Artificial selection on trichome number in Brassica rapa.

We examined genetic variation for trichome production in a rapid-cycling population of Brassica rapa by conducting a selection experiment and by growing progeny from maternal seed families. Data from the maternal families were also used to estimate the genetic correlations between trichome number and (1) number of days to first flower and (2) flower production. For seven generations, 10% of the plants were selected from low, high and control lines with 100 individuals per line per generation. The number of trichomes on the right edge of the first leaf was 20.8 ± 13.4 (mean + SD; n=100) in the base population, and had by the final generation reached 93.9 ± 28.7 (n=100) in the high line and 0.9 ±2.6 (n=100) in the low line. Control line plants of the seventh generation did not differ significantly from base population plants in number of trichomes on the edge. The realized heritability of trichome number on the edge was 0.38 (based on the regression of cumulative response on cumulative selection differential). The divergence between lines in trichome production on the edge of the first leaf was associated with a divergence in trichome number on the petiole and on the top of the same leaf, and on the edge, top and petiole of the third leaf. The increase in trichome production in the high line was further associated with a significant delay in flowering time relative to the control and low lines. The estimated heritabilities of trichome number on the edge and the petiole of the first leaf and of days to first flower were not significantly different from 1.0 (based on the among maternal family component of the total variance). Trichome number on the edge showed a significant genetic correlation with trichome number on the petiole of the first leaf (r=0.80), and with number of days to first flower (r=0.31), but not with total flower production (r=0.17; n=83). The substantial genetic variation for trichome production in B. rapa is of potential value for breeding purposes and also makes B. rapa a convenient system for the study of plant-herbivore interactions.