Ma. Aizen et Kb. Searcy, SELECTIVE FRUIT FILLING IN RELATION TO POLLEN LOAD SIZE IN ALSTROEMERIA-AUREA (ALSTROEMERIACEAE), Sexual plant reproduction, 11(3), 1998, pp. 166-170
We conducted an experiment in a natural population of Alstroemeria aur
ea, a clonal perennial, to determine (1) if reproduction was resource
limited, and (2) if fruits would be selectively filled based on differ
ences in pollination intensity when pollen loads were adequate for ful
l seed set. Under these conditions, differences in pollination intensi
ty are unlikely to affect seed number, but could affect seed quality,
providing an interesting test of the gametophytic competition hypothes
is. To test for resource limitation, percent fruit maturation, number
of seeds per fruit and average seed weight were compared to paired con
trols for ramets in which all but one fruit was removed. To test the e
ffect of pollination intensity on selective resource allocation, three
types of pollination treatments were performed: (1) all flowers of th
e single inflorescence received a low pollen load, (2) all flowers rec
eived a high pollen load, (3) alternate flowers of the inflorescence r
eceived either a high or a low pollen load. We determined the percenta
ge of fruit that reached maturity, counted the number of seeds and ovu
les and calculated the average seed weight for all capsules in each tr
eatment. Resources appeared to limit reproduction in this population s
ince seed number and weight were significantly higher than in controls
when competing capsules were removed. At the whole ramet level, a fou
r fold difference in pollen loads had no significant effect on any of
the parameters measured. However, when pollination intensity varied wi
thin an inflorescence, the number of seeds per fruit increased by abou
t 10% in flowers that received the higher pollen load. We observed the
same trend in each of 2 years, but the increase was significant in on
ly 1 year. The differences,, although not great, were only slightly sm
aller than when all competing fruits were removed, and were consistent
with selective re source allocation based on pollination intensity in
dependent of seed set.