Effects Of Experimental Methods On The Estimation Of Pollen Limitation

Researchers often use supplemental pollination to detect the degree of pollen limitation.If supplemental pollination on plants(flower or inflorescence)produced more fruits or seeds than the natural ones,then the reproductive output is regarded as limited by pollen.However,the appropriateness and accuracy of this approach was questioned by many researchers.Among others,Knight et al showed that nearly half of the experiments on pollen supplementation conducted on 263 species only looked into a single flower level,which does not take into account the role of resource allocation.At the same time,the pollen quality in supplemental pollination should matter.The traditional way of conducting supplemental pollination is to use pollen from other conspecific plants,namely outcross pollen.However,in terms of self-compatible plants,their mating systems are mostly a mixture of selfing and outcrossing.A large number of outcross pollen provided by artificial pollination is presumably a better source of gamete compared to self-pollen.In this case,plants might favor outcrossing in order to avoid inbreeding depression and improve the quantity and quality of offspring.So,traditional supplemental pollination experiments may confuse the role of pollen quality and quantity in assessing pollen limitation.This study explores pollen limitation in Sagittaria trifolia Linn.in two aspects,namely experimental levels(at the level of a single flower and an inflorescence)and pollen quality(self pollen and outcross pollen).My hypothesis was that supplemental pollination on a single flower level would overestimate the degree of pollen limitation than inflorescence level due to resource redistribution from other flowers that are not supplemented at the single flower level.Secondly,traditional method of applying outcross pollen would result in more offspring due to higher pollen quality,irrespective of pollen quantity,and thus also overestimate the degree of pollen limitation compared with self-pollen.The experimental results showed that the experimental level indeed caused varying degrees of pollen limitation.Specifically,supplemental pollination at a single flower level has lead to resource reallocation among flowers.However,the pollen-supplemented flower produced less seeds than the average of pollen-supplemented inflorescence,which was contrary to my hypothesis.It may be due to a reverse direction of resource allocation to the flowers that were not pollen supplemented when treatment was at a single flower level.I attributed the resource redistribution to the fact that control flowers compared with supplemented ones in the same inflorescence were in a natural state,accepted adequate pollen in a more gradual manner and might effectively avoid pollen overcrowding and possible interference.This was actually echoed by the results that natural flowers produced more seeds than supplemental ones.At the same time,the type of supplemental pollen also could affect the degree of pollen limitation.Adding outcross pollen produced more seeds than self pollen,and resulted in a higher germination rate of the seeds.Interestingly,the seed production from self-pollen supplementation did not vary significantly from flowers that were not supplemented at all,which proves that this system is limited by the quality of pollen rather than quantity.My results suggest that when it comes to making evaluations on pollen limitation,future studies should try to conduct experiments at the entire inflorescence or plant level to avoid bias introduced by resource reallocation,and also consider comparative experiments of supplementing self and outcross pollen to differentiate the causation of pollen limitation.Only when we have a more accurate and objective understanding on pollen limitation,can we put forward better scientific and reasonable countermeasures to solve ecological problems related to pollen limitation.