Physical aspects of cavitation and freezing in conifer xylem
Part I details work on cavitation, refilling and permeability of conifer xylem. Part II shows how conifers may avoid widespread cavitation during freezing of the xylem sap. PART I Cavitation was induced in sapwood samples by sealing wood in semi-permeable membranes and immersion in osmotic solutions. Emptying of up to 20-30% of the tracheids occurred for absolute pressures of between 1.0 x 105 N. m-2and -4 x 105 N m-2but very little further emptying occurred when the pressure was lowered to -9 x 105 N m-2. The limit of emptying found for this range of pressures indicates that one type of tracheid (earlywood or latewood) preferentially cavitates. It is suggested that the latewood cavitates more readily than the earlywood. Experimental tests showed that widespread refilling of cavitated tracheids only occurred when the xylem sap pressure was raised above the pressure in the cavitated tracheid lumens. Six theories to explain refilling of tracheids at the top of tall trees are considered. Of these capillary condensation, root pressure, refilling by ray parenchyma cells and capillarity are discarded on theoretical grounds. Experiments showed that temperature changes and temperature gradients did not cause refilling in equilibrated samples or in intact stem and root systems. Water absorption at the leaves appears to be the most likely method of refilling. Steady state liquid permeability measurements showed that xylem permeability decreased exponentially with increasing void volume for Douglas fir and Sitka spruce. The curves are probably not applicable to living trees because some liquid flowed through partially refilled tracheids. PART II Pressure rises up to 39 x 105 N m-2 and low freezing rates (1.75 to 2.3 pm s-1) were measured in freezing sapwood samples using implanted pressure transducers and thermistors. Electron micrographs of slowly frozen wood show bordered pit aspiration is not widespread during radial freezing. A finite element model of freezing in a single tracheid predicts pressure rises similar to those measured experimentally. This model also predicts that 5 to 8% of the water in the lumen migrates to other areas during freezing. However, in contradiction to experimental results, a model of whole stem freezing predicts freezing rates four times faster than the measured values. The low freezing rates measured experimentally indicate that bubbles only nucleate in the outermost layers of the xylem during freezing. However, if 5 to 8% of the lumen water migrates during freezing, low pressures on thawing may still cause cavitation. For this to be avoided it is suggested that water migrates from freezing tracheids to already frozen tracheids through narrow cell wall capillaries.