Pregnancy reduces critical thermal maximum, but not voluntary thermal maximum, in a viviparous skink

Upper thermal limits are commonly measured in ectotherms; however, the effects of life-history stages, and in particular pregnancy in viviparous species, are rarely considered. In this study, we examined whether two measures of upper thermal tolerance varied among life-history stages in a viviparous New Zealand skink (Oligosoma maccanni). First, we measured voluntary thermal maxima (VT_max) and critical thermal maxima (CT_max) for four groups: late-pregnant females, newly postpartum females, males, and neonates. Second, we examined whether exposing skinks to their CT_max in late pregnancy affected success of pregnancies or size and performance of offspring. We found that both VT_max and CT_max differed among the four groups, although only CT_max differed enough to detect specific pairwise differences. Pregnant skinks and neonates had a significantly lower CT_max than postpartum skinks, and pregnant skinks also had a lower CT_max than males. Effect sizes were very large between groups, where CT_max differed significantly, and borderline large for VT_max between male and neonate skinks and between postpartum and pregnant females. Pregnancy success, and the size and sprint speed of resulting neonates were not affected by thermal-tolerance tests on late-pregnant females. The reduction in CT_max we observed in pregnant skinks could indicate that at high temperatures, pregnant skinks do not have the same ability to keep up with oxygen demands as non-pregnant skinks-possibly reflecting reduced ventilation capacity simultaneous with high oxygen demands from embryos as well as maternal tissues. These findings are consistent with some studies, showing that reduced oxygen availability can reduce thermal tolerance in reptiles.