Use of clones increases the power of physiological experiments on coastal Douglas‐fir

The statistical benefit of reduced variability in experiments with clones was quantified with cold hardy Douglas‐fir ( Pseudotsuga menziesii var. menziesii [Mirb.] Franco) ramets from six clones from a single full‐sib family and a check group of open‐pollinated seedlings from a bulk seed collection from the same geographic seed source. All groups were cold deacclimated under controlled conditions. The variability in response for physiological attributes (cold hardiness, root growth potential, days to bud‐break) was compared within and among groups by using statistical power analyses. Sample sizes required to detect significant differences of varying magnitude between two hypothetical treatment means were calculated. Differences among clones in the average response exhibited for each physiological attribute were large, but within‐clone variability was low, relative to the check group of seedlings. Selection of plant material for a hypothetical experiment from this population of several identifiable clones would have consistently resulted in an experiment with greater power than an experiment using these ramets without the clone identities. From a statistical perspective, the best approach to reduce the number of replicates needed to detect treatment differences was selection of experimental plant material from a single clone, especially with prior screening for the most homogeneous clone for the physiological attributes and time periods of interest. However, from a biological perspective, use of a single clone should be approached with caution because of the lack of representation of natural population variability and the possible inability to broadly apply experimental results.