Numerical modeling of meteorological and topographical effects on pollen shed, dispersion, and viability

For open-pollinated species, the success of a pollen grain pollinating a receptive plant is a function of when it is shed, its dispersion in the atmosphere, and the chance of germination upon deposition. Most studies of pollen dispersion have focused on a source and surrounding fields in relatively flat terrain, but dispersion to distances over 20 km has been reported in regions of complex terrain. For this reason, a more detailed examination of atmospheric dispersion over long distances and the components of dispersion -- shed, transport, and successful pollination -- was undertaken. This dissertation uses numerical models to predict these components in order to improve simulations of dispersion over long distances. Chapter One provides an overview of work previously done in pollen dispersion. Chapter Two describes a model that uses empirical relationships between pollen release and meteorological variables to predict diurnal variations in maize pollen shed. The model captured the general pattern of shed and predicted the time of peak shed on most days. In Chapter Three, a meteorological model was combined with a Lagrangian dispersion model to predict bentgrass pollen dispersion over complex terrain. The models predicted pollen deposition in locations where outcross was previously reported over 20 km from the source. However, pollen was not predicted to be deposited on upwind slopes of mountains or in valleys more than a few kilometers from the source. Chapter Four examines how thermally-generated updrafts resulting from surface sensible heat flux and mechanically-generated updrafts associated with complex topography affect dispersion. We determined the gaps in pollen dispersion found in Chapter Three resulted from additional lift created by convergence on the windward side of mountains and thermally-induced flow that traveled up valley walls and mountain slopes. Chapter Five examines how maize pollen viability is affected when pollen grains are lifted higher in the atmospheric boundary layer. Viability was predicted to be as much as 30% higher when the atmospheric conditions along a pollen grain's trajectory are used compared to estimates of viability using conditions at the Earth's surface. Chapter Six summarizes the dissertation. The results of this work are expected to improve future models of pollen dispersion, leading to more accurate predictions of outcrossing. Also, the methods described here are not limited to maize or bentgrass pollen, but can be modified for applications to other biological particles that rely on atmospheric conditions for dispersion.