Re-crystallisation temperature (T_rc) in an antarctic springtail

Cryobiologists have traditionally assumed that the temperature of crystallisation (T_c) or supercooling point (SCP) of a chill-tolerant insect is not a stochastic event, i.e. that it is a biologically meaningful indicator of phenotypic characteristics, be they exogenous influences (e.g. acclimation/acclimatization) or endogenous factors (e.g. life history stage, moult state). Recent work by Wilson et al. (11) has suggested that SCPs – at least in non-biological samples – are more stochastic than previously thought. Here, this question is tested indirectly by the repetitive freezing of individuals of the Antarctic springtail, Cryptopygus antarcticus. The springtails were each supercooled ten times in succession to determine their recrystallisation temperatures (T_rc). SCPs were found to be deterministic i.e. related to their initial T_c. Despite the mortality of re-crystallised samples, 70% showed <1°C difference between T_c and T_rc1 and 95% showed <5°C difference. T_c and T_rc1 were significantly correlated. Variability in re-crystallisation temperatures is hypothesised to be predominantly the result of differences in nucleator content and changes in body fluid osmolality during the experimental exposures. Factors affecting the relative variability of SCPs are discussed.