A greenhouse pot experiment was conducted to elucidate the pattern of
decreasing seed set and seed dry weight found from the basal to the upper
spikelet in Lolium perenne L. ears and—very
markedly—acropetally within a spikelet. The changes in fresh and dry
weight of the seed after anthesis were monitored for 12 positions within the
ear. To determine the duration of seed growth, the spatial and temporal
patterns of flowering and ripening along the ear were assessed. The proximal
florets of the central spikelets flowered first, and the upper spikelets in
the ear flowered before the basal spikelets, with flowering proceeding
acropetally within a spikelet. The upper spikelets ripened earlier than the
basal ones, and the seeds within a spikelet ripened simultaneously. These
patterns of flowering and ripening along the ear resulted in different
durations of seed growth along the ear. Thus from the proximal to the distal
seed within a spikelet the duration of growth decreased by 26%, the
rate of growth fell by 48% and seed set and seed dry weight also fell
sharply. Of the total variation in maximum, final seed dry weight within the
ear, 4% could be attributed to the difference in seed dry weight
between spikelets and 89% to differences in seed dry weight within
spikelets. Differences in duration and rate of seed growth and seed set were
also much smaller between spikelets than within spikelets. About 60% of
the differences in seed dry weight along the ear were attributable to
differences in growth rate and about 30% to differences in the duration
of seed growth. The relative growth rate did not differ between seeds in
different positions. The main factor determining seed growth rate was the dry
weight of the ovule at anthesis, which declined sharply from the proximal to
the distal floret within a spikelet. Analysis of the fructose, glucose,
sucrose and starch concentrations showed that seed growth was not limited by
the availability of sucrose nor by the ability of the seeds to convert sucrose
into starch. These conclusions were supported by results from a second
greenhouse experiment in which all ovules in 50% of the spikelets were
removed at anthesis. Dry weight of the remaining seeds increased by only
15%.
Self-incompatibility in ryegrass VI. Self seed-set and incompatibility genotype in Lolium perenne L