Epsilon Open Archive

Abstract

Willow (Salix spp.) grown on arable land as short-rotation coppice (SRC) produces renewable energy in the form of woody biomass. This perennial crop has a high ratio of energy output to input and a good environmental profile. However, weed control is mostly dependent on herbicide use. Therefore, this thesis examined the possibility to further improve the environmental profile of willow SRC by omitting the use of herbicides during establishment. If genetic variation in willow competitiveness to weeds exists, more weed-competitive cultivars might be bred. However, in a study performed at three different sites in southern Sweden, only small differences were found between 12 clones tested for their ability to compete with weeds. Depending on site, weeds reduced stem biomass yield by between 68 and 94% after the first harvest cycle and increased plant mortality at all sites. The practice of cutting the first-year shoots either reduced or did not affect the ability of the willow plants to compete with weeds. Hence, this measure should be omitted provided this is compatible with other management actions. A study on the efficiency and economic returns from four different non-chemical weed control methods during willow establishment of two different cultivars showed that it is possible to establish an agriculturally and economically viable willow plantation without the use of herbicides. The most promising non-chemical weed control method involved repeated passes with a row crop cultivator equipped with torsion weeders, while the least promising method was a living clover cover crop.
Yield data were obtained from the weed competition study for the willow SRC clones when subjected to thorough weeding. Cultivars Sven and Tordis were found to be among the highest yielding at all three sites, although site x clone interactions were found. However, these two clones did not yield significantly more than two more recently bred clones, Klara and Linnea, at any site.
Biomass estimates from destructive and non-destructive methods have been shown to differ and the magnitude of these differences may depend on clone. A study with six different clones showed that assumptions regarding harvest height and dry matter content of clones might explain part of these differences.