Lowe, AndrewGardner, Michael GeorgeOttewell, Kym MarieBreed, Martin F.2014-02-052014-02-052013http://hdl.handle.net/2440/81911Fragmented tree populations are not expected to be as susceptible to small population paradigm effects (e.g. genetic drift) that generally dominate conservation genetics and restoration as many other taxa. The reasons for this are that trees tend to (1) undergo regular far-reaching gene flow, even in fragmented landscapes, (2) have many overlapping generations and (3) be long lived relative to when most habitat fragmentation occurred. These traits result in tree populations having great genetic inertia and thus they tend to maintain genetic diversity (as measured by numbers of alleles) despite significant habitat fragmentation. However, trees are not resistant to changes in the genetic diversity of their progeny (as measured by observed heterozygosity) as a result of habitat fragmentation. Habitat fragmentation can alter the mating patterns of individual trees by changing pollination dynamics (e.g. levels of selfing, pollen diversity) and these mating patterns directly influence the genetic makeup of progeny. Tree progeny are predicted to be particularly sensitive to mating pattern changes of their maternal plant since most tree species predominantly outcross, leading them to accumulate more genetic load than plants that regularly selfpollinate. Consequently, reduced pollen diversity is likely to reduce pollen competition or reduce heterosis effects within the observed generation; more selfing is expected to increase inbreeding depression. Furthermore, for trees, these patterns remain to be examined in an experimental or quantitative way. Furthermore, discussions of these trends have often relied on theoretical arguments, rather than empirical data, paving the way for experimental investigations. Consequently, it was the primary goal of this thesis to examine some of these gaps in knowledge in an experimental and quantitative way. Specifically the aims of this thesis were to: 1. Examine and quantify the impact of fragmentation and tree density on mating patterns, and how this may vary with pollinators of differing mobility 2. Determine the theoretical expectations and perform empirical tests of mating pattern-fitness relationships in trees 3. Explore the plant genetic resource management implications that arise from the observations in aims 1 and 2 In general the results showed that stands of trees that have experienced habitat fragmentation or are present in lower densities express a quantifiable negative shift in their mating patterns (i.e. they tend to self more and receive less pollen diversity). More mobile pollinators appear to buffer trees from these negative shifts in their mating patterns. I present a theoretical guide to the mating pattern-fitness relationships in terms of habitat fragmentation. I found that an increase in selfing and a decrease in pollen diversity are both important factors that could be quantified as impacting on fitness of established seedlings. Taken together, these findings suggest that seeds collected from larger, less fragmented and higher density stands have higher fitness. Consequently, collecting seeds from these stands should lead to better outcomes of ex situ and in situ conservation, restoration and revegetation plantings.plant mating systems; gene flow; inbreeding; forests; habitat fragmentation; conservation biology; garden experiments; pollen competitionThesis20131009124252