Frozen mountain pine needles: The endodermis discriminates between the ice-containing central tissue and the ice-free fully functional mesophyll.

Conifer (Pinaceae) needles are the most frost-hardy leaves. During needle freezing, the exceptional leaf anatomy, where an endodermis separates the mesophyll from the vascular tissue, could have consequences for ice management and photosynthesis. The eco-physiological importance of needle freezing behaviour was evaluated based on the measured natural freezing strain at the alpine treeline. Ice localisation and cellular responses to ice were investigated in mountain pine needles by cryo-microscopic techniques. Their consequences for photosynthetic activity were assessed by gas exchange measurements. The freezing response was related to the microchemistry of cell walls investigated by Raman microscopy. In frozen needles, ice was confined to the central vascular cylinder bordered by the endodermis. The endodermal cell walls were lignified. In the ice-free mesophyll, cells showed no freeze-dehydration and were found photosynthetically active. Mesophyll cells had lignified tangential cell walls, which adds rigidity. Ice barriers in mountain pine needles seem to be realised by a specific lignification patterning of cell walls. This, additionally, impedes freeze-dehydration of mesophyll cells and enables gas exchange of frozen needles. At the treeline, where freezing is a dominant environmental factor, the elaborate needle freezing pattern appears of ecological importance.

Freeze dehydration vs. supercooling of mesophyll cells: Impact of cell wall, cellular and tissue traits on the extent of water displacement.

The extent of freeze dehydration of mesophyll cells in response to extracellular ice varies from supercooling to severe freezing cytorrhysis. The structural factors involved are poorly understood. In a comparison of mesophyll cells of 11 species, the factors "cell wall", "cellular" and "tissue" traits were investigated. The extent of freeze dehydration was quantified as reduction in the sectional area during controlled freezing in the presence of ice. The cell wall thickness, cell size, cell area and the relative area of intercellular spaces were determined. The modulus of elasticity was determined by psychrometry. To grasp the relationships between factors and with freeze dehydration, we applied a principal component analysis. The first two components explain 84% of the variance in the dataset. The first principal component correlated negatively with the extent of freeze dehydration and relative area of intercellular spaces, and positively with the squared cell wall thickness to cell size ratio, elasticity and cell wall thickness. The cell size parameters determined the second principal component. Supercooling appeared preferable in cells with a high squared cell wall thickness to cell size ratio and a low relative area of intercellular spaces. Such factors are hypothesised to affect the magnitude of negative turgor pressure being built up below the turgor loss point. Negative turgor pressure slows dehydration by reducing the water potential gradient to the extracellular ice. With high levels of freeze dehydration, sufficient intercellular spaces for extracellular ice accommodation are needed. The low relative area of intercellular spaces increases cell-to-cell contact area and could support tissue stability.