The responses of clones of a range of willow species to waterlogging and total submersion were compared in glasshouse experiments using specially constructed perspex tanks. Up to 13 weeks of waterlogging had little effect on form and growth rates of above-ground parts of two clones of Salix caprea L. and one clone of S. cinerea L. A second clone of S. cinerea responded differently to the other three clones by significantly increasing its leaf area and stem dry weight, suggesting that it is likely to be more tolerant of waterlogging.
Plant parts below the waterline in the waterlogging treatment responded very differently to the same parts on control plants. All waterlogged plants developed hypertrophied lenticels within four weeks, followed by significant increases in stem diameter. In each clone most of the pre-existing root system died by the end of the waterlogging treatment. All clones formed adventitious roots which varied considerably between the species and to a lesser extent between clones within each species. In S. cinerea many short, fine, much-branched roots grew vertically upwards until they reached the water surface. In S. cinerea clone 90 they originated from both the pre-existing woody roots and from the hypertrophied lenticels on the submerged portions of the stems. In S. cinerea clone 12 they developed only from the hypertrophied lenticels. New roots produced by waterlogged S. caprea clones were long, sparsely branched and floated on or near the surface of the water. As in S. cinerea clone 90 they originated from both the pre-existing woody roots and the submerged hypertrophied lenticels. Whereas roots of the control plants of all clones were mostly ectomycorrhizal, those of the waterlogged plants had no mycorrhizae. In a second waterlogging experiment S. × sericans, S. viminalis and S. fragilis produced adventitious roots more like those of S. caprea than S. cinerea. Plants of all clones removed from the waterlogging treatment after 16 days and kept with their soil at field capacity thereafter produced many new roots from their old root balls. After 32 days waterlogging no such recovery occurred.
Total submersion caused much more severe damage to the aerial parts of all tested species than waterlogging. After eight days submersion the shoot tips of S. fragilis were beginning to rot and by day 25 those of S. × sericans and S. viminalis were rotting too. Subsequently shoots developed from axillary buds which remained unaffected by rot throughout the experiment. In S. cinerea shoot growth ceased after eight days and terminal buds were formed which remained dormant but healthy throughout the experiment. Submersion caused progressive loss of pre-existing leaves in all species. Some newly formed leaves of each species developed epinastic deformities after > 16 days of submersion. All species except S. cinerea showed significant reductions in root dry weight after 32 days of submersion.
In a second submersion experiment responses of three clones each of S. purpurea and S. viminalis were compared. Shoot growth was reduced significantly in four of the six clones after 14 days. No further reductions occurred in the next 14 days of submersion. Root dry weight also decreased, but it took 28 days of submersion to produce the greatest effect. The faster growing clones of each species were more adversely affected by submersion in terms of both reduced shoot growth and root growth than those which grew more slowly.
Considerable intraspecific variability in responses to waterlogging and submersion have been shown in these experiments by clones not selected for their variation in flooding tolerance. This suggests that selection and breeding could lead to the development of flood-tolerant cultivars of species such as S. cinerea and S. caprea which are not noted for their ability to withstand inundation.