The ecology and evolution of wind pollination

The evolution of wind pollination (anemophily) has occurred at least 65 times in the flowering plants and over 10% of angiosperm species are wind pollinated. However the pollination and mating of anemophily species is poorly understood, particularly in comparison with animal-pollinated species. My thesis employs a range of approaches and tools to examine the evolution and ecology of wind pollination. These include comparative analyses, theoretical modeling, field and glasshouse experiments, the use of genetic markers and quantitative genetics. Experimental studies on diverse taxa were used to address questions concerned with the efficacy of outcrossing mechanisms, the ecological and demographic context of pollination and mating, and the plasticity of sex allocation. Comparative analyses indicated that wind pollination is correlated with unisexual flowers, reduced ovule number, small unshowy flowers, an absence of nectar, and open habitats. These analyses also demonstrated that anemophily originates more often in lineages with unisexual flowers. This suggests that wind pollination evolves in diclinous taxa as a mechanism of reproductive assurance because autonomous selfing is mechanically precluded. Empirical data on stigmatic pollen loads in 19 anemophilous species challenge the widespread assumption that anemophilous plants commonly have uniovulate flowers because they capture few pollen grains. Further, a model based on floral costs and the aerodynamics of pollen capture demonstrated that when flowers are inexpensive it is optimal to produce many flowers each with few ovules, because this allows more efficient sampling of the airstream. Manipulative field experiments on seven Carex species indicated that neither monoecy nor protogyny, two putative outcrossing mechanisms, are effective at limiting selfing. Based on these results I suggest that geitonogamy can provide reproductive assurance in anemophilous species with unisexual flowers. Field experiments and the application of sex-specific markers in Rumex nivalis revealed that the local neighbourhood of maternal plants affects pollination intensity and progeny sex ratios. Finally, I demonstrated that plant density in Ambrosia artemisiifolia affects stigmatic pollen loads but not outcrossing rates. Through a quantitative genetics experiment in A. artemisiifolia, I detected significant genetic variation for plasticity in sex allocation, potentially enabling adaptive adjustment of sex allocation to local environmental conditions.Ph