Hydraulic architecture was studied in shrub species differing in rooting depth in a cold desert in Southern Argentina. All species exhibited strong hydraulic segmentation between leaves, stems and roots with leaves being the most vulnerable part of the hydraulic pathway. Two types of safety margins describing the degree of conservation of the hydraulic integrity were used: the difference between minimum stem or leaf water potential (Y) and the Y at which stem or leaf hydraulic function was reduced by 50% (Y -Y50), and the difference between leaf and stem Y50. Leaf Y50 -stem Y50 increased with decreasing rooting depth. Large diurnal decreases in rootspecific hydraulic conductivity suggested high root vulnerability to embolism across all species. Although stem Y50 became more negative with decreasing species-specific Ysoil and minimum stem Y, leaf Y50 was independent of Y and minimum leaf Y. Species with embolism-resistant stems also had higher maximum stem hydraulic conductivity. Safety margins for stems were >2.1 MPa, whereas those for leaves were negative or only slightly positive. Leaves acted as safety valves to protect the integrity of the upstream hydraulic pathway, whereas embolism in lateral roots may help to decouple portions of the plant from the impact of drier soil layers.

Physiological adjustments to enhance tolerance or avoidance of summer drought and winter freezing were studied in shallow- to deep-rooted Patagonian cold desert shrubs. We measured leaf water potential (Ψ(L)), osmotic potential, tissue elasticity, stem hydraulic characteristics, and stomatal conductance (g (S)) across species throughout the year, and assessed tissue damage by subzero temperatures during winter. Species behavior was highly dependent on rooting depth. Substantial osmotic adjustment (up to 1.2 MPa) was observed in deep-rooted species exhibiting relatively small seasonal variations in Ψ(L) and with access to a more stable water source, but having a large difference between predawn and midday Ψ(L). On the other hand, shallow-rooted species exposed to large seasonal changes in Ψ(L) showed limited osmotic adjustment and incomplete stomatal closure, resulting in turgor loss during periods of drought. The bulk leaf tissue elastic modulus (ε) was lower in species with relatively shallow roots. Daily variation in g (S) was larger in shallow-rooted species (more than 50 % of its maximum) and was negatively associated with the difference between Ψ(L) at the turgor loss point and minimum Ψ(L) (safety margin for turgor maintenance). All species increased ε by about 10 MPa during winter. Species with rigid tissue walls exhibited low leaf tissue damage at -20 °C. Our results suggest that osmotic adjustment was the main water relationship adaptation to cope with drought during summer and spring, particularly in deep-rooted plants, and that adjustments in cell wall rigidity during the winter helped to enhance freezing tolerance.

Plants can avoid freezing damage by preventing extracellular ice formation below the equilibrium freezing temperature (supercooling). We used Olea europaea cultivars to assess which traits contribute to avoid ice nucleation at sub-zero temperatures. Seasonal leaf water relations, non-structural carbohydrates, nitrogen and tissue damage and ice nucleation temperatures in different plant parts were determined in five cultivars growing in the Patagonian cold desert. Ice seeding in roots occurred at higher temperatures than in stems and leaves. Leaves of cold acclimated cultivars supercooled down to -13 °C, substantially lower than the minimum air temperatures observed in the study site. During winter, leaf ice nucleation and leaf freezing damage (LT50 ) occurred at similar temperatures, typical of plant tissues that supercool. Higher leaf density and cell wall rigidity were observed during winter, consistent with a substantial acclimation to sub-zero temperatures. Larger supercooling capacity and lower LT50 were observed in cold-acclimated cultivars with higher osmotically active solute content, higher tissue elastic adjustments and lower apoplastic water. Irreversible leaf damage was only observed in laboratory experiments at very low temperatures, but not in the field. A comparative analysis of closely related plants avoids phylogenetic independence bias in a comparative study of adaptations to survive low temperatures.