Tansley Review No. 119

Vascular plants have shown a strong evolutionary trend towards increasing length in xylem conduits. Increasing conduit length affects water transport in two opposing ways, creating a compromise that should ultimately define an optimal conduit length. The most obvious effect of increased length is to decrease the sequential number of separate conduits needed to traverse the entire pathway, and thereby to reduce the number of wall‐crossings and the hydraulic resistance to flow within the xylem. This is an essential evolutionary pressure towards the development of the vessel, a conduit of multicellular origin whose length is not restricted by developmental constraints. The vessel has been an essential component in all plant lineages, achieving transport tissues with very high specific conductivity. A countering effect, however, arises from the partitioning of the cavitation response, a process whereby individual xylem conduits drain of water and lose conducting capacity. Flow in the xylem is down a gradient of negative pressure, which is necessarily most negative in the distal regions (i.e. near the foliage). Cavitation can be caused directly by negative pressures, and results in a total loss of the hydraulic conductance of the individual conduits within which it occurs. If cavitation is triggered by low pressure experienced only at the very distal end of a long conduit, the conduit nevertheless loses its conducting capacity along its entire length. Pathways composed of long conduits will therefore suffer greater total conductance loss for equivalent pressure gradients, because the effects of cavitation are not effectively restricted to the tissue regions within which the cavitation events are generated. By contrast, short conduits can restrict cavitation to distal regions, leaving trunk and root tissues less seriously affected. The increased total conductance loss of a system made entirely of very long conduits translates into a lower maximum rate of water transport in the xylem. The loss in hydraulic capacity associated with failure to partition the flow pathway fully, and locally contain the effects of cavitation, theoretically reaches a maximum of 50% for the extreme case in which a single set of conduits traverses the entire pathway. Shorter conduits confine individual cavitation events to smaller regions and permit the pathway as a whole to have a more gradual conductance loss in conjunction with the pressure gradient. A compromise exists between (1) minimizing total conductance loss from cavitation via fine partitioning of the pathway with many tiers of short conduits, and (2) reducing total wall resistance via coarse partitioning with a few tiers of long conduits. An analysis is presented of the optimal number of end walls (i.e. mean conduit length relative to total pathway length) to maximize transport capacity. The principle of optimal containment of cavitation also predicts that conduits should not be of equal length in all portions of the pathway. The frequency of end walls should rather be proportional to the magnitude of the water‐potential gradient at each point, and conduits should be longest in the basal portion (roots) and progressively shortened as they move up the stems to the foliage. These concepts have implications for our understanding of the contrasting xylem anatomies of roots and shoots, as well as the limits to evolution for increased hydraulic conductance per xylem cross‐sectional area. They also indicate that to model the hydraulic behaviour of plants accurately it is necessary to know the conduit length distribution in the water flux pathway associated with species‐specific xylem anatomy. CONTENTS Summary 196 I. INTRODUCTION 196 II. EVOLUTIONARY TRENDS IN CONDUIT DIMENSIONS 199 III. MAXIMUM XYLEM TRANSPORT IN THE PRESENCE OF CAVITATION 203 IV. INCLUDING VESSEL LENGTH IN A TRANSPORT MODEL 205 V. CONDUIT LENGTH IN MODERN TAXA: IMPLICATIONS FOR TRANSPORT 210 VI. CONCLUSIONS 212 VII. APPENDIX: ANALYTICAL SOLUTIONS AND PROOFS 213 Acknowledgements 216 References 216