for hepatic gluconeogenesis is activated by pregnancy, lipid mobilization from adipose tissue should have been enhanced, which would thereby increase circulating lipolytic products (glycerol and FFAs) and ketone bodies. Increased lipolytic products and ketone bodies may have been siphoned off across the placenta for fetal growth during pregnancy, although their placental transfer, except for that of ketone bodies, is quantitatively low (Herrera 2002). Notably, single time-point measurements of blood parameters only provide a “snapshot” of metabolism. Further studies are necessary to examine differences between pregnant and nonpregnant bears in the blood kinetics of these parameters and of hormonal control in energy metabolism.

Although pregnancy did not significantly affect daily losses of body mass (estimated from the difference in body mass from November-December to January) in this study, pregnant and lactating bears have been reported to undergo severe losses of body mass (Samson and Huot 1995) and muscle protein (Tinker et al. 1998; Harlow et al. 2002), which suggests that protein catabolism was enhanced in those bears. Consequently, one could expect a corresponding increase in nitrogenous end- products, including urea and creatinine, and urea/creatinine ratios in pregnant bears compared to nonpregnant individuals. However, this study demonstrated that the opposite was true. This paradoxical result can be explained by the fact that the enhancement of protein turnover avoids urea production in hibernating bears (Hellgren 1995). The rate of protein turnover increases 3- to 5-fold during hibernation (Lundberg et al. 1976), and the carbon from labeled alanine that was injected during hibernation entered only plasma proteins, not urea (Ahlquist et al. 1984). Furthermore, Barboza et al. (1997) showed that almost all of the urea that was produced was reused in hibernating bears. These studies suggest that enhancements in protein turnover and urea recycling reduce blood urea levels in hibernating bears. Compared to nonpregnant bears, pregnant animals need to synthesize more amino acids and proteins to maintain the pregnancy and fetal growth. Thus, it is conceivable that these physiological demands accelerate protein turnover and urea recycling, thereby reducing blood urea and creatinine levels and the urea/creatinine ratio. Testosterone reduces blood urea concentrations in male bears during hibernation (Nelson et al. 1978). Similarly, it is possible that some anabolic hormones, such as prolactin and progesterone, suppress blood urea levels in females; this is supported by the fact that a reduction in blood urea also was observed in pseudopregnant bears. However, our interpretation of the current results takes into account that water was not restricted during this study. Although drinking behavior and urination were not directly observed during daily checkups, if pregnant and pseudopregnant females consumed more water, as compared to nonpregnant bears, this action might have compensated for their body mass loss and may have affected blood levels of urea and creatinine. Therefore, the current findings must be confirmed by a reproductive experiment under water-deprived conditions.

In conclusion, we have demonstrated for the first time that reproductive status affects the body temperature and blood biochemical profiles of hibernating black bears. Although pregnancy during hibernation requires additional energy to maintain a high body temperature and high blood glucose levels for fetal growth, the effect on loss of mass was limited, at least during the first one-half of the gestational period, which indicates that the bears had increased the efficiency of thermogenesis and energy metabolism. More amazing is that bears were better at avoiding accumulation of nitrogenous end- products, which is a key feature of bear hibernation (Nelson et al. 1984), during pregnancy. It is suggested that these physiological changes during gestation were mainly caused by corpus luteum-derived factors, but they also were facilitated by some placental factors. However, it is still not clear which endocrinological factors are involved in regulation, or how they act. Further study is needed to address these issues.

Supporting Information

Supporting Information S1.—Patterns in body temperature during hibernation in nonpregnant (n = 3), pseudopregnant (n = 3), and pregnant (n = 2) bears. Tvs: ventral subcutaneous temperature; Tns: neck subcutaneous temperature; Ta: ambient temperature. Both pregnant bears gave birth to 2 cubs on 1 February (dotted line).

Found at DOI: 10.1644/12-MAMM-A-246. 1.S1

Supporting Information S2.—Statistical data for time-course analyses of blood biochemical values.

Found at DOI: 10.1644/12-MAMM-A-246.1.S2


We thank K. Sakamoto for loaning us experimental devices. We also thank the staff at the Ani Mataginosato Bear Park for then-generous support. This study was supported by a Grant-in-Aid for Scientific Research from the Ministry of Education, Sports, Science, and Technology of Japan (21598355) and a Grant-in-Aid for Young Scientists (B; 23780277).


  • Associate Editor was Loren D. Hayes.

Literature Cited

View Abstract