Sampling techniques for measurement of fibre length in eucalyptus species

Summary Individual fibre lengths were measured on the projected images of macerated tissue samples of seven Eucalyptus species grown in a replicated species trial in Zambia. Analyses of variance and variance components were used to study variation between species, replications, distances from the pith, radial positions, microscope slides and individual fibres. The overall mean was 0.93 mm and little variation was attributable to the main factors (fixed effects); two-thircs of total variation was due to differences between fibres within samples. Variance components for random factors were used to compare different sampling strategies; the size of acceptable difference for hardwoods (0.10 ... 0.14 mm) is considerably less than that for conifers and samples of 80 ... 200 fibres may be necessary. Differences between microscope slides were unimportant and two or more slides would be necessary only to facilitate location of the large numbers of fibres or to provide a check on operators.Laboratory work was carried out at Tree Improvement Research Centre, Kitwe, Zambia.