The objective of the research was to evaluate the effects on pollen of
a selection conducted in maize (Zed mays L.) for the reaction to absc
isic acid (ABA) applied to silks. Two single crosses were used as sour
ces. Selected populations were developed by sib-pollinating ears treat
ed with ABA (10(-3) M), while non-selected populations were obtained b
y sib-pollinating untreated ears. The analysis of ABA concentration of
the treated silks indicated that the selective agent was in part able
to enter the silks and to move inward. The pollen performances of sel
ected and non-selected populations were then compared for evaluating t
he selection responses. Selected populations were superior for the pol
len reaction in vivo evaluated as number of kernels earl obtained by p
ollinating silks of the parental F(1)s sprayed with ABA. Selected popu
lations were also superior for pollen tube length in vitro without ABA
(i.e., for a trait reflecting the per se pollen vigour). In contrast,
selected and non-selected populations did not significantly differ fo
r the pollen reaction in vitro evaluated as grain germination and tube
length in the presence of ABA. Results indicate that the selection pr
ocedure was effective in improving both the pollen reaction in vivo to
ABA applied to silks and the per se pollen vigour.