Tables
Table 1: Total selection on (a) leaf-out day and (b) first
flowering day, and direct linear, nonlinear and correlational selection
for both traits (c). Model
estimates with 95% bias-corrected and accelerated (BCa) bootstrap
intervals. Estimates with BCa intervals that do not overlap zero are inbold . Fitness (the response variable, the number of intact
seeds) was relativized and leaf-out day, first flowering day, and plant
size (above-ground volume) were standardized to 0 mean and unit standard
deviation before analysis. Plant size was ln-transformed before
standardization. The nonlinear (quadratic) model estimates represent ½
of the magnitude of nonlinear selection (Stinchcombe et al. 2008)