Freeze-thaw-induced embolism in Pinus contorta: centrifuge experiments validate the 'thaw-expansion hypothesis' but conflict with ultrasonic emission data.

*The 'thaw-expansion hypothesis' postulates that xylem embolism is caused by the formation of gas bubbles on freezing and their expansion on thawing. We evaluated the hypothesis using centrifuge experiments and ultrasonic emission monitoring in Pinus contorta. *Stem samples were exposed to freeze-thaw cycles at varying xylem pressure (P) in a centrifuge before the percentage loss of hydraulic conductivity (PLC) was measured. Ultrasonic acoustic emissions were registered on samples exposed to freeze-thaw cycles in a temperature chamber. *Freeze-thaw exposure of samples spun at -3 MPa induced a PLC of 32% (one frost cycle) and 50% (two cycles). An increase in P to -0.5 MPa during freezing had no PLC effect, whereas increased P during thaw lowered PLC to 7%. Ultrasonic acoustic emissions were observed during freezing and thawing at -3 MPa, but not in air-dried or water-saturated samples. A decrease in minimum temperature caused additional ultrasonic acoustic emissions, but had no effect on PLC. *The centrifuge experiments indicate that the 'thaw-expansion hypothesis' correctly describes the embolization process. Possible explanations for the increase in PLC on repeated frost cycles and for the ultrasonic acoustic emissions observed during freezing and with decreasing ice temperature are discussed.