Homeothermic variation during gestation
During the survey of Torpor during Gestation, it became increasingly clear that it would be difficult to precisely determine at what stage of development the torpor was occurring, particularly for placentals. I decided to also collect studies of gestations that appear to be completely homeothermic to see if there were subtle variations which would give some information on developmental stage(s). This might some guidance in interpreting the torpor studies. It might also be useful to study reports of species that never enter torpor, but where pregnant females show different temperature behaviors at the same as non-pregnant females.
For example, one common behavior is what Kozak (1997) termed “gestational hypothermia,” defined as lower average daily temperatures. Roberts (2019) in her dissertation on temperature behaviors of pregnant vervet monkeys included a table of species in which this has been observed. This table also shows that for species in which daily variation in temperature was also measured, all showed small decreases in this variation compared to non-pregnant females, i.e. pregnant females were somewhat more homeothermic than non-pregnant females. One such species is the house mouse (Mus muculus). Since it is the most studied mammal under controlled laboratory conditions, it is likely to provide the most insight on gestational hypothermia
Rodentia
Gamo et al. (2013) monitored body temperature and activity in laboratory mice prior to mating and during pregnancy. Daily mean temperature and mean temperatures while active (nocturnal) or resting (diurnal) are shown in the figure at right. Prior to mating, temperatures when active averaged about 1.0oC higher than when resting. The typical gestation time for mice is around 20 days. The average day of conception likely occurred .
Figure 6 of Gamo et al. 2013
early in the period they have labeled as “mating.” During the period of Days 7-11 post-conception, corresponding to the gastrula through pharyngula stages, mean daily temperature changed very little and was about 0.5oC higher than prior to pregnancy, and the difference between active and rest mean temperatures were about half as much, 0.5oC, as prior to pregnancy, i.e., they were more homothermic. Thereafter the mean temperatures decreased about 0.1oC per day until shortly before parturition and differences between active and rest means were even less, about 0.3oC. During the latter part of pregnancy when the mean daily temperatures were decreasing, there was much less activity at night. Gestational hypothermia, average daily temperatures less than prior to mating, could be said to occur the last 6 days prior to parturition
In a study of the role of various genes in circadian oscillation of fat deposition and utilization in female mice, Wharfe et al. (2016) also reported that daily changes in body temperature had two components, a circadian oscillation (minimum around noon and maximum around midnight) of the temperatures averaged over a few hours and more rapid fluctuation, presumably related to periods of activity and rest. The daily mean of the circadian oscillation was 0.6oC higher at 6-10 days post conception compared to prior to mating and then decreased about 0.05oC per day until just prior to parturition. The amplitude of the circadian oscillation was 0.5oC prior to mating, 0.3oC on Day 6, decreasing thereafter to just 0.2oC. These results are similar to those of Gamo et al., except that was not gestational hypothermia late in pregnancy. Despite the echidna being much more heterothermic than mice in the earliest stages of gestation, the general pattern does not appear to be all that different from around gastrulation onward.
​
Kittrell and Satinoff (1988) reported that laboratory rats (Rattus norvegicus domestica) have considerably smaller temperature variations during gestation than prior to mating or during lactation. Eliason and Fewell (1997) observed inter-animal variability on Day 10 of gestation was less than half that prior to pregnancy. Mean temperature dropped about 0.7oC between Day 10 and Day 20, just prior to parturition.
Figures 4A and 4B of Wharfe et al. (2016)
Figure 1a of Eliason and Fewell (1997)
Cairns et al. (2005) found little temperature difference (<0.2oC) in untreated rats between prior to pregnancy and 4, 9, or 10 days after conception. Subsequently the temperature decreased to about 0.5oC just prior to birth, then increased about 1oC by Day 7 of lactation. Mac Aulay et al. (1993) observed that the primitive streak forms at Day 8.5 after conception. It thus appears that for at least until a day and a half after the start of gastrulation rats are homeothermic.
Figure 2 of Cairns et al. (2005)
Williams et al. (2011) studied free-ranging artic ground squirrels (Urocetellus parryii) in the foothills of the Brooks Range in northern Alaska. The figure below left shows the temperature profile for more than a year for a representative female and male. Females enter the hibernacula (A) and almost immediately enter hibernation (B) in early fall with brief normothermic periods until the final arousal (C) and leaving the hibernacula (D) in early spring early spring. Males enter the hibernacula (A) later than females, but then have a period of lower temperature variation before beginning hibernation(B). Males also arouse (C) earlier from hibernation but remain in the hibernacula for a period before leaving at around the same time as the female arousal. Based on an observed gestation time of around 25 days, mating happens shortly after the female’s arousal and emergence from the burrow. Parturition date can be determined from an abrupt 1-1.5oC rise in temperature, which was also observed in captive females housed at a constant temperature. Note that the daily fluctuation in temperature in temperature during gestation is considerably less during gestation than after parturition. In the figure below right the mean of the average daily body temperature for 36 females was plotted for days prior to the assumed parturition date. From Day -25 (likely shortly after conception) to Day -21 temperature increased about 0.3oC, changed little over the next six days, before slowly decreasing until parturition. If the timing of the developmental stages is similar to that of rats with a similar gestation length, then at least part of the G/P would be overlapped by the constant temperature period.
