One distinctive aspect of plants is that their life cycle shows ‘alternation of generations’, associated with development of multicellular bodies in both diploid (sporophyte) and haploid (gametophyte) stages. In flowering plants, the short-lived haploid gametophyte can be either male or female, producing gametes for double fertilization, resulting in the progeny seed. The male gametophyte, primarily encountered as pollen and carrying sperm cells, can be under strong selective pressure during its brief period of independence, competing with large numbers of other male gametophytes, particularly during pollen tube growth at the recipient flower. Understanding the biological processes underlying pollen function is not only important for a basic understanding of plant reproduction, but is also relevant to pollen's critical role in agriculture - e.g., the production of seeds.

The Fowler research group, alongside key collaborators, is using traditional genetics, supported by computational and quantitative approaches, to investigate pollen biology in maize. The simple genetics of the haploid male gametophyte results in a key characteristic: mutants that affect pollen function deviate from Mendelian inheritance, which can easily assessed on the maize ear using linked kernel phenotypes. Our recent results have used an ear/kernel phenotyping platform (EarVision), combined with a large population of GFP-marked insertion alleles, to dramatically scale up our ability to identify genes contributing to pollen fitness, and measure their quantitative effect. By combining phenotyping outputs with statistical modeling, we have also tested the hypothesis that alleles with reduced pollen fitness can influence the spatial distribution of progeny kernels on an ear. Finally, our pollen fitness dataset has enabled successful development of machine-learning models that predict gene-specific phenotypic effects from genome-scale data. The modeling results are pointing towards expression profiling data as a particularly information-rich feature for identifying genes that contribute to pollen fitness.