Holland J Nathaniel, Chamberlain Scott A
Department of Ecology and Evolutionary Biology, Rice University, MS-170, 6100 South Main Street, Houston, Texas 77005, USA.
Ecology. 2007 Mar;88(3):706-15. doi: 10.1890/06-1283.
Central to the ecology and evolution of a broad range of plants is understanding why they routinely have submaximal reproduction manifested as low seed : ovule and fruit : flower ratios. We know much less about the processes responsible for low seed : ovule ratios than we do for fruit : flower ratios. Current hypotheses for low seed : ovule ratios are largely drawn from those for fruit : flower ratios, including proximate (ecological) causes of pollen limitation, resource limitation, and pollen quality, as well as the ultimate (evolutionary) hypothesis of "bet hedging" on stochastic pollination. Yet, such mechanisms operating on fruit : flower ratios at the whole-plant level may not best explain low seed : ovule ratios at the individual-flower level. We tested each of these proximate and ultimate causes for low seed : ovule ratios using the specialized pollination mutualism between senita cacti (Pachycereus schottii) and senita moths (Upiga virescens). Seed : ovule ratios were consistently low (approximately 0.61). Such excess ovule production by senita likely has a strong genetic component given the significant differences among plants in ovule number and the consistency in ovule production by plants within and among flowering seasons. Excess ovule production and low seed : ovule ratios could not be explained by pollen limitation, resource limitation, pollen quality, or bet hedging. Nevertheless, phenotypic selection analyses did show significant selection gradients for increased ovule number, suggesting that other evolutionary processes may be responsible for excess ovule production and low seed : ovule ratios. In contrast, low fruit : flower ratios at the whole-plant level were explained by an apparent equilibrium between pollen and resource limitation. Thus, mechanisms responsible for low fruit : flower ratios at the whole-plant level are not necessarily in accord with those of low seed : ovule ratios at the individual-flower level. This suggests that we may need to adopt a more pluralistic approach to seed : ovule ratios and consider alternative hypotheses, including a greater array of proximate and ultimate causes. Initial results of this study suggest that floral allometry, selection on correlated floral traits, stigma clogging with pollen grains, and style clogging with pollen tubes may provide promising avenues for understanding low seed : ovule ratios.
对于众多植物的生态与进化而言，核心在于理解为何它们常常表现出繁殖水平未达最大值，具体体现为低种子：胚珠比以及低果实：花比。相较于果实：花比，我们对导致低种子：胚珠比的过程了解得更少。当前关于低种子：胚珠比的假说大多借鉴了关于果实：花比的假说，包括花粉限制、资源限制和花粉质量等直接（生态）原因，以及随机授粉时“风险分摊”的终极（进化）假说。然而，在整株植物水平上作用于果实：花比的此类机制，可能并非个体花朵水平上低种子：胚珠比的最佳解释。我们利用仙人球（Pachycereus schottii）与仙人球蛾（Upiga virescens）之间的特殊传粉互利关系，对导致低种子：胚珠比的这些直接和终极原因逐一进行了测试。种子：胚珠比始终较低（约为0.61）。鉴于不同植株间胚珠数量存在显著差异，且同一植株在不同花期以及不同植株之间胚珠产量具有一致性，仙人球这种过量的胚珠产生可能具有很强的遗传成分。花粉限制、资源限制、花粉质量或风险分摊均无法解释过量的胚珠产生以及低种子：胚珠比。尽管如此，表型选择分析确实显示出胚珠数量增加存在显著的选择梯度，这表明其他进化过程可能导致了过量的胚珠产生以及低种子：胚珠比。相比之下，整株植物水平上低果实：花比可由花粉和资源限制之间的明显平衡来解释。因此，整株植物水平上导致低果实：花比的机制未必与个体花朵水平上低种子：胚珠比的机制一致。这表明我们可能需要对种子：胚珠比采取更为多元的方法，并考虑其他假说，包括更多种类的直接和终极原因。本研究的初步结果表明，花部异速生长、对相关花部性状的选择、花粉粒堵塞柱头以及花粉管堵塞花柱可能为理解低种子：胚珠比提供有前景的途径。