Williams et al. (2011) Figure 1A
Willaims et al. (2011) Fig 2B
Similar results were reported by Kuroyanagi et al. (2022) for both free-ranging and captive Japanese wood mice (Apodemus speciosus). The figure at right shows the temperature as a function of days before and after parturition, which was marked by an abrupt 1oC rise in daily maximum temperature indicated parturition. Mean copulation date was taken to be Day -21. The daily range of temperature was less during gestation than either before or after. Mean daily temperature rose for about 5 days after copulation, mainly due to a raising daily minimum temperature. For the next 5 days, the mean daily temperature remained near constant for about 5 days. During the last 10 days before parturition the mean daily temperature dropped steadily. Again at least part of G/P likely overlapped the constant temperature period.
Kuroyanagi et al. (2022). X-axis: Days prior to or after partruition
The previously discussed rodents are mostly active at night. A diurnally active species, the Nile grass rat (Arvicanthus niloticus) was studied by Schrader et al. (2009). The figure below shows the change in mean temperature throughout the night (black bar) and daytime (white bar) for reproductive females on three different days during gestation and for non-reproductive females at the corresponding times during the year. Daytime activity and temperature were higher in pregnant females throughout pregnancy compared to non-pregnant females. Nighttime temperatures did not decrease as much in pregnant females, particularly at the tenth day of pregnancy. Taking account both the daytime and night time temperatures, pregnant females were most homeothermic on Day 10.
Shrader et al. (2009) Part of Figure 4
Primates
Roberts (2019) Part of Figure 2, birth indicated by arrow
Roberts (2019) measured the internal temperatures of vervet monkeys (Chlorocebus pygerythrus) in the semi-arid Karoo region of South Africa. There was no difference between males and non-pregnant or lactating females. However, during pregnancies temperatures averaged about 0.40C cooler before abruptly rising at parturition. She illustrated for two females in the figure at right that this gestational hypothermia begins to occur around the time of conception, which she calculated based on an assumed gestational time of 165 days. However, this estimate came from a study by Rowell et al. (1970) of a captive colony in Kampala, Uganda, near the equator. In Uganda, births occurred throughout the year, while in the Karoo all the births were in the warmer months of October through January. In the later cases, conception would occur in the dry, colder months, when development might occur more slowly. In the figures at right, the x axis starts at 20 days before the expected start of pregnancy. If gestation time is longer in Karoo, the period of around 20 days when the temperature was higher and relatively stable could be occurring after conception.
Artiodactyla
Schmidt et al. (2020) measured the body temperature wild female musk ox (Ovibus muschatus) in northern Greenland. The gestation period overlaps the period of lowest ambient temperatures and deepest snow cover when food availability is at its lowest. Animals were captured in early October and reproductive status determined by vaginal ultrasound and serum samples. Temperature was monitored with intravaginal implants and activity monitored by collared-mounted GPS transmitters. In their Figure 2A shown at right, circles (with different colors for different groups) show the individual temperature measurements. Solid lines and associated shaded areas show a smooth fit to the data with 95% confidence intervals for different groups. Although the number of animals studied was small, the temperature behavior was clearly distinct based on reproductive status.
For non-pregnant animals (with calves) that survived the, winter(purple, n=2), the average daily body temperature steadily decreased to about 37.4C in March when ambient temperatures were the lowest, about -20C. Thereafter, body temperatures increased during the period when snow depth was decreasing. These animals were accompanied by an offspring at the time of implant. These observations support the idea that winter hypothermia allows winter survival after depletion of energy reserves from prior gestation and lactation.
Figure 2a of Schmidt et al.(2020)
In contrast, pregnant females that survived the winter (yellow, n=3) maintained a near-constant average daily temperature (about 38.2C) during gestation. Daily fluctuations in temperatures were less than for non-pregnant females. Body temperatures from February through parturition were also monitored by the same methods for pregnant oxen (green, n=3) in captivity in Alaska where ambient temperatures were higher and food was freely available. These animals had temperatures similar those of pregnant animals in the harsher conditions in Greenland.
Two of the Greenland females (brown, one initially pregnant, one not) had increasingly lower temperatures before dying.
Based on a gestation time of around 235 days, I estimated that the average gestation at the time the temperature monitoring began at around 40 days. Since the temperature difference between pregnant and non-pregnant animals was already about a third of the later maximum difference, it appears likely that the period of maintaining the more-constant body temperatures in the animals that gave birth must have begun early in gestation.
Katsumata et al. (2006) measured daily (before active period) rectal temperatures and serum progesterone levels (every 2-4 weeks) for two female killer whales (Orcinus orca). One whale was followed through three full-term (around 530 days) pregnancies with healthy deliveries, and the other for two pregnancies, in which the first resulted in a full-term live birth, but the baby died 30 minutes later, and the second resulted in a stillbirth at around 2/3 though gestation, and the mother died 3 days later.
Three successful pregnancies- Figure 3 of Katsumata et al. (2006)
Two unsuccessful pregnancies- Figure 2 of Katsumata et al. (2006)
The temperature data were averaged over two-week intervals. Early in successful pregnancies both temperature (0.4-0.6oC) and progesterone (around 2-fold) increased, with temperature reaching a maximum in the second or third interval and then falling, reaching near pre-pregnancy level shortly before birth. During the temperature ascending phase, there was a correlation between temperature and progesterone levels, but not in the descending phase.
For the female with the two unsuccessful pregnancies, in the first pregnancies after the initial rise in temperature, the temperature did not decrease for about 20 days. Temperature behavior appeared normal in the 2nd pregnancy until about halfway through gestation when the temperature became elevated before the stillbirth and maternal death.