Investigating the Pollination Process in Wind-Pollinated Plants
Date
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Abstract
Despite its frequent evolution in the flowering plants, wind pollination has long been considered inefficient. Key traits of the wind pollination syndrome, high pollen production and low ovule number, have been thought to compensate for high pollen losses and a low probability of pollen capture. However, these traits may instead reflect selection by various processes to facilitate pollination, including the aerodynamics of wind pollination, selection related to sex system, correlated selection on pollen size and number, and the influence of local pollination conditions. Broadly, my thesis explores influences on wind pollination from both male and female perspectives, using field-collected data on pollen production, size and capture. Pollen transfer efficiency, a measure of pollen transport loss, was similar between 19 wind-pollinated species and values previously reported for animal-pollinated species, suggesting wind pollination is not less efficient than animal pollination. Pollen size and number were negatively correlated in 6 of 23 species, indicating potential size-number tradeoffs. In all species, pollen size was significantly less variable than pollen number, suggesting pollen size is under stronger stabilizing selection. Pollen size but not production had a significant phylogenetic component, potentially because selection on the aerodynamics of pollen size is likely closely tied to plant architecture. Hermaphroditic species produced significantly more pollen per anther than monoecious, but not dioecious species, though this may be related to mating system. Average stigmatic pollen loads exceeded ovule number for all 23 wind-pollinated species, suggesting that low ovule number resulting from a low probability of pollen receipt is unlikely. Instead, low ovule number may allow for increased pollen competition or represent a strategy to better sample the airstream by producing many inexpensive flowers. Stigma length and the timing of sampling within the flowering season strongly influenced pollen loads. Hermaphroditic species captured more pollen than monoecious or dioecious species; however, much of this pollen may be self-pollen. Together, my results demonstrate the context-dependence of pollen receipt in wind pollination and challenge the assumption that wind pollination is inefficient, suggesting that high pollen production and low ovule number are unlikely to arise simply from a low probability of successful pollination.

