Supercooling cells of frost hardy palm leaves: Quantified percentage of frozen water and displacement from thermodynamic equilibrium

Trachycarpus fortunei is the most frost-hardy palm. Its leaves undergo supercooling, previously believed to involve a sublethal high temperature exotherm (HTE), followed by a period of no cellular water loss without freezing, which is terminated by a low temperature exotherm (LTE). However, this knowledge is based on laboratory studies only, and how leaves freeze in nature remains unknown. Also, experimental evidence is missing whether the LTE is caused by intracellular ice formation. We hypothesized that intracellular freezing is the cause of frost damage, and that it occurs after a certain amount of displacement from ideal equilibrium freezing. Observations on how T. fortunei plants freeze in the field, allowed appropriate laboratory simulations. The percentage of frozen water (pfW) as a function of temperature and the displacement from ideal equilibrium freezing were estimated based on high-resolution differential scanning calorimetry. Cryo-microscopic examinations allowed the localization of ice during HTE and LTE. In the field, leaves froze at −3.3±1.0 °C, similar to non-supercooling species. Using Snomax®, appropriate control of HTE was achieved in laboratory on detached leaves. After the HTE, ice was localized exclusively in the vascular bundles. Thereafter, only a small percentage of water (6–14 %) was frozen. Interestingly, we found further, previously unknown ice formation processes between the HTE and LTE, indicating moderate freeze dehydration in addition to supercooling. Freeze dehydration was time dependent and increased under slow cooling (higher frost-dose). Regardless of season, cooling rate, or displacement from ideal equilibrium, the LTE occurred at −15.6 °C, matching with the killing temperature. Post-LTE, intracellular ice formation in mesophyll cells was identified as the cause of frost damage. Slower cooling reduced the displacement from ideal equilibrium, but did not change LTE, suggesting an inactivation or activation of molecular components involved in triggering intracellular ice formation.