In order to study the adaptive significance of variation in Bellis perennis L., an experiment was carried out in which plants from 19 populations – drawn from four habitat groups (I, lawns; II, grazed areas; III, seasonally mown and/or grazed areas; and IV, areas not seasonally reduced in height by grazing and/or mowing) - were transplanted into a lawn subjected to regular mowing. Vegetative growth (as revealed by increment of plant width, dry wt and number of rosettes) and reproduction (as estimated by the number of capitula which survived mowing and produced fruit) were investigated over a 17 month period. As most plants in the experiment grew successfully and approximately 60 % of the 95 individuals (including plants from all four habitat groups) produced fruiting capitula, there was no evidence of distinct ‘races’ of B. perennis in the regularly defoliated habitats of lawn and pasture studied. However, the more successful plants, in terms of numbers of fruiting capitula, were found in samples collected from lawns and Port Meadow: pasture. Four individuals (three from lawns and one from Port Meadow) yielded approximately 50% of all the capitula which fruited. Within ‘successful’ populations there were plants which produced few or no fruiting capitula. Various hypotheses are presented to account for population variation.
Estimates of coefficients of selection, comparing different features of growth and reproduction, may yield different values. For example, even though different individuals are apparently equally successful in vegetative growth under experimental conditions, i.e. with coefficients of selection of 0, a comparison of reproductive success for the same plants may give a value of 10.